Bacteriophages: Biology, Technology, Therapy 9783319419855, 9783319419862, 9783319419879

Citation preview

David R. Harper · Stephen T. Abedon Benjamin H. Burrowes Malcolm L. McConville Editors

Bacteriophages Biology, Technology, Therapy

Bacteriophages

David R. Harper • Stephen T. Abedon • Benjamin H. Burrowes • Malcolm L. McConville Editors

Bacteriophages Biology, Technology, Therapy

With 173 Figures and 44 Tables

Editors David R. Harper Evolution Biotechnologies Bedfordshire, UK Benjamin H. Burrowes Evolution Biotechnologies Georgetown, TX, USA

Stephen T. Abedon Department of Microbiology The Ohio State University Mansfield, OH, USA Malcolm L. McConville Evolution Biotechnologies Bedfordshire, UK

ISBN 978-3-319-41985-5 ISBN 978-3-319-41986-2 (eBook) ISBN 978-3-319-41987-9 (print and electronic bundle) https://doi.org/10.1007/978-3-319-41986-2 © Springer Nature Switzerland AG 2021 All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG. The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

As the ongoing crisis of antimicrobial resistance (AMR) threatens to bring an end to the era of routine control of bacterial diseases, there is great interest in developing other approaches to controlling such infections. One of the oldest of these, the use of bacteriophages (viruses that can target and destroy bacteria) as therapeutic agents, is experiencing a resurgence of interest and is now considered a promising approach to countering AMR. First developed 100 years ago, this approach, known as phage therapy, was set aside in Western Europe and the USA when the use of chemical antibiotics became widespread. Now, the pressing need for new ways to control such resistant bacteria is resulting in very real and rapid progress in developing phage therapy, alongside a range of technologies based around bacteriophages, in medicine and elsewhere. Bacteriophages: Biology, Technology, Therapy is intended to cover all major, current aspects of work with bacteriophages, from their basic biology to clinical trials of phage therapeutics and from early history to nanotechnology. In so doing, the intention is to provide a single, readily citable source covering the biology of bacteriophages and bacteriophage infection, their use across a wide range of technologies, and their evolving use as therapeutic agents.

v

Contents

Volume 1 Part I Introduction to Bacteriophages: Biology, Technology, Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

...............................

3

..............................

17

Structure and Function of Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . Marta Sanz-Gaitero, Mateo Seoane-Blanco, and Mark J. van Raaij

19

......................

93

Temperate Phages, Prophages, and Lysogeny . . . . . . . . . . . . . . . . . . . . Joanna Łoś, Sylwia Zielińska, Anna Krajewska, Zalewska Michalina, Aleksandra Małachowska, Katarzyna Kwaśnicka, and Marcin Łoś

119

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction . . . . Christine L. Schneider

151

Genetics and Genomics of Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . Aidan Casey, Aidan Coffey, and Olivia McAuliffe

193

........................

219

Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Michael A. Brockhurst, Britt Koskella, and Quan-Guo Zhang

231

Introduction to Bacteriophages David R. Harper Part II

Bacteriophage Biology

Adsorption: Phage Acquisition of Bacteria John J. Dennehy and Stephen T. Abedon

Bacteriophage Discovery and Genomics Graham F. Hatfull

Bacteriophage Ecology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . John J. Dennehy and Stephen T. Abedon

253

vii

viii

Contents

Bacteriophage Pharmacology and Immunology . . . . . . . . . . . . . . . . . . Krystyna Dąbrowska, Andrzej Górski, and Stephen T. Abedon

295

Phage Infection and Lysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . John J. Dennehy and Stephen T. Abedon

341

Part III

385

History of Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . . . .

The Discovery of Bacteriophages and the Historical Context . . . . . . . . William C. Summers

387

........

401

Early Therapeutic and Prophylactic Uses of Bacteriophages Nina Chanishvili and Zemphira Alavidze

Volume 2 Part IV

Bacteriophage Technology . . . . . . . . . . . . . . . . . . . . . . . . .

431

Isolation of Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Frits van Charante, Dominique Holtappels, Bob Blasdel, and Benjamin H. Burrowes

433

Bacteriophage Use in Molecular Biology and Biotechnology . . . . . . . . . Nathan Brown and Chris Cox

465

Detection of Bacteriophages: Phage Plaques . . . . . . . . . . . . . . . . . . . . . Stephen T. Abedon

507

Detection of Bacteriophages: Statistical Aspects of Plaque Assay . . . . . Stephen T. Abedon and Tena I. Katsaounis

539

Detection of Bacteriophages: Electron Microscopy and Visualisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . David M. Belnap

561

Detection of Bacteriophages: Sequence-Based Systems . . . . . . . . . . . . . Siân V. Owen, Blanca M. Perez-Sepulveda, and Evelien M. Adriaenssens

621

Novel Approaches for Detection of Bacteriophage . . . . . . . . . . . . . . . . . Carrie L. Pierce, Jon C. Rees, and John R. Barr

645

Bacteriophages in Nanotechnology: History and Future . . . . . . . . . . . . Paul Hyman and Jenna Denyes

657

..................

689

Bacteriophage Manufacturing: From Early Twentieth-Century Processes to Current GMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Krzysztof Regulski, Patrick Champion-Arnaud, and Jérôme Gabard

699

The Selection and Optimization of Phage Hosts Jason J. Gill

Contents

ix

Intellectual Property Issues for Bacteriophages . . . . . . . . . . . . . . . . . . . Martin R. MacLean and David R. Harper

731

Bacteriophage as Biocontrol Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . David R. Harper

751

Part V Agriculture, Food, and Environmental Use of Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bacteriophages as Bio-sanitizers in Food Production and Healthcare Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sudhakar Bhandare and Lawrence Goodridge

767

769

Biofilm Applications of Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . . . Catarina Milho, Maria Daniela Silva, Sanna Sillankorva, and David R. Harper

789

...................

823

Crop Use of Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jeffrey B. Jones, Antonet M. Svircev, and Aleksa Ž. Obradović

839

Food Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lars Fieseler and Steven Hagens

857

Bacteriophage Utilization in Animal Hygiene . . . . . . . . . . . . . . . . . . . . Sarah Klopatek, Todd R. Callaway, Tryon Wickersham, T. G. Sheridan, and D. J. Nisbet

891

Part VI

919

Industrial Processes Involving Bacteriophages Marcin Łoś

Therapeutic Use of Bacteriophages

..................

Current Updates from the Long-Standing Phage Research Centers in Georgia, Poland, and Russia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ryszard Międzybrodzki, Naomi Hoyle, Fikria Zhvaniya, Marzanna Łusiak-Szelachowska, Beata Weber-Dąbrowska, Małgorzata Łobocka, Jan Borysowski, Zemphira Alavidze, Elizabeth Kutter, Andrzej Górski, and Lasha Gogokhia

921

................

953

Enzybiotics: Endolysins and Bacteriocins . . . . . . . . . . . . . . . . . . . . . . . Ryan D. Heselpoth, Steven M. Swift, Sara B. Linden, Michael S. Mitchell, and Daniel C. Nelson

989

The Use of Bacteriophages in Veterinary Therapy Robert J. Atterbury and Paul A. Barrow

Phage Display Technology and the Development of Phage-Based Vaccines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1031 Joe A. Fralick and Jason Clark

x

Contents

Phage Therapy Collaboration and Compassionate Use . . . . . . . . . . . . . 1069 Jessica C. Sacher and Jan Zheng Clinical Trials of Bacteriophage Therapeutics . . . . . . . . . . . . . . . . . . . . 1099 Shawna McCallin and Harald Brüssow Selection of Disease Targets for Phage Therapy David R. Harper

. . . . . . . . . . . . . . . . . . 1129

Regulatory Considerations for Bacteriophage Therapy Products: USA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1151 Roger D. Plaut and Scott Stibitz Regulatory Aspects of the Therapeutic Use of Bacteriophages: Europe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1165 Eric Pelfrene, Zigmars Sebris, and Marco Cavaleri Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1179

About the Editors

Dr. David R. Harper has a scientific background in virology and microbiology and is the CEO of Evolution Biotechnologies. He was the founder, CEO, and CSO of Biocontrol Limited and CSO of AmpliPhi Biosciences and managed the first modern regulated clinical trial to demonstrate the efficacy of bacteriophages. He is a member of the Alternatives to Antibiotics panel, a collaboration between the Wellcome Trust and the UK Department of Health. Dr. Stephen T. Abedon has been a member of the faculty of the Ohio State University, Department of Microbiology, for over 25 years. He has been studying bacteriophages for 35-plus years and has over 100 phage-based publications including 9 monographs or equivalents on phages which he has edited (1), co-edited (7), or single authored (1), along with 3 more edited volumes currently in progress. Additional information and resources can be found at abedon.phage.org. Dr. Benjamin H. Burrowes received his Ph.D. in medical microbiology from Texas Tech University Health Sciences Center in 2010, where his research focus was the development of therapeutic bacteriophages with extended host range using in vitro evolution. Upon leaving Texas, Dr. Burrowes worked at Biocontrol Inc. (now AmpliPhi Biosciences Corporation) developing multiple human phage therapy preparations targeting several major bacterial pathogens. Ben left GeneWEAVE some time ago and now works as Senior Consultant at Evolution Biotechnologies, Georgetown, TX, USA, working on novel phage therapeutics, and as a Senior Scientist at the Center for Phage Technology, Texas A&M University, College Station, TX, USA. Dr. Malcolm L. McConville is a microbiologist with over 30 years industrial experience. Having gained a B.Sc. (Cardiff University, UK) and Ph.D. (University of Reading, UK), he has worked in the contract biomedical research industry and for AmpliPhi Biosciences Corporation (formerly Biocontrol Ltd) on the commercial development of bacteriophage therapy products. He is currently Trials Director at Evolution Biotechnologies.

xi

Contributors

Stephen T. Abedon Department of Microbiology, The Ohio State University, Mansfield, OH, USA Evelien M. Adriaenssens Microbiology Research Group, Institute of Integrative Biology, University of Liverpool, Liverpool, UK Zemphira Alavidze Phage Therapy Center, Tbilisi, Georgia Eliava BioPreparations, Tbilisi, Georgia Robert J. Atterbury School of Veterinary Medicine and Science, University of Nottingham, Leicestershire, UK John R. Barr Division of Laboratory Sciences, National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, GA, USA Paul A. Barrow School of Veterinary Medicine and Science, University of Nottingham, Leicestershire, UK David M. Belnap School of Biological Sciences and Department of Biochemistry, University of Utah, Salt Lake City, UT, USA Sudhakar Bhandare Department of Food Science, McGill University, Montreal, QC, Canada Bob Blasdel Laboratory of Gene Technology, Leuven, Belgium Vésale Pharma, Noville-Sur-Mehaigne, Belgium Jan Borysowski Department of Clinical Immunology, Transplantation Institute, Medical University of Warsaw, Warsaw, Poland Michael A. Brockhurst Department of Animal and Plant Sciences, University of Sheffield, Sheffield, UK University of York, York, UK Nathan Brown Department of Infection, Immunity, and Inflammation, University of Leicester, Leicester, UK xiii

xiv

Contributors

Harald Brüssow Host-Microbiome Interaction Group, Institute of Nutritional Science, Nestlé Research Center, Lausanne, Switzerland Benjamin H. Burrowes Evolution Biotechnologies, Georgetown, TX, USA Todd R. Callaway Department of Animal and Dairy Science, University of Georgia, Athens, GA, USA Aidan Casey Department of Food Biosciences, Teagasc Food Research Centre, Cork, Ireland Marco Cavaleri Office of Anti-infectives and Vaccines, Human Medicines Evaluation Division, European Medicines Agency, Amsterdam, The Netherlands Patrick Champion-Arnaud Pherecydes Pharma SA, Romainville, France Nina Chanishvili George Eliava Institute of Bacteriophage, Microbiology and Virology (EIBMV), Tbilisi, Georgia Jason Clark Fixed-Phage Ltd, Glasgow, UK Aidan Coffey Department of Biological Sciences, Cork Institute of Technology, Cork, Ireland Chris Cox Cobio Diagnostics, Golden, CO, USA Krystyna Dąbrowska Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wrocław, Poland John J. Dennehy Biology Department, Queens College and The Graduate Center of the City University of New York, New York, NY, USA Jenna Denyes NSPM, Meggen, Switzerland Lars Fieseler Zurich University of Applied Sciences, Institute of Food and Beverage Innovation, Wädenswil, Switzerland Joe A. Fralick Department of Immunology and Molecular Microbiology, Texas Tech University Health Sciences Center, Lubbock, TX, USA Jérôme Gabard Pherecydes Pharma SA, Romainville, France Jason J. Gill Department of Animal Science, Center for Phage Technology, Texas A&M University, College Station, TX, USA Lasha Gogokhia Department of Medicine, Division of Gastroenterology and Hepatology, Weill Cornell Medicine, New York, NY, USA Lawrence Goodridge Canadian Research Institute for Food Safety, Department of Food Science, University of Guelph, Guelph, ON, Canada

Contributors

xv

Andrzej Górski Bacteriophage Laboratory, Phage Therapy Unit, Hirszfeld Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wrocław, Poland Department of Clinical Immunology, Transplantation Institute, Medical University of Warsaw, Wrocław, Poland Steven Hagens Micreos Food Safety, Gelderland, Wageningen, The Netherlands David R. Harper Evolution Biotechnologies, Bedfordshire, UK Graham F. Hatfull Department of Biological Sciences, University of Pittsburgh, Pittsburgh, PA, USA Ryan D. Heselpoth Institute for Bioscience and Biotechnology Research, University of Maryland, Rockville, MD, USA Dominique Holtappels Laboratory of Gene Technology, Leuven, Belgium Naomi Hoyle Eliava Phage Therapy Center, Eliava Foundation, Tbilisi, Georgia Paul Hyman Department of Biology/Toxicology, Ashland University, Ashland, OH, USA Jeffrey B. Jones University of Florida, Gainesville, FL, USA Tena I. Katsaounis Department of Mathematics, The Ohio State University, Mansfield, OH, USA Sarah Klopatek Department of Animal Science, University of California, Davis, Davis, CA, USA Britt Koskella Department of Integrative Biology, University of California, Berkeley, Berkeley, CA, USA Anna Krajewska Department of Bacterial Molecular Genetics, Faculty of Biology, University of Gdansk, Gdansk, Poland Elizabeth Kutter Phagebiotics Research Foundation, Olympia, WA, USA Katarzyna Kwaśnicka Department of Molecular Biology, Faculty of Biology, University of Gdansk, Gdansk, Poland Sara B. Linden Institute for Bioscience and Biotechnology Research, University of Maryland, Rockville, MD, USA Małgorzata Łobocka Autonomous Department of Microbial Biology, Faculty of Agriculture and Biology, Warsaw University of Life Sciences, Warsaw, Poland Department of Microbial Biochemistry, Institute of Biochemistry and Biophysics, Polish Academy of Sciences, Warsaw, Poland

xvi

Contributors

Joanna Łoś Department of Bacterial Molecular Genetics, Faculty of Biology, University of Gdansk, Gdansk, Poland Phage Consultants, Gdansk, Poland Marcin Łoś Department of Bacterial Molecular Genetics, Faculty of Biology, University of Gdansk, Gdansk, Poland Phage Consultants, Gdansk, Poland Marzanna Łusiak-Szelachowska Bacteriophage Laboratory, Hirszfeld Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wroclaw, Poland Martin R. MacLean Mathys and Squire LLP, London, UK Aleksandra Małachowska Department of Genetics and Biosystematics, Faculty of Biology, University of Gdansk, Gdansk, Poland Olivia McAuliffe Department of Food Biosciences, Teagasc Food Research Centre, Cork, Ireland Shawna McCallin Department of Fundamental Microbiology, University of Lausanne, Lausanne, Switzerland Zalewska Michalina Department of Bacterial Molecular Genetics, Faculty of Biology, University of Gdansk, Gdansk, Poland Ryszard Międzybrodzki Bacteriophage Laboratory, Phage Therapy Unit, Hirszfeld Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wroclaw, Poland Department of Clinical Immunology, Transplantation Institute, Medical University of Warsaw, Warsaw, Poland Catarina Milho CEB – Centre of Biological Engineering, LIBRO – Laboratório de Investigação em Biofilmes Rosário Oliveira, University of Minho, Braga, Portugal Michael S. Mitchell Institute for Bioscience and Biotechnology Research, University of Maryland, Rockville, MD, USA Daniel C. Nelson Institute for Bioscience and Biotechnology Research, University of Maryland, Rockville, MD, USA D. J. Nisbet Food and Feed Safety Research Unit, USDA/ARS, College Station, TX, USA Aleksa Ž. Obradović Faculty of Agriculture, University of Belgrade, BelgradeZemun, Serbia Siân V. Owen Microbiology Research Group, Institute of Integrative Biology, University of Liverpool, Liverpool, UK

Contributors

xvii

Department of Biomedical Informatics and Laboratory of Systems Pharmacology, Harvard Medical School, Boston, MA, USA Eric Pelfrene Office of Anti-infectives and Vaccines, Human Medicines Evaluation Division, European Medicines Agency, Amsterdam, The Netherlands Blanca M. Perez-Sepulveda Microbiology Research Group, Institute of Integrative Biology, University of Liverpool, Liverpool, UK Carrie L. Pierce Division of Laboratory Sciences, National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, GA, USA Roger D. Plaut Center for Biologics Evaluation and Research, US Food and Drug Administration, Silver Spring, MD, USA Jon C. Rees Division of Laboratory Sciences, National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, GA, USA Krzysztof Regulski Pherecydes Pharma SA, Romainville, France Jessica C. Sacher Phage Directory, Atlanta, GA, USA Marta Sanz-Gaitero Department of Macromolecular Structure, Centro Nacional de Biotecnologia (CNB-CSIC), Madrid, Spain Christine L. Schneider Carroll University, Waukesha, WI, USA Zigmars Sebris Regulatory Affairs Office, Human Medicines Evaluation Division, European Medicines Agency, Amsterdam, The Netherlands Mateo Seoane-Blanco Department of Macromolecular Structure, Centro Nacional de Biotecnologia (CNB-CSIC), Madrid, Spain T. G. Sheridan Nell Hodgson Woodruff School of Nursing, Emory University, Atlanta, GA, USA Sanna Sillankorva CEB – Centre of Biological Engineering, LIBRO – Laboratório de Investigação em Biofilmes Rosário Oliveira, University of Minho, Braga, Portugal INL International Iberian Nanotechnology Laboratory, Braga, Portugal Maria Daniela Silva CEB – Centre of Biological Engineering, LIBRO – Laboratório de Investigação em Biofilmes Rosário Oliveira, University of Minho, Braga, Portugal Scott Stibitz Center for Biologics Evaluation and Research, US Food and Drug Administration, Silver Spring, MD, USA William C. Summers Yale University, New Haven, CT, USA

xviii

Contributors

Antonet M. Svircev Agriculture and Agri-Food Canada, Vineland Station, ON, Canada Steven M. Swift Institute for Bioscience and Biotechnology Research, University of Maryland, Rockville, MD, USA Frits van Charante Ghent University, Ghent, Belgium Mark J. van Raaij Department of Macromolecular Structure, Centro Nacional de Biotecnologia (CNB-CSIC), Madrid, Spain Beata Weber-Dąbrowska Bacteriophage Laboratory, Phage Therapy Unit, Hirszfeld Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wroclaw, Poland Tryon Wickersham Department of Animal Science, Texas A&M University, College Station, TX, USA Quan-Guo Zhang College of Life Sciences, Beijing Normal University, Beijing, China Jan Zheng Phage Directory, Atlanta, GA, USA Fikria Zhvaniya Eliava Phage Therapy Center, Eliava Foundation, Tbilisi, Georgia Sylwia Zielińska Department of Bacterial Molecular Genetics, Faculty of Biology, University of Gdansk, Gdansk, Poland Phage Consultants, Gdansk, Poland

Part I Introduction to Bacteriophages: Biology, Technology, Therapy

Introduction to Bacteriophages David R. Harper

Contents Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 On the Nature of Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Antibiotics: From Savior to Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Antibiotics: A Biological Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Abstract

As the ongoing crisis of antimicrobial resistance (AMR) threatens to bring an end to the era of the routine control of bacterial diseases, there is great interest in developing other approaches to controlling such infections. One of the oldest of these, the use of bacteriophages (viruses that can target and destroy bacteria) as therapeutic agents, is experiencing a resurgence of interest and is now considered a promising approach to countering AMR. First developed 100 years ago, this approach, known as phage therapy, was set aside in Western Europe and the USA when the use of chemical antibiotics became widespread. Now, the pressing need for new ways to control such resistant bacteria is resulting in progress in developing phage therapy, alongside a range of technologies based around bacteriophages, in medicine, and elsewhere. Bacteriophages: Biology, Technology, Therapy is intended to cover all aspects of work with bacteriophages, from basic biology to clinical trials and from early history to nanotechnology.

D. R. Harper (*) Evolution Biotechnologies, Bedfordshire, UK e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_48

3

4

D. R. Harper

Bacteriophages Bacteriophages are viruses that infect bacteria. However, within that simple phrase lies a whole host of possibilities. Despite earlier observations of a variety of antibacterial activities (Abedon et al. 2011), bacteriophages were first characterized by Frederick Twort, in 1915 (Twort 1915), and again, independently, by Felix d’Herelle in 1917 (d’Herelle 1917) (see chapter ▶ “The Discovery of Bacteriophages and the Historical Context”; Duckworth 1976). At the time, bacterial diseases were a major health problem and a frequent cause of death, so it is unsurprising that bacteriophages, named for their activity as “eaters of bacteria,” were evaluated as potential therapeutic agents (d’Herelle 1919). At a time when the best available chemical antibiotics included both mercuric and arsenical compounds, there was a clear need for new, less toxic approaches. It has been claimed that, by 1925, more than 150 papers had been published on the therapeutic use of bacteriophages (Eaton and Bayne-Jones 1934). In 1925 the approach was publicized in the Pulitzer Prize winning novel Arrowsmith, written by Sinclair Lewis (1925) (see chapter ▶ “Early Therapeutic and Prophylactic Uses of Bacteriophages”). Unfortunately, despite the widespread use, the nature of bacteriophages remained unclear for many more years. During this first “bacteriophage era,” many groups offered therapeutics based on little more than optimism. An example is the identification by one provider that “Bacteriophage therapy is indicated for” a wide range of conditions including herpes (a viral infection), eczema and urticaria (immunemediated skin rashes), and even gallstones (Fig. 1). Added to this there was a lack of understanding of the high specificity of bacteriophages, again leading to inappropriate use. Indeed, at the time there was a little understanding of the diversity of bacteriophages or even that of bacteria themselves. Another complicating factor was the impurity of early bacteriophage preparations, which had a range of immunomodulatory effects due to components derived from the bacterial host that were present in the mixture (Krueger and Scribner 1941). As late as 1941, an influential review stated that, contrary to the opinions of some that it was an organism, “Phage is a protein of high molecular weight” (Krueger and Scribner 1941) – which was published actually some time after the first images of these bacterial viruses had been seen with the newly developed electron microscope (Ruska 1986). With such a misunderstanding of the fundamental nature of the agent, it is perhaps inevitable that much early work was at best inconclusive, and at worst deeply flawed. Given the above, early reviews, pooling the results of a very mixed range of work, concluded that there was limited evidence of efficacy for bacteriophage therapy (Eaton and Bayne-Jones 1934; Krueger and Scribner 1941). With the arrival of the first modern antibiotics and, in particular, the development of wide-spectrum agents such as penicillin, bacteriophages were sidelined in favor of the new drugs – at least in Western Europe and the USA.

Introduction to Bacteriophages

5

Fig. 1 An advertisement for therapeutic bacteriophages from the 1920s. (Reprinted with the kind permission of Dr. J. Soothill)

On the Nature of Bacteriophages While bacteriophages fell out of use as therapeutics, many factors, including the accessibility of their hosts, combined to make them valuable tools in the then emerging science of molecular biology (Cairns et al. 2007) (see chapter ▶ “Bacteriophage Use in Molecular Biology and Biotechnology”). The loosely constituted “phage group” of researchers used bacteriophages, notably the T4 Myovirus, as model systems to elucidate many of the fundamentals of genetics. These included the definitive demonstration that DNA is the genetic material; identifications of the genetic code; showing the function of messenger RNA; and the discovery of genetic recombination (Harper et al. 2015; Mathews 2015). In the course of this work, the mechanisms and biology of bacteriophage infection became far more fully understood (see chapters ▶ “Phage Infection and Lysis” and ▶ “Bacteriophage Use in Molecular Biology and Biotechnology”). Although many types of bacteriophages exist, more than 90% belong to the three families known as “tailed phages” (Fig. 2) – the families Myoviridae, Podoviridae, and Siphoviridae, collectively the members of the order Caudovirales (see chapter ▶ “Structure and Function of Bacteriophages”). These relatively large and complex viruses have double-stranded DNA genomes and infect a very wide range both of bacteria and the superficially similar but biologically distinct Archaea.

6

D. R. Harper

Fig. 2 The tailed bacteriophages. (Source: Harper and Enright 2011)

In the course of infection, the bacteriophage particle (the virion) drifts until it encounters a potential bacterial host cell. Receptors on the surface (usually at the tail tip) encounter specific receptors on the outer surface of the host cell. If these are of the correct type, interaction of the tail fibers with the cell surface receptors initiates a cascade of events resulting in insertion of the phage genome (and sometimes accessory molecules) into the host cell cytoplasm, where the bacteriophage life cycle initiates (see chapter ▶ “Phage Infection and Lysis”). Infection does not, however, always involve the killing of the bacterial host. Only in productive infections is the host cell is taken over, becoming a phage factory, and only in lytic infections is the host then rapidly killed to release a large number of progeny bacteriophages (Fig. 3). While a few types of bacteriophages can be productive without becoming lytic, the productive infections caused by all tailed bacteriophages are lytic. Alternatively, bacteriophages can enter into a lysogenic infection (see chapter ▶ “Temperate Phages, Prophages, and Lysogeny”), where the viral genome is maintained in a semi-quiescent state within the host cell, awaiting some triggering event (Fig. 4). Such bacteriophages which can enter into such lysogenic cycles are referred to as temperate. Tailed temperate phages enter into a lytic infection cycle upon activation, whether spontaneously or in response to a specific stimulus. There are some bacteriophages which can establish chronic, nonlethal but nevertheless highly productive infections. Typically these are filamentous bacteriophages of the family Inoviridae. These are widely used in molecular biology and protein chemistry but are rarely considered for therapeutic use (see chapters ▶ “Phage Display Technology and the Development of Phage-Based Vaccines” and ▶ “Bacteriophages in Nanotechnology: History and Future”). For phage therapy, bacteriophages that are genetically unable to enter a lysogenic state (referred to as virulent or obligately lytic) are preferred.

Introduction to Bacteriophages

7

Fig. 3 Bacterial lysis induced by bacteriophage infection. (From: Brown 2003)

Fig. 4 Bacteriophage life cycles. (Source: Harper and Enright 2011)

Antibiotics: From Savior to Crisis The first of the modern antibiotics to be used widely was the sulfonamides, or “sulfa drugs,” used extensively during World War 2. Work on these began with prontosil, developed by Bayer in 1932. This was a precursor (prodrug) to the active form, sulfanilamide, which was only produced within the body when the drug was administered to patients. Penicillin had actually been discovered earlier, in 1928 (Fleming 1945), and had even been used therapeutically in 1930 (Howie 1986). However, it was not until the work of Florey and Chain from 1939 onwards that penicillin could be produced in sufficient quantities to enter widespread use (Chain et al. 1940), ushering in the era of chemical antibiotics (Fig. 5).

8

D. R. Harper

Fig. 5 Advertisement for penicillin from World War 2

The development of antibiotics was supported by the new approach of controlled clinical trials, which was not the case for bacteriophage therapeutics at the time they were initially being developed. Indeed, early work with antibiotics defined much of what we now think of as clinical trials (Streptomycin in Tuberculosis Trials Committee 1948). For many years they were seen as the answer to bacterial disease.

Introduction to Bacteriophages

9

Each of the original antibiotics was modified and modified again to produce a range of drugs with altered or enhanced activity, forming the basis of the modern antibiotic pharmacopoeia. In Western Europe and in the USA, chemical antibiotics became the approach of choice. There can be no doubt that at that time antibiotics were truly wonder drugs, and they went on to save millions upon millions of lives. Despite many citations, it has now been conclusively disproven that Dr. William H. Stewart, the US Surgeon General from 1965 to 1969, ever actually said “It is time to close the book on infectious diseases and declare the war against pestilence won” (Spellberg and Taylor-Blake 2013). However, that statement certainly reflects the widespread belief at that time that antibiotics were then and would remain the only answer to bacterial disease that humanity needed or would ever need. Unfortunately this wasn’t true. Some bacteria are naturally resistant to many antibiotics, whether by structural features such as the complex cell wall of Gram-negative bacteria or by growth in structures such as biofilms (Harper et al. 2014b). But key to current concerns is acquired antibiotic resistance, which has the potential to transfer between even unrelated bacterial species. Antibiotic resistance was discussed by Alexander Fleming, the discoverer of penicillin, in his acceptance speech for the Nobel Prize in 1945. Penicillin resistance is often cited as arising shortly after its first widespread use in 1943, but there are reports of resistant Staphylococcus as early as 1940 arising from early, experimental uses (Ventola 2015). Similarly, resistance has appeared for every class of chemical antibiotics developed to date (Table 1), with pan-drug resistant (PDR) bacteria (resistant to all known antibiotics) now being reported. As can be seen from Table 1, historically the mean time from introduction to the appearance of resistance is approximately 6 years. For those few antibiotics introduced since 2000, the mean time is less than 2 years. Even antibacterial treatments long thought immune to such concerns, including metals such as copper and silver (Hobman and Crossman 2015) and many (though not yet all) antiseptics (Lachapelle et al. 2013), have been countered by multiple mechanisms of resistance. Many of the determinants of such resistance can, as noted, be transferred between bacteria on mobile genetic elements, including by some types of bacteriophage. This ability for resistance to spread under the selective pressure resulting from antibiotic use has led to the evolution of the “superbugs,” bacteria resistant to many or even all known antibiotics. Underlying this consequence is a simple truth: Life evolves. Chemicals do not. In time, resistant forms can become a widespread and serious clinical problem, as with MRSA or XDR TB (extremely drug-resistant tuberculosis). With the arrival of bacterial strains showing resistance to all known antibiotics (pan-drug resistance or PDR), there is the potential for the spread of untreatable bacterial disease (WHO 2017). It is this genuinely worrying possibility that underlies the many concerns expressed, from the authoritative, such as the statements by the Director-General of the World Health Organization (WHO) regarding the risks of a post-antibiotic era (Chan 2016), to the many lurid headlines in mainstream media and elsewhere.

10

D. R. Harper

Table 1 First reported appearance of resistance to some common antibiotics Antibiotic Penicillin Tetracycline Erythromycin Methicillin Gentamicin Vancomycin Imipenem Ceftazidime Levofloxacin Linezolid Daptomycin Ceftaroline All available antibiotics (PDR bacteria)

Introduction 1943 1950 1953 1960 1967 1972 1985 1985 1996 2000 2003 2010

First clinical resistance 1940–1950s 1959 1968 1962 1979 1988 1998 1987 1996 2001 2005 2011 2004

Data derived from Ventola (2015)

Multiple priority lists have been prepared, identifying highly resistant species of bacteria for which there is an urgent need for new antibacterial approaches (CDC 2013; WHO 2017). The risk is real, and we need to counter it. As resistance to even novel chemical antibiotics shows, bacteria can and will evolve to counter any such drug. However, while resistance is widespread, this does not mean that all antibiotics have, for now, stopped working. One potential positive is that resistance to one antibiotic type may sometimes confer a degree of susceptibility to other antibiotics – the so-called see-saw effect (Barber et al. 2014; Yang et al. 2010). However, this is by no means to be relied upon, and resistance to all known antibiotics has been reported. It should be noted, in particular, that the first observation of clinical resistance does not mean that such resistance is universal, nor even widespread. Initially, such resistance tends to be quite limited but can spread. Bacteria have many mechanisms to share such resistance, not only within but also between bacterial species (Hiltunen et al. 2017). Once resistance has been observed, it will remain a concern. Furthermore, if resistance has appeared once in any strain of a bacterial species, then it is perfectly possible for it to appear again, independently. The situation has been summarized as follows: “We are losing our first-line antibiotics. This makes a broad range of common infections much more difficult to treat. Second- and third-choice antibiotics are more costly, more toxic, need much longer durations of treatment, and may require administration in intensive care units” (Chan 2016). More extreme resistance is seen in the case of pan-drug resistance, where even the antibiotics of last resort no longer work. As noted, multidrug resistance, in most cases, still leaves some windows of vulnerability – some drugs that continue to work, at least for now. But those drugs can be difficult and expensive to use, and identifying those drugs which still work

Introduction to Bacteriophages 18 16

11

16 14

14 12

10

10

7

8

5

6 4

2

2 0 1983-1987

1988-1992

1993-1997

1998-2002

2003-2007

2008-2012

Fig. 6 Approvals of antibiotics (US). (Data from The Pharmaceutical Journal, Vol. 291, 520)

can be time-consuming. In severe infections the time to screen for such vulnerabilities in the infecting bacteria can be very limited. Attempts to conserve antibiotics through controlled prescribing (“antibiotic stewardship”) are under way, along with efforts to limit their use in noncritical areas such as use as growth promoters in farmed animals. However, while worthwhile, these efforts are not as effective as might be hoped due to a combination of commercial concerns, a lack of global adherence to proposed guidelines, and an understandable desire among patients for any potentially effective treatment. There is a very real danger that our ability to find an effective antibiotic to counter routine bacterial infections could soon become a thing of the past. And there is another pressure. For big pharma companies, antibiotics are not the best area in which to be active. The financial returns from a short course of treatment are limited compared to “lifestyle drugs” such as the cholesterol-lowering statins that may be taken every day for life. This is reflected in the rapid and sustained decline in the number of antibiotics making their way through the development process and onto the market (Sukkar 2013) (Fig. 6): Despite early optimism, bacteria can kill and will continue to kill, unless we have ways to stop them. In the words of the Director-General of the World Health Organisation (WHO), “the world is heading towards a post-antibiotic era in which common infections will once again kill . . . This may even bring the end of modern medicine as we know it. We need to act now to make sure this does not happen. This is a crisis, it is global” (Chan 2016).

Antibiotics: A Biological Approach The crisis of antibiotic resistance has been recognized for over 25 years (Neu 1992). Unsurprisingly, a wide range of studies have addressed approaches to counter this. One recent study on this topic considered a very wide range of approaches (Czaplewski et al. 2016), from immunomodulatory agents to metal chelation, and

12

D. R. Harper

reported that bacteriophages were considered to be one of the most promising approaches. Critically, our understanding of the nature of bacteriophages, as well as the tools that we have developed to further understand them, has improved considerably since the early days of phage therapy (as above). It is understood that bacteriophages are extremely abundant. They have been found in all niches of prokaryotic life, outnumbering their hosts approximately 10:1 with an estimated 1031 on Earth (Rohwer and Edwards 2002), making them the most plentiful form of life on the planet. Bacteriophages offer an unimaginably large resource not only for combatting infections but for many other things besides, such as understanding ecology at small and large scales as applied to microbial communities and their role in the global ecology (see chapter ▶ “Bacteriophage Ecology”). Bacteriophages also provide many of the basic materials, reagents, models, tools, and techniques for molecular biology, genetics, and bioinformatics (see chapters ▶ “Bacteriophage Use in Molecular Biology and Biotechnology”, and ▶ “Detection of Bacteriophages: Sequence-Based Systems”), and, of course, bacteriophages have many uses in biotechnology (see chapters ▶ “Bacteriophage as Biocontrol Agents”, and ▶ “Bacteriophages in Nanotechnology: History and Future”). Bacteriophages have unique strengths as antibacterials. Lytic bacteriophages will infect their host, and only where that host is present will undergo repeated rounds of replication, amplifying themselves to a level matching that of the infecting bacteria even from a tiny initial dose (Marza et al. 2006; Wright et al. 2009). This in situ amplification, occurring only as and where needed, permits very low dosing levels – down to a billionth of the dose used with a conventional antibiotic. This, combined with the high specificity of bacteriophages, results in low levels of toxicity. Other unique strengths of the approach include the ability of bacteriophages to evolve in response to resistance generated by their target bacteria (see chapter ▶ “BacteriaPhage Antagonistic Coevolution and the Implications for Phage Therapy”) and their ability to control bacterial biofilms (Harper et al. 2014b) (see chapter ▶ “Biofilm Applications of Bacteriophages”). Against this, bacteriophages are large nucleoprotein complexes that can diffuse poorly and may be immunogenic when used systemically. Compared to conventional antibiotics, they can have limited stability, and delivery can be challenging. A key point is that no bacteriophage therapeutic has yet completed the complex and demanding regulatory process and found its way to market in the USA or Western Europe. However, progress is being made. Phage therapy has a long and somewhat cloudy history as noted above. Unfortunately, it has long been seen by many in the light of overoptimistic claims for efficacy arising from early work in the 1920s and 1930s (Eaton and Bayne-Jones 1934; Krueger and Scribner 1941). However, even after chemical antibiotics came into widespread use, phage therapy had never fallen out of use in some locations, notably in Eastern Europe (see chapter ▶ “Current Updates from the Long-Standing Phage Research Centers in Georgia, Poland, and Russia”).

Introduction to Bacteriophages

13

In 1926, George Eliava, a microbiologist from what was then the Soviet Republic of Georgia, had visited d’Herelle in Paris. When he returned to Georgia, he worked on setting up the centralized bacteriophage institute which now bears his name. Work continued after Eliava’s death in the Stalinist purges, and the institute that he founded took a leading position in preparing and providing millions of bacteriophage preparations for use in the Soviet Union and elsewhere in Eastern Europe. Here, phage therapy is seen to have nothing to prove and is used as an accepted part of medical practice. However, what studies were carried out reflected local priorities. Results were generally published in Russian or Georgian and did not align with the methods of the clinical trials by then being carried out in Western Europe and the USA. By the time that the emerging antibiotic crisis was attracting attention in the 1990s, there was far more knowledge of bacteriophage biology on which to base new work on phage therapy, from both east and west. In the 1980s, a British group led by William Smith conducted a series of carefully designed, well-controlled experiments that provided a basis for understanding how phage therapy could work in animal systems (Smith and Huggins 1982, 1983; Smith et al. 1987). The title of the first such paper, Successful Treatment of Experimental Escherichia coli Infections in Mice Using Phage: Its General Superiority Over Antibiotics, indicates the hopeful outcome of this work. Additional animal studies followed over the following years, the results of which in general supported the optimistic message (Harper et al. 2014a). During the following decade, with the need for new approaches now clear, a number of small companies began working to revive phage therapy in Western Europe and the USA. While the Eastern European work provided a basis for believing in the technology, bringing it west proved problematic. As well as operational difficulties, it soon became clear that western regulators were unable to accept work carried out to such different standards (see chapter ▶ “Regulatory Aspects of the Therapeutic Use of Bacteriophages: Europe”), and some companies simply started anew. It was these companies that were to carry out the first modern clinical trials of phage therapy (see chapter ▶ “Clinical Trials of Bacteriophage Therapeutics”). The first such trial was a phase 1 (safety) trial carried out in 1999 in London. While the results were never published in full, this work was followed by other safety trials, in Switzerland (Bruttin and Brüssow 2005), Bangladesh (McCallin et al. 2013), Belgium (Rose et al. 2014), and the USA (Rhoads et al. 2009). Given what is known of the high specificity of bacteriophages – and the small input doses down to picogram range (Marza 2006) permitted by the use of a therapeutic agent that can amplify itself at need – it is unsurprising that safety did not and indeed does not appear to be the major concern. One problem, however, is that a pure phase 1 clinical trial treatment is given to healthy individuals. If the recipients do not have the right bacterial infection, amplification cannot and does not occur, so therapeutic dosing levels (which are generated in vivo by this amplification) simply cannot happen, thus rendering the resulting safety data of limited utility. A limited number of clinical trials have taken place in patients with the relevant infection, thus allowing extended assessment of safety as well as of efficacy.

14

D. R. Harper

Here, while one trial has reported positive outcomes, results have been more mixed. Topical use against ear infections appeared to be successful (Wright et al. 2009), while oral dosing for gut infections was less so (Sarker et al. 2016) (see chapter ▶ “Selection of Disease Targets for Phage Therapy”). Other results are awaited at the time of writing. However, it is clear that the number of modern trials is very limited, and far more work needs to be done, including the large, phase 3 clinical trials that support progression to market. Some groups have elected to bypass the trial process, at least initially, focusing on “expanded access” (in the USA) and the Declaration of Helsinki (in Europe) that permit uses for individual, seriously ill patients (Schooley et al. 2017). In such cases, compiling data to support more generalized use can be challenging. Other groups are taking a more personalized approach, seeking to establish approved phage banks from which phages can be tested on bacteria from individual patients and individualized mixtures prepared on a customized basis (Pirnay et al. 2011). However, the regulatory basis for such an approach to enter widespread use has yet to be established. While multiple approaches are under evaluation, there is clearly a great deal of effort going into making phage therapy a global reality. When this will happen is not yet certain, but it is clear that new approaches to combat the continuing crisis of antimicrobial resistance are needed and that phage therapy has the potential to be a major part of this (Czaplewski et al. 2016). Bacteriophages: Biology, Technology, Therapy is intended to cover all major, current aspects of work with bacteriophages, from their basic biology to clinical trials of phage therapeutics and from early history (see chapter ▶ “The Discovery of Bacteriophages and the Historical Context”) to nanotechnology (see chapter ▶ “Bacteriophages in Nanotechnology: History and Future”). In so doing, the intention is to provide a single, readily citable source covering the biology of bacteriophages and bacteriophage infection, their use across a wide range of technologies, and their evolving use as therapeutic agents.

References Abedon ST, Thomas-Abedon C, Thomas A, Mazure H (2011) Bacteriophage prehistory: is or is not Hankin, 1896, a phage reference? Bacteriophage 1:174–178 Barber KE, Ireland CE, Bukavyn N, Rybak MJ (2014) Observation of “seesaw effect” with vancomycin, teicoplanin, daptomycin and ceftaroline in 150 unique MRSA strains. Infect Dis Ther 3:35–43 Brown JC (2003) Virology. In: The encyclopedia of life sciences. Wiley, Chichester. https://doi.org/ 10.1038/npg.els.0000435. www.els.net Bruttin A, Brüssow H (2005) Human volunteers receiving Escherichia coli phage T4 orally: a safety test of phage therapy. Antimicrob Agents Chemother 49:2874–2878 Cairns J, Stent GS, Watson JD (eds) (2007) Phage and the origins of molecular biology. CSHL Press, New York CDC (2013) Antibiotic resistance threats in the United States, 2013. https://www.cdc.gov/ drugresistance/threat-report-2013/index.html Chain E, Florey HW, Adelaide MB et al (1940) Penicillin as a chemotherapeutic agent. Clin Orthop Relat Res 295:3–7

Introduction to Bacteriophages

15

Chan M (2016) WHO Director-General briefs UN on antimicrobial resistance. http://www.who.int/ dg/speeches/2016/antimicrobial-resistance-un/en/ Czaplewski L, Bax R, Clokie M, Dawson M, Fairhead H, Fischetti VA, Foster S, Gilmore BF, Hancock REW, Harper D, Henderson IR, Hilpert K, Jones BV, Kadioglu A, Knowles D, Ólafsdóttir S, Payne D, Projan S, Shaunak S, Silverman J, Thomas CM, Trust TJ, Warn P, Rex JH (2016) Alternatives to antibiotics–a pipeline portfolio review. Lancet Infectious Diseases 16: 239–251 d’Herelle F (1917) Sur un microbe invisible antagonistic des bacilles dysenteriqes. C R Acad Sci Paris 165:373–375 d’Herelle F (1919) Sur le role du microbe bacteriophage dans la typhose aviare. C R Acad Sci Paris 169:932–934 Duckworth DH (1976) Who discovered bacteriophage? Bacteriol Rev 40:793–802 Eaton MD, Bayne-Jones S (1934) Bacteriophage therapy: review of the principles and results of the use of bacteriophage in the treatment of infections. JAMA 103:1769–1776; 1847–1853; 1934–1939 Fleming A (1945) https://www.nobelprize.org/nobel_prizes/medicine/laureates/1945/fleming-lec ture.pdf Harper DR (2011) Viruses: Biology, Applications and Control. Garland Science, New York Harper DR, Enright MC (2011) Bacteriophages for the treatment of Pseudomonas aeruginosa infections. J Appl Microbiol 111:1–7 Harper DR, Burrowes BH, Kutter EM (2014a) Bacteriophage: therapeutic uses. In: The encyclopedia of life sciences. Wiley, Chichester. www.els.net Harper DR, Parracho HMRT, Walker J et al (2014b) Bacteriophages and biofilms. Antibiotics 3:270–284 Harper DR, McConville M, Anderson FJ, Enright MC (2015) Antimicrobial phages. In: Tang YW et al (eds) Molecular medical microbiology, 2nd edn. Elsevier, San Diego Hiltunen T, Virta M, Laine AL (2017) Antibiotic resistance in the wild: an eco-evolutionary perspective. Philos Trans R Soc Lond Ser B Biol Sci 372:20160039 Hobman JL, Crossman LC (2015) Bacterial antimicrobial metal ion resistance. J Med Microbiol 64:471–497 Howie J (1986) Penicillin: 1929–40. Br Med J 293:158–159 Krueger AP, Scribner EJ (1941) The bacteriophage: its nature and its therapeutic use. JAMA 116:2160–2167 Lachapelle J-M, Castel O, Casado AF et al (2013) Antiseptics in the era of bacterial resistance: a focus on povidone iodine. Clin Pract 10:579–592 Lewis S (1925) Arrowsmith. Harcourt Brace and Company, New York Mathews CK (2015) Bacteriophage T4. In: eLS. John Wiley & Sons Ltd, Chichester. http://www. els.net Marza JA, Soothill JS, Boydell P, Collyns TA (2006) Multiplication of therapeutically administered bacteriophages in Pseudomonas aeruginosa infected patients. Burns 32:644–646 McCallin S, Sarker SA, Barretto C et al (2013) Safety analysis of a Russian phage cocktail: from metagenomic analysis to oral application in healthy human subjects. Virology 443:187–196 Neu HC (1992) The crisis in antibiotic resistance. Science 257:1064–1073 Pirnay JP, De Vos D, Verbeken G et al (2011) The phage therapy paradigm: pret-a-porter or sur-mesure? Pharm Res 28:93493–93497 Rhoads DD, Wolcott RD, Kuskowski MA et al (2009) Bacteriophage therapy of venous leg ulcers in humans: results of a phase I safety trial. J Wound Care 18:240–243 Rohwer F, Edwards R (2002) The phage proteomic tree: a genome-based taxonomy for phage. J Bacteriol 184:4529–4535 Rose T, Verbeken G, Vos DD et al (2014) Experimental phage therapy of burn wound infection: difficult first steps. Int J Burns Trauma 4:66–73 Ruska E (1986) Nobel lecture. The development of the electron microscope and of electron microscopy. http://www.nobelprize.org/nobel_prizes/physics/laureates/1986/ruska-lecture. html

16

D. R. Harper

Sarker SA, Sultana S, Reuteler G et al (2016) Oral phage therapy of acute bacterial diarrhea with two coliphage preparations: a randomized trial in children from Bangladesh. EBioMedicine 4:124–137 Schooley RT, Biswas B, Gill JJ et al (2017) Development and use of personalized bacteriophagebased therapeutic cocktails to treat a patient with a disseminated resistant Acinetobacter baumannii infection. Antimicrob Agents Chemother 61:e00954–e00917 Smith HW, Huggins MB (1982) Successful treatment of experimental Escherichia coli infections in mice using phage: its general superiority over antibiotics. J Gen Microbiol 128:307–318 Smith HW, Huggins MB (1983) Effectiveness of phages in treating experimental Escherichia coli diarrhoea in calves, piglets and lambs. J Gen Microbiol 129:2659–2675 Smith HW, Huggins MB, Shaw KM (1987) The control of experimental Escherichia coli diarrhoea in calves by means of bacteriophages. J Gen Microbiol 133:1111–1126 Spellberg B, Taylor-Blake B (2013) On the exoneration of Dr. William H. Stewart: debunking an urban legend. Infect Dis Poverty 2:3 Streptomycin in Tuberculosis Trials Committee (1948) Streptomycin treatment of pulmonary tuberculosis. A Medical Research Council investigation. Br Med J 2:769–782 Sukkar E (2013) Why are there so few antibiotics in the research and development pipeline? Pharm J 291:520 Twort FW (1915) An investigation on the nature of ultramicroscopic viruses. Lancet 1915:1241 Ventola CL (2015) The antibiotic resistance crisis. Part 1: causes and threats. P T 40:277–283 WHO (2017) WHO publishes list of bacteria for which new antibiotics are urgently needed. http:// www.who.int/mediacentre/news/releases/2017/bacteria-antibiotics-needed/en/ Wright A, Hawkins C, Anggard EA, Harper DR (2009) A controlled clinical trial of a therapeutic bacteriophage preparation in chronic otitis due to antibiotic-resistant Pseudomonas aeruginosa; a preliminary report of efficacy. Clinical Otolaryngology 34: 349–357 Yang SJ, Xiong YQ, Boyle-Vavra S et al (2010) Daptomycin-oxacillin combinations in treatment of experimental endocarditis caused by daptomycin-nonsusceptible strains of methicillin-resistant Staphylococcus aureus with evolving oxacillin susceptibility (the “seesaw effect”). Antimicrob Agents Chemother 54:3161–3169

Part II Bacteriophage Biology

Structure and Function of Bacteriophages Marta Sanz-Gaitero, Mateo Seoane-Blanco, and Mark J. van Raaij

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview of Phage Families . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Leviviridae Family of Single-Stranded RNA Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Microviridae Family of Single-Stranded DNA Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Inoviridae Family of Filamentous, Single-Stranded DNA Phages . . . . . . . . . . . . . . . . . . . . . The Cystoviridae Family of Double-Stranded RNA Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview of Bacteriophages Containing Double-Stranded DNA Genomes . . . . . . . . . . . . . . . . . . . . The Tectiviridae Family . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Corticoviridae Family . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview of Order Caudovirales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Caudovirales Head Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Caudovirales Head Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Podoviridae Tail Assembly and Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Siphoviridae Tail Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Siphoviridae Tail Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Myoviridae Tail Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Myoviridae Tail Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions and Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

20 21 28 33 36 40 43 44 47 51 52 54 59 62 65 71 73 80 82 82

Abstract

Bacteriophages, or phages, are viruses with an exquisitely evolved structure to accomplish their goals. These goals are recognizing a suitable host bacterium, profiting from the host metabolism, and producing multiple progeny phages that are stable enough to survive until they find a new host bacterium to infect. Their M. Sanz-Gaitero · M. Seoane-Blanco · M. J. van Raaij (*) Department of Macromolecular Structure, Centro Nacional de Biotecnologia (CNB-CSIC), Madrid, Spain e-mail: [email protected]; [email protected]; [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_1

19

20

M. Sanz-Gaitero et al.

genomes consist of single-stranded RNA, double-stranded RNA, single-stranded DNA, or double-stranded DNA, depending on phage type. They store their genome in highly symmetric protein capsids to protect it from degradation. Often these capsids are icosahedral, but helical and other shapes are also used. Tectiviridae and Corticoviridae have an internal lipid membrane, while Cystoviridae sport an outer membrane layer. Phages with tails, belonging to the Caudovirales order, are the most commonly encountered bacteriophages and have icosahedral or prolate capsids. In addition to the capsid, phages need a host cell recognition apparatus. The small icosahedral Leviviridae have a single minor capsid protein for this purpose. More complex phages dedicate multiple proteins to host cell recognition, and examples of this are the helical Inoviridae and the icosahedral Tectiviridae, Corticoviridae, and Cystoviridae. The Caudovirales have highly efficient tail protein complexes for DNA transfer. These tails are flexible (Siphoviridae), extensible (Podoviridae), or contractile (Myoviridae). Apart from elements designed for genome protection, host recognition, and genome transfer, more complicated phage particles may contain proteins for environmental sensing, binding to suitable matrices where host bacteria are likely to be encountered, and other functions.

Introduction The replication cycle of a bacteriophage consists of several steps: (i) recognition of a suitable bacterium to infect, (ii) the transfer of the genomic material into the host, (iii) subversion of the host metabolic machinery to produce a multitude of new phage particles, (iv) escape from the confines of the host cell by lysis or secretion, and (v) the wait for an encounter with a new, suitable, host bacterium (see chapter ▶ “Phage Infection and Lysis”). To perform these tasks, bacteriophages have evolved metastable particles. They need to survive between infections, protecting their genomic material inside a sturdy protein capsid. At the same time, these capsids contain exposed or exposable sites for the recognition of new hosts. When this recognition happens, the capsids are poised for important conformational changes necessary to transfer their genomic material into the bacterium. Due to their small size, from a few tens of nm to a few hundred nm in their largest dimension, phages have relatively small genomes which cannot code for a lot of proteins (see chapters ▶ “Genetics and Genomics of Bacteriophages” and ▶ “Bacteriophage Discovery and Genomics”). Therefore, phage capsids are generally symmetrical, with many copies of the same protein. Icosahedral shapes are the most common, but helical phages also exist. Some phages decorate their capsids with proteins that are thought to bind molecules present in matrices where host bacteria are likely to be found, for instance, mucins for binding in animal lungs or intestine. Phages may also have structures designed for environmental sensing; displaying certain proteins only when encountering a suitable host under the right conditions where successful infection is likely.

Structure and Function of Bacteriophages

21

Another constraint on phage structure is the cell wall of the bacterial host, including glycans, outer membrane proteins, flagella, and other appendages. The phage must be adapted to recognize one or more of these molecules efficiently and conclusively. On the one hand, phages must avoid transferring their genomic material into a host cell that is unsuitable for replication, and on the other, for maximum evolutionary success, they should not miss any suitable hosts. The phage must also carry a mechanism to help its genetic material traverse the cell wall and membranes. Once the genome has entered the bacterium, the phage can take over host cell metabolism. Most phages only transfer a few proteins into the cell with their nucleic acid, relying on the host cell for production of the rest of their gene products. In this chapter, we discuss the virion structures of bacteriophages of the Leviviridae, Microviridae, Inoviridae, Cystoviridae, Tectiviridae, Corticoviridae, Siphoviridae, Podoviridae, and Myoviridae families (Fig. 1). We will also relate these structures to function. A glossary of terms relevant to phage structure and function can be found in Table 1.

Overview of Phage Families The Leviviridae (Fig. 1a) are a family of small icosahedral bacteriophages with a monopartite, single-stranded, linear, plus-strand RNA genome that serves as a messenger RNA, encoding only four proteins: the coat protein, the replicase, the maturation protein, and the lysis protein. There are two genera: Levivirus (containing the species MS2 and BZ13) and Allolevivirus (containing the species Qβ and F1). The Microviridae (Fig. 1b) are a family of small icosahedral bacteriophages with a single-stranded, circular DNA genome. Virions contain the plus-strand of the DNA, which means the minus-strand has to be generated intracellularly to be used as a template for the generation of messenger RNAs by transcription. The Microvirinae are the most representative genus of the Microviridae and include the wellstudied ϕX174 but also phages G4 and α3. Other genera are the less well-studied Gokushovirinae, Pichovirinae, Aravirinae, and Stokavirinae. The Inoviridae (Fig. 1c) are a family of rod-shaped filamentous viruses with a circular, single-stranded DNA molecule (plus-strand). They have a helical structure and do not lyse the bacterium after infection but use the host cell to continually produce progeny virions by extrusion through the host membrane. Accepted genera are Inovirus (including the well-known cloning vehicle and phage display tool M13), Fibrovirus, Habenivirus, Lineavirus, Plectrovirus, Saetivirus, and Vespertiliovirus. The Cystoviridae (Fig. 1d) are the only known family of dsRNA viruses that infect bacteria. They have a nucleocapsid containing three double-stranded RNA segments: a small, a medium, and a large RNA. The nucleocapsid is covered by a lipid membrane layer. They are only known to infect Pseudomonas bacteria, and only a few species are known, belonging to a single genus: Cystovirus.

22

M. Sanz-Gaitero et al.

Fig. 1 Schematic drawing of the levivirus MS2 (a), the microvirus ϕX174 (b), the inovirus M13 (c), the cystovirus ϕ6 (d), the tectivirus PRD1 (e), the corticovirus PM2 (f), the myovirus T4 (g), the siphovirus T5 (h), and the podovirus T7 (i). Phages are not drawn to scale, and the inovirus in panel C should be longer, with a longer genome (red) and many more yellow major coat protein subunits. The type of genome of each phage is indicated under their drawings

The Tectiviridae (Fig. 1e) are a family of double-stranded DNA phages that infect Gram-negative bacteria. These phages have an icosahedral capsid (structurally related to adenovirus capsids) that is decorated with spikes at the fivefold vertices. They do not have tails, but the capsid encloses an internal host-derived membrane. Upon infection, this membrane, together with associated proteins, is extruded and functions as an extensible appendage for DNA transfer. The Corticoviridae (Fig. 1f) are also icosahedral viruses with a circular double-stranded DNA genome.

Structure and Function of Bacteriophages

23

Table 1 Glossary of terms relevant to phage structure and function Term α-Helix

Asymmetric unit

Avidity

β-Hairpin β-Sheet (also β-pleated sheet) β-Strand

Baseplate Capsid Capsid expansion

Capsid maturation

Cementing protein

Chaperone protein Chronic infection

Coat protein

Connector protein

Contractile tail Decoration protein

Definition Secondary structure element of a protein chain consisting of a righthanded helix, in which the backbone nitrogen donates a hydrogen bond to the carbonyl oxygen of the amino acid four residues found before it Minimal unit of a symmetric structure that, when repeated in space according to the different symmetry elements, can generate the entire structure Increased binding strength due to multiple copies of the same protein or protein domain binding to multiple copies of a ligand attached to each other or to a surface (like a bacterial membrane) Minimal β-sheet made up of two antiparallel β-strands connected to each other by a loop region Flat or curved part of a protein structure made up of multiple β-strands interacting with each other via backbone hydrogen bonds; β-sheets can be parallel, with the strands running in the same direction, or antiparallel, with the strands running in alternate directions Secondary structure element of a protein chain consisting of an extended region of the protein chain, interacting with other β-strands via backbone hydrogen bonds to form a β-sheet Structure found at the end of a phage tail that orchestrates the interaction with host bacteria and the start of DNA transfer into the host Protein shell that contains the genetic material of the phage The process of a phage capsid enlarging as it is filled by DNA and internal scaffolding proteins are removed. It is accompanied by a thinning of the capsid wall (see also capsid maturation) The broader process of a phage capsid changing conformation after assembly and into an infectious phage. It entails capsid expansion, but also processing such as proteolytic processing of capsid proteins, when applicable (see also capsid expansion) Proteins that bind to the exterior part of the capsid at regular locations, usually between coat protein subunits, to stabilize the capsid wall (also called staple protein) Protein that helps one or more other proteins fold and assemble into the correct arrangement but does not form part of the final structure Phage infection that does not lead to host cell lysis, but transforms the host into a phage factory, constantly extruding newly formed phage particles. Plaques can still form due to drastically reduced rates of bacterial cell division compared to uninfected cells Protein making up the capsid of a phage. Multiple copies of the same coat protein subunits make up a symmetrical icosahedral, prolate, or helical capsid (see also major capsid protein) Protein that connects the head to the tail of tailed bacteriophages. Twelve circularly arranged copies of the connector protein form the special or unique vertex in the head and function as the gateway for DNA entry and release (also called portal protein) Double-layered tail of the Myoviridae, of which the outer sheath layer can contract to drive the inner tail tube into the host cell wall Protein bound to the outside of the phage capsid or tail without an apparent function for the structural stability of the phage. They likely (continued)

24

M. Sanz-Gaitero et al.

Table 1 (continued) Term

DNA transfer DNA translocation Domain swapping Enveloped virus Extensile tail Fiber diffraction

Filamentous phage

Fusogenic protein or peptide Head-full packaging

HK97 fold

Icosahedron

Immunoglobulin domain Jelly roll

Lipopolysaccharide

Major capsid protein

Definition have a function in host binding or in helping the phage retain position in favorable environments such as an animal lung or gut Movement of DNA out of the phage and into the host bacterium during infection Assisted incorporation of newly synthesized phage DNA into the preformed capsid by packaging proteins Instance where one or more secondary structure elements (often a β-strand) insert themselves into the fold of a neighboring protein Virus with a membrane covering its protein capsid. Additional viral proteins are inserted into the membrane Tail formed during the infection process by proteins that come from inside the capsid. Podoviridae and Microviridae use this mechanism Technique where aligned biomacromolecules are irradiated and the diffraction patterns recorded to make inferences about the underlying structure of sample Phage that has a long, thin, and to some extent flexible shape, i.e., those belonging to the Inoviridae family. An interesting consequence of this shape is that the length of the virus varies with genome size Protein or peptide that mediates fusion between viral and host cell membranes Mechanism where DNA is translocated into the phage capsid until a certain pressure is reached, i.e., the capsid (head) is full. It means that somewhat more than a complete genome is packaged and that different phage particles have different genome, terminally redundant, ends (in an alternative mechanism, packaging starts at a specific genome site, the packaging signal, and ends when a termination site is encountered) Basic fold of the major capsid proteins of all members of the Caudovirales order, suggesting that all these viruses, and including the Herpesviridae family, have a common ancestor Geometric figure with a high internal symmetry, consisting of 12 vertices, 20 triangular faces, and 30 ridges. Each vertex contains a fivefold symmetry axis, each face a threefold symmetry axis, and each ridge a twofold symmetry axis Protein domain with a fold similar to that encountered in antibodies and other immune system proteins. The fold consists of two β-sheets forming a sandwich with a specific topology (ABED-CFG) Fold encountered in the capsid protein of many viruses, including phages belonging to the Microviridae, Tectiviridae, and Corticoviridae families. The fold consists of two β-sheets forming a sandwich with a specific topology (BIDG-CHEF) Outer membrane component of Gram-negative bacteria that is also called endotoxin. They consist of a lipid A membrane anchor, an oligosaccharide inner and outer core, and an O-antigen Viral capsid protein with the highest copy number, i.e., the protein that forms the capsid faces for icosahedral viruses (see also coat protein) (continued)

Structure and Function of Bacteriophages

25

Table 1 (continued) Term Metastable particles

O-antigen

Packaging signal

Peptidoglycan

Pilus

Plasmid

Portal protein Procapsid Prolate capsid

Receptor-binding protein Scaffolding protein

Special vertex

Spike protein Staple protein

Stem-loop

Definition Metastable particles are stable phage particles which break apart (usually in a specific way) in response to a certain trigger, such as receptor binding or pH The O-antigen is the part of the bacterial lipopolysaccharide that is exposed to the environment. It is made up of repeating oligosaccharide units and is specific to the bacterial species. Phage receptor-binding proteins often specifically recognize the oligosaccharide units Site on the viral genome that serves to initiate the assembly of the capsid around it or as a recognition site for proteins to start translocating the genome into a preformed phage capsid Peptidoglycan, or murein, forms a meshed layer outside the bacterial plasma membrane. It is composed of glycan polymers cross-linked by peptide bridges. Bacterial species have different peptidoglycan structures A pilus (plural pili) is a thin appendage of bacteria used in DNA transfer between them (conjugation). Phages can use them as specific attachment sites and use the fact that they are regularly retracted to get close and enter the bacterial cell wall Circular double-stranded DNA structure present in many bacteria that is small compared to the bacterial genome and that replicates independently. Plasmids often encode genes that are useful to bacteria only in certain circumstances See connector protein Immature capsid of a bacteriophage, generally not yet containing genomic material. Procapsids may still contain scaffolding proteins Phage capsid with an elongated shape. The caps on the ends are icosahedral, but extra rings of protein subunits are inserted between them, yielding the elongated shape Phages use receptor-binding proteins to specifically recognize elements on the surface of suitable host bacteria. These proteins are usually trimeric, i.e., made up of three identical protein chains Protein structures that serve to assemble coat protein units into a complete capsid. Phages may have internal and/or external scaffolding proteins Vertex of icosahedral and prolate capsids where the portal protein is located; the other 11 vertices contain pentameric capsid proteins with the HK97 fold. This vertex initiates capsid assembly, is where DNA translocation into the capsid takes place and where DNA transfer into the host originates (also called unique vertex) Proteins forming spikes on the viral surface. These proteins are generally homo-trimeric and involved in receptor binding and/or membrane fusion Proteins that bind to the exterior part of the capsid at regular locations, locking together coat protein subunits, to stabilize the capsid wall (see also cementing protein) RNA secondary structure in which two nearby, short complementary regions pair up to form a small double helix (the stem), connected by a (continued)

26

M. Sanz-Gaitero et al.

Table 1 (continued) Term

Structural biology

Tail fiber

Tail sheath Tail tube

Tailspike

Tape measure (protein)

Teichoic acid

Terminase proteins

Transmembrane domain Triangulation number Uncoating Unique vertex

Definition loop, formed by the nucleotides between the two complementary sequences Science that studies the ultrastructure of biomacromolecules and their assemblies and complexes with ligands, for example, phages, phage proteins, and their complexes with host cell components Fibrous structures attached to the bottom of a phage tail that are responsible for host cell binding. Many phages may contain two kinds of tail fibers, long tail fibers for primary, reversible, host cell interaction and short tail fibers for secondary, irreversible binding The contractile outer layer of the double-layered tail of phages from the Myoviridae family The inner tube of the double-layered tail of phages from the Myoviridae family. Phages from the Siphoviridae family only have a tail tube, without the contractile sheath Stubby extensions to the tails of many members of the Caudovirales order that bind host receptor molecules (often oligosaccharides) and in many cases hydrolyze them to allow the phage tail access to the host cell membrane Phages of the Siphoviridae and Myoviridae families have copies (presumably six) of a long α-helical tape measure protein spanning the length of their tail tubes. They appear to regulate tail length. PRD1, a phage of the Tectiviridae family, has tape measure proteins (P30) spanning the inside of its icosahedral capsid. P30 is proposed to regulate capsid size Anionic glycopolymer on the outside of most Gram-positive bacteria. Gram-positive bacteria lack an outer membrane. Instead, their peptidoglycan layer is modified with teichoic acid chains (wall teichoic acid). Lipoteichoic acid, anchored in the membrane, is usually also present. Teichoic acids are copolymers of glycerol or ribitol phosphate and carbohydrates Terminase proteins bind to the special vertex of a phage head for DNA packaging. They couple ATP hydrolysis to DNA translocation into the preformed capsid Part of a protein, usually an α-helix, that traverses the lipid bilayer Number that describes the arrangement of protein subunits into a symmetric icosahedral or prolate virus capsid Removal of protein and/or membrane that covers the genetic material of a virus upon entering a host cell See special vertex

The Caudovirales order contains three different families of tailed bacteriophages, all with an icosahedral or prolate capsid along with a linear double-stranded DNA genome. The Myoviridae (Fig. 1g) have a long contractile tail; the Siphoviridae (Fig. 1h) a long, flexible tail; and the Podoviridae (Fig. 1i) a short tail. The outer tail sheath of a myovirus contracts upon infection to drive the inner tail tube through the

Structure and Function of Bacteriophages

27

bacterial outer membrane and peptidoglycan layer and delivers the phage DNA directly into the cytoplasm. In the case of a podovirus, core proteins located inside the capsid pass through the short tail to form a protective tube traversing the periplasmic space, again allowing safe passage of the phage DNA. Siphoviral DNA transfer has not been studied that well. The Plasmaviridae are a family of membrane-enveloped viruses that infect bacteria without a cell wall. Only one species, L2, is known. It has a genome consisting of a 12 kb molecule of circular, supercoiled double-stranded DNA. It is pleomorphic, i.e., of variable shape, but no detailed studies about the structure have been reported, and they will not be further discussed here. Because many bacteriophages are relatively easy to produce, are innocuous to the experimenter, and give valuable general information about virus biology, they have been studied extensively by structural biology (see Box 1). In this chapter, we explain what is known about how the phage particles are assembled, what the final infectious particles look like in atomic detail, and how they are designed to efficiently infect the next available host bacterium. Box 1: Structural Biology

Structural biology is dedicated to the structural determination of biological molecules and complexes. The main techniques employed are electron microscopy, X-ray crystallography, and NMR spectroscopy, which can lead to highresolution atomic models. Small-angle X-ray scattering (SAXS) and atomic force microscopy (AFM) should also be mentioned. SAXS is used to obtain envelopes in solution (Korasick and Tanner 2018), and AFM allows the study and manipulation of individual molecules (Moreno-Madrid et al. 2017) but is limited to low resolution. Negative staining electron microscopy is a relatively easy method to obtain two-dimensional images of a phage at resolutions of 5–10 nm, often allowing a rapid and reliable classification of the phage family (Ackermann and Prangishvili 2012). Obtaining many two-dimensional negatively stained electron microscopy images allows three-dimensional reconstruction, although distortions introduced by the stain limit the attainable resolution to 1–2 nm. Stainless cryo-electron microscopy avoids these distortions, and recent developments in detector technology and image processing have allowed a resolution revolution (Kühlbrandt 2014; Henderson 2015). Three-dimensional reconstruction of electron density maps of biological macromolecules, including bacteriophages, to resolutions where reliable atomic models can be built (0.2 to 0.5 nm), is now possible (see chapter ▶ “Detection of Bacteriophages: Electron Microscopy and Visualization”). X-ray crystallography is the oldest and still most-used technique to obtain atomic models of biological macromolecules. Well-diffracting crystals need to be obtained, which is not always straightforward or possible. Furthermore, the “phase problem” needs to be solved for accurate calculation of electron density (continued)

28

M. Sanz-Gaitero et al.

Box 1: (continued)

maps (Llamas-Saiz and van Raaij 2013). Electron density maps are calculated by summing three-dimensional wave functions. These wave functions have a direction, an amplitude, and a phase, all necessary to sum them correctly. While the direction and the amplitude can be inferred from the diffraction pattern, the phase cannot and has to be estimated using special techniques involving related protein structures and/or heavy atom derivatives (Taylor 2010). Nevertheless, if these conditions are met, it leads to high-quality atomic models, hence its popularity. Several small bacteriophages have been crystallized, for example, the levivirus MS2 (Golmohammadi et al. 1993), allowing structure solution by X-ray crystallography (at 0.3 nm resolution or better). More often, however, structural proteins of bacteriophages have been crystallized outside the context of an intact phage particle. Usually, this means that higher resolution can be attained (0.1–0.2 nm). The resulting structures can be fitted into lower-resolution electron microscopy maps, allowing for atomic representation of entire phages or phage organelles like tails or baseplates (Taylor et al. 2016). X-ray fiber diffraction is useful for determining helical parameters and structural modeling (Marvin 2017). NMR spectroscopy (Higman 2013) is a very powerful technique for solving detailed structures and especially the dynamics of monomeric proteins up to around 30 kD in size, but more difficult for the generally larger and multimeric structural proteins of bacteriophages. Solid-state NMR spectroscopy has been used to solve the structure of magnetically aligned inovirus particles (Thiriot et al. 2004).

The Leviviridae Family of Single-Stranded RNA Phages The Leviviridae are spherical viruses with a genome consisting of a 4 kb linear single strand of RNA. The first complete genome sequenced was that of the levivirus MS2 (Fiers et al. 1976). Leviviruses gain entrance to the host cell cytoplasm via attachment to surface pili structures (Bollback and Huelsenbeck 2001). Their capsids have an icosahedral structure with a triangulation number T = 3 (for an explanation of triangulation numbers, see Box 2). Leviviridae genomes only encode four proteins: a maturation protein or minor capsid protein that is involved in host cell recognition, a coat protein that is the major capsid protein, an RNA-dependent RNA replicase subunit, and a lysis protein (in case of Levivirus) or a read-through protein in case of Allolevirus (Bernhardt et al. 2001). The Allolevirus read-through protein is the result of occasional read-through of the capsid protein stop codon. Each Allolevivirus is estimated to contain three to ten copies of the read-through protein (replacing the normal, shorter, coat protein). The reading frame for the lysis protein of leviviruses overlaps the coat protein and replicase genes in a different reading frame. In the case of Allolevivirus, the maturation protein is also responsible for the lysis function.

Structure and Function of Bacteriophages

29

Assembly The capsids of Leviviridae co-assemble with their genomic RNA molecules (Stockley et al. 2016). Each RNA molecule contains multiple packaging sites and specific sequences that recruit coat proteins. Folding of the RNA molecule into its secondary and tertiary structure then helps to orient coat protein molecules relative to each other and form the icosahedral capsid. Contacts between coat protein molecules probably also play an important role, but assembly is initiated by the genomic RNA molecule (Fig. 2). An RNA sequence at the start of the replicase subunit gene has a high affinity for the coat protein and initiates assembly (Rumnieks and Tars 2014). This interaction is phage-specific (i.e., the coat protein of a phage will only interact with the correct sequence in its own genome), so that only the right genome is coated with its protein. Additional coat protein dimers interact with A-X-X-A sequences in the loops of stem-loop structures of the RNA (A stands for adenosine, X for any nucleoside). This process is driven by the specific three-dimensional structure of the RNA molecule. Stem-loops just upstream and downstream of the putative initiator stem-loop are near each other in the structure, with most of the other well-ordered stem-loops also at the same side of the virus (Dai et al. 2017). Leviviridae do not encode scaffolding proteins. Instead, the folded genomic RNA molecule functions as an internal scaffold, around which the capsid is assembled. In the final structure, the RNA molecule shows some flexibility, while the capsid is wellordered. This suggests that the RNA molecule orients the coat protein dimers in roughly the right positions, avoiding off-pathway octahedral assemblies (Plevka et al. 2008). Protein-protein contacts between coat protein dimers then take care of the precise relative orientations, leading to a well-ordered icosahedral coat protein structure.

Fig. 2 Assembly of the levivirus MS2. (a) Conversion of the symmetric CC coat protein dimer to an asymmetric AB dimer through interaction with an RNA hairpin loop. (b) Threefold and fivefold symmetric capsid formation intermediates formed by combinations of asymmetric AB dimers and symmetric CC dimers interacting with genomic RNA (not shown). These intermediates can further assemble and grow into complete T = 3 capsids. (c) Three-dimensional model of the viral RNA (yellow). Part of the protein capsid is shown in cyan; the maturation protein is shown in red, from a model in the supplementary information of Dai et al. (2017)

30

M. Sanz-Gaitero et al.

Structure The Leviviridae capsid contains 178 copies of the coat protein plus a single copy of the maturation protein (Gorzelnik et al. 2016). Each coat protein adopts one of three conformations (A, B, or C) (Golmohammadi et al. 1993) and organizes into AB and CC dimers in the capsid (Fig. 3). The α-helical part of the maturation protein takes up the position of one of the CC dimers of the coat protein in the otherwise icosahedrally symmetric T = 3 capsid. The β-structured part of the maturation protein sticks out into solution and is likely the domain that recognizes the host receptor. The insertion of the maturation protein in the capsid opens a gap right next to it; this probably helps the RNA to leave the capsid and enter the host bacterium (Gorzelnik et al. 2016). The central part of the MS2 coat protein (Ni et al. 1995) is a five-stranded antiparallel β-sheet that faces the interior of the virus (Fig. 3a). The amino-terminal region forms a β-hairpin that covers part of the β-sheet on the outside, while the carboxy-terminal end is helical and makes extensive interactions with the neighboring monomer, covering the part of its β-sheet not covered by its own β-hairpin. The

Fig. 3 Structure of the levivirus MS2. (a) Structure of the coat protein dimer viewed from the outside of the capsid (PDB entry 1MSC). Protein chains are colored green and cyan. Amino-termini (Nt) and carboxy-termini (Ct) are indicated. (b) Structure of the maturation protein (PDB entry 5TC1), viewed from the outside of the capsid (red). (c) Structure of the MS2 virus viewed down the fivefold symmetry axis (PDB entry 1MS2). The capsid has triangulation number T = 3. (d) Structure of the MS2 virus viewed down the threefold symmetry axis. AB trimers are in green/ cyan, CC trimers in magenta. PDB is the Protein Data Bank (https://www.ebi.ac.uk/pdbe/)

Structure and Function of Bacteriophages

31

resulting structure is a stable, interlocked dimer. Dimerization is also favored by the β-sheets of the two monomers aligning in such a way that a shared antiparallel, ten-stranded β-sheet results. The Qβ phage coat protein (Allolevirus genus) has an identical fold, although the sequence identity is only 25% (Golmohammadi et al. 1996). The coat proteins are assembled so that five AB dimers surround the fivefold symmetry axes (Fig. 3c). Three AB dimers and three CC dimers alternate around the threefold axes, making them pseudo-sixfold. Finally, the twofold symmetry axes coincide with the twofolds of each CC dimer. The maturation protein consists of an α-helical domain and a β-sheet domain (Fig. 3b). The α-helical domain contains a bundle of six α-helices. The β-sheet domain has six antiparallel strands with an α-helix on the side and a helix-loop-helix motif covering the sheet on the outside of the virus (Dai et al. 2017). The α-helical domain is inserted into the protein capsid, while the β-sheet domain is exposed, presumably ready to interact with the bacterial pilus. The single-stranded genomic RNA in the virus has a complicated secondary and tertiary structure (Fig. 2c). There are many short-range interactions, leading to most of the RNA molecule being double-helical. Long-range tertiary interactions (base pairing or kissing loops) are also present and important for the RNA molecule adopting its final shape. In total, most of the bases are involved in interactions with other nucleotides. Exceptions are the loop regions on the outside, which contact the coat protein capsid instead. The spherical shape of the genomic RNA is probably partially imposed by the RNA structure itself and partially by the coat protein. The maturation protein interacts with an RNA stem-loop region at the 30 end of the genome. It has been proposed that infection takes place when the pilus retracts, with a virus particle bound to it. The virus, being too big to enter the host, gets stuck on the surface, while the maturation protein is pulled inside, together with the bound genomic RNA (Dai et al. 2017). Box 2: Symmetry in Viruses

Triangulation numbers of icosahedral capsids. Icosahedra can be thought to be made up of pentagonal tiles at each of the 12 vertices and a variable number of hexagonal tiles on the faces. A T = 1 capsid has no hexagonal tiles, while a virus with a high triangulation number has many hexagonal tiles. To determine the triangulation number, T, one starts at a fivefold symmetry axis (i.e., in the middle of a pentagonal tile) and jumps via the shortest route to the next pentagonal tile, via the hexagonal tiles (if they are present). Straight jumps count for h and jumps to the side for k. For a T = 1 capsid, only one jump is necessary, so h = 1 and k = 0. For a T = 13laevo capsid, three straight jumps and one jump to the left are needed, so h = 3 and k = 1 (the mirror image of a T = 13laevo capsid would be a T = 13dextro capsid). The triangulation number is given as T = h2 + hk + k2 (Fig. 4a). The figure illustrates some examples. Further reading can be found in (Caspar and Klug 1962) and (Johnson and Speir 1997). (continued)

32

M. Sanz-Gaitero et al.

Fig. 4 Symmetry in viruses. (a) Triangulation numbers of icosahedral capsids. T = 13 l stands for T = 13laevo. (b) Triangulation numbers of prolate capsids. (c) Parameters of helical structures

Box 2: (continued)

Triangulation numbers of prolate capsids. Although the caps of prolate (elongated) heads can be described with triangulation numbers as above, the midsection is elongated and made up of nonsymmetric triangles. Calculating the triangulation number Q of the facets with uneven sides now needs h1, h2, k1, and k2, with Q = h1h2 + h1k2 + k1k2. In the case of bacteriophage T4, h1 = 3, h2 = 1, k1 = 4, and k2 = 2 (Fig. 4b). In this case, for the caps, the triangulation number T equals 13. For the side facets, the triangulation number Q equals 20. For a more thorough explanation, see Prasad and Schmid (2012). Parameters of helical structures. Helices can be right-handed (clockwise screw moving away from the observer looking along the helix) or left-handed (anticlockwise screw moving away from the observer looking along the helix). The helical pitch P is defined as the distance that the helix travels in one complete turn (Fig. 4c). A helix can consist of more than one intertwined helix of subunits, i.e., have more than one start. For instance, a dsDNA helix can be considered a two-start helix. The twist or tilt angle can be defined as the number of degrees the helix turns between subunits.

Structure and Function of Bacteriophages

33

The Microviridae Family of Single-Stranded DNA Phages The Microviridae are a family of small, icosahedral, bacteriophages with a plusstranded, circular DNA genome (Doore and Fane 2016). They initiate infection via attachment to host cell lipopolysaccharide molecules (Inagaki et al. 2003). The genome is around 5 kb and codes for 11 gene products: A (DNA replication protein), A, B (internal scaffolding protein), C (DNA replication protein), D (external scaffolding protein), E (lysis protein), F (coat protein), G (spike protein), H (DNA pilot protein), J (DNA-binding protein), and K. Genes A, E, and K are nonessential when the virus is grown in the lab. The A protein results from internal translation in gene A, in the same frame as the parent protein. Microvirus capsids have a triangulation number T = 1.

Assembly The assembly of microviruses can be divided into early and late stages (Doore and Fane 2016). Early stage assembly is mediated by the internal scaffolding protein B. Five copies of the small α-helical B protein bind with their carboxy-terminus to the underside of a capsid protein F pentamer, recruiting one copy of the H protein per pentamer and inducing a conformational change (Fig. 5). The resulting F5B5H1 particle then recruits a spike protein G pentamer, which binds on top of the F pentamer, forming the G5F5B5H1 particle (Fig. 5). Late stage assembly is mediated by the external scaffolding protein D; 240 copies of the α-helical D protein organize 12 G5F5B5H1 particles into procapsids. Four copies of the D protein are arranged in two distinct asymmetric dimers (D1D2 and D3D4), each of which contacts a coat protein F molecule. The D1D2 and D3D4 dimers contact each other in the complex, making a tetramer. Each structural conformer of the D tetramer establishes different interactions with the coat protein beneath it and the D proteins next to it (Prevelige and Fane 2012). The lattice formed by the 240 copies of the D protein holds the procapsid together, allowing divisions between the coat protein pentamers at the threefold symmetry axes, forming 3 nm

Fig. 5 Assembly of the microvirus ϕX174. Pentamers of the F coat protein form spontaneously and bind five copies of the internal scaffolding protein B on the inside and a pentamer of the spike protein G on the outside. A single copy of the pilot protein H also binds to the inside. The external scaffolding protein D helps to organize the complexes into icosahedral procapsids, with pores large enough to allow DNA entry (PDB entry 1CD3). After DNA entry, the scaffolding proteins leave and allow the collapse of the capsid into a stable virus particle without any visible pores (PDB entry 1RB8)

34

M. Sanz-Gaitero et al.

pores (Fig. 5). These pores are necessary for DNA entry and for the exit of the internal scaffolding protein B. The single-stranded DNA is synthesized and transferred into the procapsids by the DNA packaging complex, which binds the procapsid at a twofold symmetry axis. This complex includes the bacterial host protein Rep and the viral proteins A and C. Along with the genome, 60 molecules of the small DNA-binding protein J enter the capsid and displace the internal scaffolding protein B (Bernal et al. 2003). Protein B displacement triggers its auto-proteolytic activity, targeting three Arg-Phe motifs at the carboxy-terminal end (positions 77, 93, and 109 in the case of ϕX174), which enables its escape through the 3 nm pores. The release of the internal and external scaffolding proteins facilitates the collapse of the 12 G5F5B5H1 tiles around the genome, resulting in mature virions without any major gaps (Fig. 5). Microvirinae such as ϕX174 are characterized by utilizing a dual scaffolding protein system, although other subfamilies of the Microviridae only have the internal scaffolding protein (Doore and Fane 2016).

Structure Microvirus virions consist of a T = 1 icosahedral capsid of 25 nm in diameter. The capsid is formed by 12 pentamers of coat protein F. Each fivefold symmetry axis is decorated with a pentamer of the spike protein G, which acts as receptor-binding protein. Both protein F and protein G share an eight-stranded antiparallel β-barrel core structure with BIDG-CHEF topology (Fig. 6). This β-barrel jelly roll fold is common to many small plant and animal viruses, forming the Picorna-like virus structural lineage (Abrescia et al. 2012). The F protein has two extensive insertion loops absent from protein G (Doore and Fane 2016), extending the shape of the protein into a triangle suited to cover the icosahedral viral surface and covering the top of the β-barrel with several α-helices. The β-barrels of protein G associate into a pentamer, with the BIDG-sheets interacting with each other and the CHEF-sheets on the outside. Long AB- and EF-loops decorate the CHEF-sheet surface. The structural similarity between the F and G proteins suggests they may have originated by a gene duplication event. Inside the capsid, the microvirus genome is associated with 60 copies of the positively charged DNA-binding protein J, which also attaches to the capsid protein F (Bernal et al. 2003). The virion also contains between 10 and 12 copies of the DNA pilot protein H. Protein H helps the DNA to get into the host cell, forming a decameric α-helical coiled-coil barrel (Fig. 6). This tube spans the periplasmic space of the host bacterium, enabling the translocation of the DNA into the cytoplasm (Sun et al. 2014). In the free virion, protein H has been suggested to be present as a monomer near the fivefold symmetry axis. It is not known if the coiled-coil barrel is pre-formed in the viral capsid just before DNA ejection or if it only forms once the H protein molecules traverse the outer membrane. Apart from the Microvirinae, a member of the Gokushovirinae (SpV4, infecting Spiroplasma bacteria) is the only other microvirus that has been structurally characterized (Chipman et al. 1998). Gokushovirinae lack the spike protein. Instead, their F coat proteins have a sequence insertion, with the insertions of three neighboring

Structure and Function of Bacteriophages

35

Fig. 6 Structure of microviruses. (a) Structure of the F coat protein, a jelly roll β-barrel with insertion leading to a triangle-shaped protein. Amino-termini (Nt) and carboxy-termini (Ct) are indicated. (b) Structure of the G spike protein, which also consists of a jelly roll, but with less decoration. (c) Topology diagram of the jelly roll β-barrel fold. (d) Overall structure of ϕX174 (PDB entry 1RB8). The coat protein F is shown in green, the spike protein G in blue, and partial structures of the internal H protein are in magenta. (e) Model of the Spiroplasma virus SpV4, a member of the Gokushovirinae, in the same orientation as in panel D (PDB entry 1KVP). Note the absence of the G spike protein at the fivefold symmetry axes; instead, insertions in the F sequence lead to symmetric protrusions at the threefold symmetry axes. (f) Membrane tube formed by the pilot protein H to allow DNA transfer into the bacterial cytoplasm (PDB entry 4JPP). In the crystal, the protein forms homodecameric tubes as shown; the exact stoichiometry in vivo is unknown. One of the monomers is shown in yellow, the other nine in red

coat protein molecules leading to protrusions at the threefold axes (Fig. 6). These trimeric protrusions presumably play the same receptor-binding role that the spike protein G does in the Microvirinae (Doore and Fane 2016). Box 3: A Short Note About Phage Protein Nomenclature

Phage protein nomenclature, and virus protein nomenclature in general, can be bewildering. It changes between phage families, different phages in the same family, and even between research groups working on the same phage. Phage (continued)

36

M. Sanz-Gaitero et al.

Box 3: (continued)

proteins are named with letters (A, B, C, etc.) or numbers (I, II, III, etc.), as gene products (gp1, gp2, gp3, etc. or gpA, gpB, etc.), with the indication p or P (p1, p2, etc. or P1, P2, etc.), with a name or with other denominations. Here, we have tried to use the nomenclature that is more generally accepted for each phage family.

The Inoviridae Family of Filamentous, Single-Stranded DNA Phages The Inoviridae are a family of long, thin, filamentous bacteriophages containing a positive sense, single-stranded, circular DNA genome (Fig. 7). They reproduce without killing their host, instead causing a chronic infection (Rakonjac et al. 2017). There are seven known genera, of which Inovirus is the best known. Inoviruses are long flexible filamentous viruses, measuring about 7 nm in diameter and 1 μm in length, with different types infecting Gram-negative and Gram-positive bacteria. The type species is Enterobacteria phage M13, but f1 and fd are also wellknown. Their genomes are between 4 and 9 kb in length. The Escherichia coli Ff phages (for F-pilus-specific filamentous phage) f1, fd, and M13 are genetically more than 98% identical. They have been useful for molecular biology and biotechnology applications, due to the ease with which their genome can be modified. Unlike for most other viruses, capsid size is not strictly limiting, because larger DNA molecules can generate longer viruses. Up to 12 kb of foreign DNA can be inserted in the viral genome (Marvin 1998). They also

Fig. 7 Overall structure and assembly of the Inoviridae. An inovirus part-way through assembly (left) and a mature inovirus are depicted (right), separated by a horizontal arrow. The circular singlestranded DNA is drawn as a black bar. Most of the phage-encoded proteins, the structural proteins p3, p6, p7, p8, and p9, and the nonstructural proteins p1, p4, p5, and p11 implicated in the process are indicated

Structure and Function of Bacteriophages

37

provide a convenient way to generate specific single-stranded DNA molecules, which used to be essential for efficient DNA sequencing (see chapter ▶ “Bacteriophage Use in Molecular Biology and Biotechnology”). They have also been used as vectors for gene transfer and vehicles for phage display. In phage display, peptides are presented on the virion surface fused to the viral coat proteins, allowing their interaction with other molecules and the selection of strong binders from libraries containing many variants (see chapter ▶ “Bacteriophages in Nanotechnology: History and Future”). Their non-lytic infection mechanism allows the long-term maintenance of bacterial clones producing mutant viruses. Below, we will discuss the assembly and the structure of the E. coli Ff phages as an example for all members of the Inoviridae family.

Assembly Filamentous phages recognize the host pilus via the receptor-binding protein p3. The p3 protein has three domains: N1, N2, and C1. The carboxy-terminal domain C1 is integrated into the virion. The second amino-terminal domain, N2, is the one responsible for interaction with the pilus (Lubkowski et al. 1999). After retraction of the pilus, the first amino-terminal domain N1 of p3 interacts with the TolA protein. The Tol complex then mediates close contact between the bacterial outer and inner membranes, allowing DNA transfer from the phage directly into the cytoplasm (Karlsson et al. 2003). The single-stranded DNA acts as a template for negative strand synthesis. This is a process independent of phage proteins, initiated by the host RNA polymerase. The RNA polymerase synthesizes a primer that is used by the host DNA polymerase III to synthesize the negative strand, obtaining a circular double-stranded DNA, which replicates by a rolling circle mechanism (Higashitani et al. 1997). The dsDNA form of the genome is known as the replicative form, while the single-stranded plus-chain is known as the infective form. To convert the replicative form to the infective form, the p2 protein is necessary (Rakonjac et al. 2011). As soon as the ss(+)DNA chain is synthesized, it is covered with dimers of p5, which collapse the circular single strand into a flexible rod of about 8 nm in diameter, preventing the synthesis of the complementary strand (Marvin 1998). The finalized p5-DNA complex forms a left-handed helix with the DNA wrapped inside the protein. In the p5-DNA complex, the only exposed zone of the genome is the packaging signal, which is a hairpin loop. To initiate assembly, the packaging signal interacts with the assembly machinery in a sequence-specific way. The structure of p5 has been determined (Fig. 8a) and revealed a β-structured dimer with a positively charged side and a negatively charged side (Su et al. 1997), but exactly how the p5 dimers associate with the single-stranded DNA is not known yet. Assembly takes place in the cytoplasmic membrane, and nascent virions are secreted from the cell as they assemble (Fig. 7). The eight phage-encoded proteins involved in assembly, including three proteins not present in the virion (p1, p4 and p11), but also the five proteins forming the virion coat (p3, p6, p7, p8, and p9), all have a transmembrane domain and are inserted in the cytoplasmic membrane before phage assembly (Rakonjac et al. 2017). Phage assembly occurs at distinct membrane

38

M. Sanz-Gaitero et al.

Fig. 8 Structure of inovirus proteins. (a) Structure of the p5 dimer (PDB entry 1GVP). Monomers are colored light and dark blue. The putative site of DNA binding is indicated. Amino-termini (Nt) and carboxy-termini (Ct) are indicated. Model of part of the helical inovirus Pf1 capsid formed by the major capsid protein p8 (PDB entry 2C0W) seen from the side (b) and from the end (c). While most copies of the p8 protein are shown in purple, a helical turn of five p8 molecules is shown in green. The orientation in panel B is the same as in Fig. 6, i.e., the p3 receptor-binding protein and the p6 protein would be on the left and the p7 and p9 proteins on the right. In panel C, p7 and p9 would be in the front and p6 and p3 in the back. (d) Structure of the p3 N1 and N2 domains (PDB entry 2G3P). The p3 N1 domain is shown in green, the N2 domain in magenta. (e) Structure of the p3 N1 domain bound to the carboxy-terminal domain of the inovirus co-receptor TolA (PDB entry 1TOL), shown in brown

assembly sites, at regions where the cytoplasmic and outer membranes are in close contact. The packaging signal located in the hairpin loop of the ssDNA-p5 complex is recognized by p7, p9, and p1, initiating the assembly. The assembly site is a transmembrane complex formed by the phage-encoded membrane proteins p1, p4, and p11 (Fig. 7). Proteins p1 and p11 are embedded in the cytoplasmic membrane, forming a multimeric complex composed of five or six copies each. Their carboxyterminal domains protrude into the periplasm and contact the outer membrane protein p4. The protein p4 integrates into the outer membrane, forming a barrelshaped homo-multimer composed of 12 to 14 subunits with a central cavity measuring 8 nm in diameter, enough to allow the assembled phage to pass through. The cytoplasmic amino-terminal domain of p1 contains a DNA-binding motif necessary for phage assembly. Phage assembly requires ATP hydrolysis (Feng et al. 1999). The protein p1 may also be necessary for the formation of the adhesion zones between the inner and outer membranes where assembly takes place (Russel et al. 1997). The host thioredoxin is also required for correct phage assembly and acts as a DNA-handling protein, not as a redox enzyme (Marvin et al. 2014).

Structure and Function of Bacteriophages

39

The small proteins p7 and p9 cap the leading, blunt end of the secreted phage. The ssDNA starts to traverse the membrane through the assembly site, causing p5 to dissociate and to be replaced by major coat protein p8. When the DNA is completely coated with p8, the minor coat proteins p6 and p3 are added, resulting in the release of the assembled phage (Marvin et al. 2014). If one of the p3, p6, p7, or p9 proteins is absent, the phage continues to elongate and stays bound to the membrane (Rakonjac et al. 2011). The p3 and p6 proteins cap the trailing, sharp end of the virion. The protein p6 is small, but p3 is a large multi-domain protein that functions as the receptor-binding protein, entering first during infection and leaving last during assembly.

Structure Inovirus virions are around 7 nm in diameter. The exact length is determined by the size of the genome (normally 5–8 kb of single-stranded DNA) and is in the order of 1 μm. The single-stranded circular DNA molecule is protected by a long cylindrical protein coat made of thousands of copies of the major coat protein p8, a small protein of only around 50 amino acids (Wang et al. 2006). This protein forms a tube around the DNA in an overlapping helical array, with the amino-terminal end of p8 located at the outside of the coat and its positively charged carboxy-terminal end at the inside. These positively charged residues interact with the DNA. The organization of the DNA is unknown; conflicting models for it have been presented (Marvin et al. 2014). What is known is that the DNA is circular, so it is flattened, and one strand goes upward in the phage and the other down. Except for five disordered negatively charged surface exposed amino-terminal residues, each p8 subunit forms a single, continuous α-helix. The central domain of p8 is hydrophobic, allowing the protein to interact with its neighboring subunits (Rakonjac et al. 2011). Different models have been proposed for the helical array of p8 subunit, based on fiber diffraction and solid-state NMR spectroscopy (Marvin et al. 2014). Cryoelectron microscopy has also been performed (Wang et al. 2006), but did not lead to an exact model. This is likely due to the difficulty in correctly averaging flexible particles and the presence of structural transitions in the particle. It has been shown that temperature affects the fiber diffraction spectra and thus the structure of the virus. In Fig. 8, a model for the fd phage based on X-ray fiber diffraction and solidstate NMR spectroscopy is shown (Marvin et al. 2006). This model is a right-handed five-start helix with a rise of about 1.6 nm and a pitch of 16 nm; there are ten subunits in a complete turn. The two ends of the virion are capped by three to five copies of each of the four minor capsid proteins, specific for infection and virus assembly (Fig. 7). The small p7 and p9 proteins are located at the end that is extruded first from the bacterial cell. The proteins p3 and p6 are located at the end that is extruded last (Marvin 1998). Once incorporated into the virion, p3 is required for the stability of the proximal end of the p8 array, and p6 is necessary for the incorporation of p3 into the virion. These minor proteins have apolar domains in their primary structure similar in length to the hydrophobic domain of p8, allowing for association with each other and with p8.

40

M. Sanz-Gaitero et al.

Crystal structures of the host cell interaction domains N1 and N2 of p3 are known (Holliger et al. 1999; Fig. 8). It is also known how the N1 interacts with the carboxyterminal D3 domain of the inovirus co-receptor TolA (Fig. 8), but not exactly how the N2 domain binds to the pilus. In fact, different filamentous phages may bind to different pili, which are correlated to different N2 structures. In phage fd, it appears that the N2 domain shields the N1 domain from interacting with TolA before N2 binds to the F-pilus, while in phage IF1, binding of N1 to TolA is not shielded before N2 binds to the the I-pilus.

The Cystoviridae Family of Double-Stranded RNA Phages The Cystoviridae are a family of enveloped viruses with a diameter of about 85 nm. Virions have a double capsid structure with an external membrane. Embedded in the membrane are trimeric spikes that decorate the particle. The genome consists of three double-stranded RNA segments, ranging from over 6 to just under 3 kb. Each genome segment encodes several proteins. So far, only one genus, Cystovirus, has been identified. The type species is ϕ6, but species ϕ7 through to ϕ13, ϕ2954, ϕNN, and ϕYY have also been identified. All known cystoviruses infect Pseudomonas species, although ϕ8 also infects E. coli and other hosts. In most cases, Pseudomonas bacteria pathogenic to plants are the host, although ϕNN has been isolated from lake water and ϕYY from hospital sewage (Mäntynen et al. 2018). Cystoviruses consist of a double-shelled capsid and an external membrane envelope. The largest RNA segment encodes the nonstructural protein P14 and the RNA-dependent RNA polymerase complex proteins P1, P2, P4, and P7; the medium-sized RNA segment codes for the membrane proteins P3, P6, P10, and P13; and the smallest RNA segment codes for the lytic protein P5 and the nucleocapsid protein P8, the nonstructural protein P12, and the membrane protein P9.

Assembly Cystoviruses commence infection by the trimeric P3 protein attaching to host cell pili, for example, ϕ6 (Bamford et al. 1976), or rough lipopolysaccharide, for example, ϕ8, ϕ12, and ϕ13 (Hu et al. 2008). The fusogenic P6 protein mediates fusion of viral membrane with the bacterial outer membrane, releasing the capsid into the periplasmic space. The P5 capsid protein then digests the peptidoglycan layer of the host, leading to endocytosis of the capsid into the cytoplasm, now covered by the host inner membrane. This membrane and the P8 protein shell are subsequently lost, but the rest of the capsid remains intact, hiding the viral genome from antiviral host factors. Polycistronic mRNAs are transcribed by the viral RNA-dependent RNA polymerase, P2, found inside the core particle and released into the cell cytoplasm. Newly synthesized proteins assemble into capsids. The P4 packaging protein translocates three plus-stranded RNA segments into the capsids, where they are converted into dsRNA by the P2 protein. Capsid maturation involves enveloping by the viral membrane (including its associated proteins), after which the host cell is lysed and mature virions are released. Presumably, free

Structure and Function of Bacteriophages

41

P5 protein serves as the phage endolysin necessary for host lysis (Caldentey and Bamford 1992). The best-studied cystovirus in terms of assembly is ϕ6 (Poranen and Tuma 2004). The assembly of phage ϕ6 consists of the generation of a procapsid, the packaging and replication of its RNA genome, the joining of the outer shell of the nucleocapsid, and the formation of its lipid envelope. Its major capsid protein, P1, when expressed in vitro, forms spherical particles and even dodecahedral cages, although they are unstable. In the presence of P4 (the single-stranded RNA packaging protein), P1 forms more stable and regularly ordered dodecahedral particles, containing 120 copies, and with P4 hexamers on the outside of the fivefold symmetry axes (Fig. 9). P4 also allows assembly at much lower concentrations of P1, suggesting it probably is important for nucleation of the assembly. Five P1 dimers assemble around the P4 hexamer. P1 dimers interact with each other to form a collapsed dodecahedral procapsid. This is facilitated by the P7 assembly cofactor which is incorporated, together with P2, at the threefold symmetry axes of the procapsid (Sun et al. 2017). A monomer of the RNA-dependent RNA polymerase, P2, associates with the assembling shell and ends up on the inside, near a threefold symmetry axis. An estimated 60 copies of P7 associate with the T = 1 P1 shell. P7 has a regulatory function in assembly and RNA packaging. The assembled procapsid has a deflated shape with deeply recessed vertices, giving the appearance of a dodecahedron (Nemecek et al. 2013). The hexameric P4 NTPases package a small (3 kb), a medium (4 kb), and a large (6 kb) positive-strand RNA molecule into the capsid sequentially (Frilander and Bamford 1995). The RNA packaging reaction is dependent on magnesium ions and requires a nucleoside triphosphate as an energy source. The packaging of the small and medium segments is efficient when they are packaged on their own, but the packaging of the large segment is very inefficient alone and appears to be dependent on the medium segment being packaged first. P2, the RNA-dependent RNA polymerase, synthesizes the complementary strand, generating the three double-stranded genome segments, some of which migrate from the threefold to the fivefold symmetry axes (Oliveira et al. 2018). During RNA incorporation and replication, the procapsid expands to reach its final icosahedral shape, increasing its volume by about 250%. The expansion of the procapsid is principally due to changes in interdimer interactions, especially at the P1B subunit.

Fig. 9 Assembly of the nucleocapsid of cystovirus ϕ6. Up to five P1 dimers (light and dark orange) associate with a P4 hexamer (green). The P1 dimers interact to form a deflated dodecahedral procapsid (PDB entry 4BTG). After RNA translocation into the capsid and synthesis of the complementary strand, the procapsid expands. Then 200 trimers of P8 assemble (light and dark blue) onto the P1 shell to complete the nucleocapsid (PDB entry 5MUU)

42

M. Sanz-Gaitero et al.

The nucleocapsid surface protein, P8, can assemble as open, irregular, shell-like structures on its own. However, onto the inflated core, P8 assembles as a regular T = 13 shell (Sun et al. 2017). The assembly is promoted by calcium ions. No phage or host cell assembly factors appear to be necessary to yield an infectious nucleocapsid particle. Around the complete nucleocapsid particle, a lipid envelope assembles afterward. The phage membrane protein P9 and the nonstructural phage protein P12 are necessary for this. The lipids of the membrane are derived from the host cytoplasmic membrane, although the assembly of the envelope occurs in the cytoplasm and does not involve budding (Poranen and Tuma 2004). The host recognition and attachment protein, P3, which is anchored to the envelope through protein P6, is expressed as a soluble protein and is the last component to be assembled onto the virions.

Structure The cystovirus internal capsid is made up of 60 dimers of the P1 protein (Sun et al. 2017), in a T = 1 geometry. These dimers are asymmetrical; they are formed by two subunits of the P1 protein (P1A and P1B) with the same α-helical fold but with small differences in tertiary structure (Oliveira et al. 2018). The P1 protein is shaped like a trapezoid tile (Fig. 10b). Most residues are in an α-helical conformation, although small β-strands are also present. Differences between the P1A and P1B subunits and between the conformations of P1A and P1B before and after capsid expansion are located mainly in a hinge region in the carboxy-terminal domain, although these differences are small compared to the differences in relative orientations of the protein subunits upon capsid expansion (Nemecek et al. 2013). The subunits of the P1 dimer are stabilized by the C-terminal tail of P4, which explains the relative rigidity of the dimer during expansion of the procapsid (Sun et al. 2017). The fivefold vertices of the P1 icosahedral inner shell are covered by hexamers of the P4 protein. The P4 hexamer is attached to the P1 dimers by up to five of the six P4 carboxy-terminal tails, circumventing the symmetry mismatch of a hexamer binding on a fivefold symmetry axis (Sun et al. 2017). The P4 hexamers are responsible for RNA packaging into the capsid. P4 is an NTPase with a nucleotide-binding Rossmann fold (Fig. 10c). The amino- and carboxy-terminal regions of the P4 protein vary between different cystoviruses, but the central enzymatic domain is conserved (El Omari et al. 2013a). The outer shell of ϕ6 is formed by 200 trimers of the P8 protein arranged as a T = 13 l shell. The P8 protein consists of two α-helical domains (an amino-terminal peripheral domain and a carboxy-terminal core domain) bound by a linker (Fig. 10d). The amino-terminal peripheral domain interacts either with the core domain of its own trimer (closed conformation) or with the core domain of an adjacent trimer (open conformation). The open conformation is the most common (540 out of the 600 P8 subunits of the fully formed P8 shell), while the closed conformation is restricted to one of the monomers of the peri-pentonal P8 trimers to avoid clashes with the P4 hexamer. The open and closed conformations differ in the linker bending, although the interactions between the peripheral and core domains are analogous (Sun et al. 2017). This feature is known as domain swapping and has been described as a mechanism for protein oligomerization (Liu and Eisenberg

Structure and Function of Bacteriophages

43

Fig. 10 (a) Schematic structure of the cystovirus ϕ6. The locations of the structural proteins P1, P2, P3, P4, P6, P7, and P8, the viral membrane, and the large (L), medium (M), and small (S) dsRNA genome segments are indicated. (b) Crystal structure of the P1 capsid protein seen from the outside of the capsid (PDB entry 4K7H). The chain is colored as a rainbow from dark blue (amino-terminal) to red (carboxy-terminal, Ct). (c) Structure of a P4 monomer (PDB entry 4BLO). The chain is colored as a rainbow from dark blue (amino-terminal) to red (carboxy-terminal). The nucleotide (ADP) is shown in black. (d) Arrangement of the P8 trimers in the nucleocapsid. A trimer with all its monomers in open conformation (red) and a trimer with two monomers in open conformation (green) and one in closed conformation (cyan) are shown (PDB entry 5MUU). The amino- and carboxy-terminal ends of one of the red monomers are indicated

2002). In phage ϕ8, the P8 shell appears to be missing (El Omari et al. 2013b). The nucleocapsid of ϕ6 is enclosed by a lipid bilayer, which interacts with P4 (and P8 when present). The membrane contains four viral integral membrane proteins: P6, P9, P10, and P13. In addition, the spike protein P3 is attached to P6 and protrudes 2 nm from the membrane surface (Jäälinoja et al. 2007; Fig. 10a).

Overview of Bacteriophages Containing Double-Stranded DNA Genomes Bacteriophages with double-stranded DNA genomes comprise more than 95% of all currently identified phages. Families for which the assembly and structure have been studied are the Tectiviridae, the Corticoviridae, and those of the Caudovirales order.

44

M. Sanz-Gaitero et al.

The Tectiviridae and Corticoviridae are icosahedral phages with an internal membrane. Phages of the Caudovirales order have icosahedral or prolate capsids and a tail. The Caudovirales are divided into three families with different tail morphologies. The Podoviridae family consists of phages with a short noncontractile tail, and the Siphoviridae family consists of phages with a long flexible tail. The last family, the Myoviridae family, consists of phages with a long, contractile tail. All these bacteriophages appear to have one vertex that is different from the other 11 (although this hasn’t been studied yet for the Corticoviridae). This specialized or unique vertex is where the DNA translocation apparatus attaches temporarily to transfer the phage DNA into the capsid. In the tailed phages, it is also where the tail is attached and where the DNA leaves the capsid to enter the host bacterium.

The Tectiviridae Family The Tectiviridae are icosahedral viruses. There are three known genera; Alphatectivirus, infecting Gram-negative bacteria; Betatectivirus, infecting Grampositive bacteria; and the newly discovered Gammatectivirus GC1 (Philippe et al. 2018). The most well-known tectivirus is PRD1, which has a pseudo T = 25 capsid of around 65 nm in diameter. PRD1 infects several Gram-negative bacterial species containing Inc-type conjugative plasmids, including Salmonella and E. coli. These Inc plasmids encode the phage receptor and DNA translocation machinery. PRD1 has a linear 15 kb double-stranded DNA genome. Upon infection, the inner membrane forms a tube, allowing for DNA transfer directly into the host cell cytoplasm.

Assembly The early operons of the viral DNA newly transferred into the host bacterium direct the production of the terminal protein P8, the viral DNA polymerase P1, and the two single-stranded DNA-binding proteins (P12 and P19) for efficient DNA replication (Butcher et al. 2012). The terminal protein P8 functions as a primer for complementary strand DNA synthesis. It also localizes the DNA replication complex to the bacterial nucleoid. The late operons include genes for the structural proteins and chaperone proteins important for their assembly (Butcher et al. 2012). PRD1 assembly begins with the expression of soluble capsid proteins such as trimers of P3, trimers of P5, and pentamers of P31. At the same time, phage membrane proteins (P7, P11, P14, P16, and P18) are produced and inserted into the cytoplasmic membrane of the host. The proper folding of some viral proteins, such as the major capsid protein P3, requires the action of either the host GroEL-GroES chaperone complex or a complex of GroEL and the viral chaperone P33. Virion assembly is initiated by pinching off a membrane patch containing phage proteins from the host membrane (Fig. 11). This budding-off from the membrane involves both nonstructural phage proteins (membrane bound P10 and soluble P17) and structural phage proteins (P3, P30, P31, and P6). Curvature is introduced into the capsid by the P31 pentamers and the membrane protein P16. Extended dimers of the protein P30, on the inside of the P3/P31 capsid,

Structure and Function of Bacteriophages

45

Fig. 11 Assembly of the tectivirus PRD1. In the left panel, a segment of the host membrane (gray) containing phage proteins is shown, already partially covered with capsid proteins. After completion of the capsid, the membrane segment is pinched off and the unique vertex closed (middle panel, unique vertex at the top). The preformed capsid is now ready for translocating the phage genome into it. After DNA translocation, the lagging copy of the covalently DNA-attached protein P8 plugs the hole in the unique vertex, and the internal pressure pushes the membrane against the inside of the capsid (right panel). Phage proteins are color-coded

are thought to be important for regulating the length of the icosahedral edges (Abrescia et al. 2004). The receptor-binding fivefold vertices are completed with the joining of the spike proteins P5 and P2, whose incorporation is P31-dependent. At the unique vertex, a complex of proteins P20 and P22 is the nucleating site for the assembly of the packaging efficiency factor P6 and packaging ATPase P9 (Hong et al. 2014). The genome is probably recruited to the unique vertex by protein P8, which is linked to the 50 ends of the DNA. ATP hydrolysis by P9 drives the translocation of the genome and the leading copy of P8 through the P20/P22 channel and into the procapsid. After packaging, the pore in the vertex is sealed by the lagging copy of P8, and the increased internal pressure expands the membrane to reach its final shape (Fig. 11).

Structure The mature virion of PRD1 has a mass of 66 MDa and contains about 18 different proteins (Abrescia et al. 2004; Fig. 12a). The external pseudo T = 25 icosahedral capsid shell is formed by 235 trimers of the major capsid protein P3. The P3 monomer contains two jelly roll domains, endowing the trimers with an almost hexagonal shape (Fig. 12b). Four P3 trimers constitute the icosahedral asymmetric unit, except for the five asymmetric units surrounding the unique vertex which consist of three P3 trimers. The amino-terminal and carboxy-terminal extensions of the 12 P3 copies in the regular asymmetric unit adopt different conformations of the amino-terminal and carboxy-terminal extensions, depending on the location within the asymmetric unit. These conformations allow differential interaction with the membrane, other subunits of the trimer, other trimers, the vertex proteins, and the protein P30. P3 trimers are arranged on a framework of 60 copies of the tape measure protein P30. P30 forms extended dimers locked together by amino-terminal hooks. These dimers cement P3 trimers along the icosahedral facet edges and interact with P16 at the vertices (Butcher et al. 2012).

46

M. Sanz-Gaitero et al.

Fig. 12 Structure of the tectivirus PRD1. (a) Overall structure of the icosahedral protein capsid viewed down a threefold axis (PDB entry 1W8X). The double-barrel structure of the P3 monomers is shown in pink and the single-barrel structure of P31 in light blue. P3 forms trimers (pseudohexagons) and P31 forms pentamers. (b) Structure of the major capsid protein P3 trimer (PDB entry 1CJD) as seen from the outside of the capsid. The three protein chains are colored green, cyan, and magenta. The amino- and carboxy-termini are at the back. (c) Structure of the asymmetric unit of the icosahedron seen from the outside with four trimers of P3 in pink, a monomer of P31 in light blue, and part of a monomer of P16 in green. (d) Structure of the carboxy-terminal end of the trimeric central spike P5 (PDB entry 1YQ8). The three monomers are colored green, red, and light blue. The amino-terminal and carboxy-terminal ends of the red monomer are indicated. (e) Structure of the lateral receptor-binding spike protein P2 (PDB entry 1N7V). The chain is colored as a rainbow from dark blue (amino-terminal) to red (carboxy-terminal). (f) Schematic representation of the viral membrane (orange) tube “injecting” the DNA (black) into the host cell. The viral capsid is shown with P3 in pink, P31 in light blue, P5 in green, and P2 in purple. The host cell inner and outer membranes are in brown

The regular fivefold vertex consists of the membrane anchor protein P16, the penton base protein P31, the receptor recognition protein P2, and the spike protein P5 (Hong et al. 2014; Fig. 12c). The penton base protein P31 has a single jelly roll domain that packs with other subunits to form the base of the vertex complex. P31 is linked to P3, the carboxy-terminal end of P30, and to P16. P2 and P5 constitute two separate spikes. The spike protein P5 is a trimer with a carboxy-terminal TNF-like domain and a stalk with, in part, a triple β-spiral fold (Merckel et al. 2005). The amino-terminal domain of the homotrimeric P5 protein is embedded in the

Structure and Function of Bacteriophages

47

pentameric penton base of the vertex, P31. P2 is an elongated monomer attached to the vertex complex at an angle. The P2 protein has a seahorse shape with an extended β-sheet tail and a β-propeller head domain (Xu et al. 2003). The head domain binds to the host receptor complex (Huiskonen et al. 2007). P3, P31, and P5 are structurally related to the adenovirus hexon, penton, and fiber proteins, respectively, but P2 is unique to the Tectiviridae (Merckel et al. 2005). The unique vertex contains the transmembrane proteins P20 and P22, the ATPase P9, and the packaging efficiency factor P6. A hexamer of the transmembrane heterodimer P20/P22 forms the central genome delivery channel (Hong et al. 2014). Part of the packaging efficiency factor P6 is anchored to the center of the transmembrane channel. The remaining region of P6 remains exterior to the membrane, associating with P9 and forming a twelvefold symmetry portal complex surrounded by ten P3 trimers. Beneath the capsid lies a membrane which, in the mature virion, follows the shape of the capsid due to the pressure of the packaged genome, the presence of viral transmembrane proteins, and the interactions with P3 (Hong et al. 2014). Half of the mass of the internal membrane has been attributed to membrane-associated proteins, either integral membrane proteins (P7, P11, P14, P16, P18, P20, P22, P32, and P34) or peripheral membrane proteins (P15). The lipid composition of the viral membrane includes predominantly phosphatidylglycerol (43%), phosphatidylethanolamine (53%), and cardiolipin (4%). The distribution of these lipids within the membrane is asymmetric. The inner leaflet contains more phosphatidylethanolamine, whose zwitterionic nature might stabilize the negative charge of the genome. The outer leaflet is enriched in phosphatidylglycerol and cardiolipin, whose negative charge might interact with the positively charged base of P3 (Cockburn et al. 2004). Inside the PRD1 virion, the double-stranded DNA is covalently bound to the terminal protein P8 at both ends through a 50 -linkage. As mentioned before, P8 acts as a primer for replication, but it is also a recognition signal for packaging. The PRD1 genome is presumably tightly wound, which results in a highly pressurized capsid interior. The resulting pressure likely facilitates the formation of the membrane tube responsible for genome transfer (Santos-Pérez et al. 2017; Fig. 12f).

The Corticoviridae Family The Corticoviridae are icosahedral viruses containing an internal lipid membrane like the Tectiviridae (Fig. 13). Only one species has so far been included in this family: Pseudoalteromonas phage PM2. PM2 particles are formed by an icosahedral capsid with an approximate diameter of 60 nm and a triangulation number T = 21d, containing spikes at the vertices. This capsid surrounds an inner lipid bilayer (the lipid core), in which eight different proteins are embedded (Kivelä et al. 2008). The lipid-protein complex is known as the lipid core. In turn, the lipid core encloses a 10 kb, circular, supercoiled molecule of double-stranded DNA (Espejo et al. 1969). The DNA encodes 21 potential open reading frames (Männistö et al. 1999).

48

M. Sanz-Gaitero et al.

Fig. 13 Simplified view of a corticovirus. The major capsid proteins P1 and P2, as well as the lipid core, are indicated. The membrane proteins P3 to P10 are shown as colored shapes, but are not distinguished individually

Assembly The corticovirus phage PM2 recognizes its Gram-negative Pseudoalteromonas hosts using its P1 spike protein (Kivelä et al. 2002). Binding of P1 to the host triggers the uncoating of the virion, allowing the lipid core to interact and fuse with the host outer membrane (Huiskonen et al. 2004). Then, protein P7 degrades the periplasmic peptidoglycan layer (Kivelä et al. 2004), allowing the highly supercoiled DNA to reach, and then pass through, the cytoplasmic membrane. It is thought that DNA replication and assembly of the virus occur near the point of infection, at a site anchored to the inside of the host cytoplasmic membrane (Brewer 1978). The DNA replicates using a rolling circle mechanism (Canelo et al. 1985). Transcription of PM2 genes is carried out by the host RNA polymerase, using the highly supercoiled PM2 DNA as a template (Zimmer and Millette 1975). One model for assembly proposes that two dimers of the transmembrane protein P3 interact with a monomer of the P6 protein, forming the protein scaffold building block (Fig. 14). Triggered by interaction with supercoiled DNA, and probably involving the P4 protein, three building blocks associate to form a subassembly corresponding to an icosahedral facet. The P6 protein determines the angle between adjacent facets and probably drives the membrane curvature. Recruitment of further protein subassemblies leads to the formation of DNA-containing vesicles covered in P3 and P6 proteins, to which the major capsid protein P2 would bind to form the virion. Specifically, in every virus facet, P3 dimers form a planar triangle of helices at the icosahedral threefold axis, to which trimers of the P2 proteins are attached. For this interaction, calcium ions are required. Finally, the pentameric P1 protein is incorporated (Abrescia et al. 2008). P3 and P6 are maintained in the mature particles (Kivelä et al. 1999), so it is thought that they are not only used for phage assembly but also to stabilize the mature virion. The release of progeny is mediated by two proteins, P17 and P18. They are synthesized at about half an hour post-infection. P17 is a holin that gets inserted in the cell cytoplasmic membrane, forming pores through which the cellular lytic factor can reach the periplasmic space and digest the peptidoglycan layer. Through these

Structure and Function of Bacteriophages

49

Fig. 14 Corticovirus assembly. (a) The protein scaffold building block is formed by two dimers of the P3 protein (dark blue) interacting with a monomer of the P6 protein (green), in the membrane. (b) Three building blocks associate to form a subassembly corresponding to an icosahedral facet. P6 proteins from two different subassemblies bind to each other, probably influenced by its interaction with supercoiled DNA. (d) The P6 protein interaction drives the membrane curvature and determines the angle between adjacent facets. DNA is shown in light blue. (e) Further protein subassemblies associate to form DNA-filled vesicles covered in P3 and P6 proteins, to which the P2 major capsid protein (red) and the P1 vertex protein (yellow) bind to form the virion

pores, the P18 protein also penetrates, reaching the outer membrane and disrupting it (Krupovic et al. 2007).

Structure The mature PM2 virion is formed by an icosahedral capsid. The facet-to-facet and vertex-to-vertex capsid dimensions are 57 and 64 nm, respectively. The surface of the capsid is formed by 600 copies of the P2 protein organized in crown-shaped trimers (Fig. 15). Each P2 monomer is composed of two jelly roll domains, so each trimer occupies the quasi-sixfold position of a pseudo T = 21 icosahedral lattice. Each vertex of the icosahedral capsid is occupied by protruding spikes formed by a pentamer of protein P1. Every P1 monomer has three domains, a protruding globular distal domain through which the phage interacts with its host, a central domain, and a proximal jelly roll domain which interdigitates with the surrounding P2 trimers to form the base of the vertex. One icosahedral asymmetric unit contains one P1 monomer and ten P2 monomers (Huiskonen et al. 2004). Underneath the protein

50

M. Sanz-Gaitero et al.

Fig. 15 Corticovirus structure. (a) Overall structure of the virion seen from the outside. The P1 protein pentamers are shown in light blue and the P2 trimers in white, light pink, pink, and purple. (b) Structure as seen from the inside of the capsid. Here, all the P2 trimers are shown in white, while the ordered parts of P3 are shown in yellow and the ordered parts of P6 in red. (c) Structure of the icosahedral asymmetric unit (PDB entry 2W0C) seen from the outside of the capsid, with the same coloring as in the previous panels. (d) Structure of the icosahedral asymmetric unit seen from the side

capsid, there is a lipid bilayer which is associated with eight different proteins (P3–P10). The lipids and the proteins form a particle filled with the dsDNA genome, called the lipid core (Kivelä et al. 2002). There are 240 copies of P3 and 60 copies of P6 in the lipid core (Abrescia et al. 2008). The P4, P5, P7, P8, and P10 proteins could not be localized in the icosahedrally averaged electron density map, so their stoichiometry has not been determined (Kivelä et al. 2008). There are four copies of P3 and one of P6 per icosahedral asymmetric unit. Both P3 and P6 comprise an ectodomain and a transmembrane helix. Protein P3 is arranged in 120 asymmetric dimers. Groups of P3 subunits are linked due to the interaction of the amino-terminal end of a P3 dimer with the protein P6, which inserts into the lipid layer along the edges of the virus facet. In each icosahedral facet, P3 dimers form planar triangles of helices at threefold axes to which P2 trimers attach, connecting the lipid core to the outer capsid (Abrescia et al. 2008). Inside the lipid core, it seems that the strongly supercoiled DNA is organized by interactions with the membrane and membrane

Structure and Function of Bacteriophages

51

proteins. The P6 transmembrane domain together with other components such as the P4 protein could mediate this interaction.

Overview of Order Caudovirales Of all identified viruses, the Caudovirales are the most numerous. They consist of an icosahedral or prolate head with a tail connected to one of the 12 vertices (Fig. 16). The Caudovirales contain bacteriophages of different sizes, both in regard of their physical size and genome length. Correspondingly, the triangulation numbers of their head domains vary. The double-stranded DNA in the capsid is densely packed into a condensate at around 0.5 g/ml (Black and Thomas 2012). The DNA does not follow the icosahedral symmetry of the outside capsid and is organized in shells with Fig. 16 Overall structure and general infection mechanism of the three kinds of Caudovirales members. Schematic structures are shown of a podovirus before (a) and after (b) genome transfer, a siphovirus before (c) and after (d) genome transfer, and a myovirus before (e) and after (f) genome transfer. Capsid proteins are in black, phage genomes in blue, and phage or host proteins that allow genome transfer are in red. The bacterial membrane is shown in brown

52

M. Sanz-Gaitero et al.

an approximate 2.5 nm spacing. The DNA may be naked, have dispersed unstructured proteins embedded within the DNA, have a small number of localized proteins, or have a significant protein core that functions as a DNA translocation device or is itself translocated into the bacterium upon infection. The tail functions as a device to recognize a suitable host cell and a conduit for efficient transfer of their double-stranded genome into it. In many cases, fibers or spikes are part of the tail complex and function to recognize receptors on the host cell (Garcia-Doval and van Raaij 2013). This receptor binding is generally reversible and is followed by an irreversible interaction with a secondary receptor. After receptor binding, the phage DNA is ejected into the cell cytoplasm and starts directing the generation of progeny phage particles. Members of the Podoviridae have an extensile tail: proteins from inside the capsid are extruded and form a tube connecting to the cytoplasm of the bacterium (Fig. 16). The Podoviridae account for about 15% of all identified Caudovirales (Ackermann and Prangishvili 2012). Members of the Siphoviridae have a noncontractile tail; proteins at the end of this tail probably contact specific complexes on the bacterial cell wall where it is close to the cytoplasm, so they can eject their DNA directly into it. The Siphoviridae family comprises more than half of the viruses in the Caudovirales order. Finally, members of the Myoviridae have a contractile outer tail sheath, driving the inner tail tube through the bacterial cell wall and making a direct connection with the cytoplasm. The Myoviridae account for about a quarter of the known members of the order Caudovirales. Hereafter, the assembly processes and structures will be described in more detail and illustrated for some well-known bacteriophages of the Caudovirales order.

Caudovirales Head Assembly Capsid formation in the Caudovirales starts with the portal protein, of which 12 copies form a dodecameric ring. Scaffold proteins assemble onto the portal protein (also called connector protein), and the capsid proteins assemble around this scaffold (Casjens and King 1975) to form the immature phage head (Fig. 17). When the head is complete, terminase subunits bind to the portal ring, forming a DNA packaging motor. This motor translocates the genome into the capsid. During this packaging, the capsid expands, thinning out its wall, and allowing the entry of more double-stranded DNA. During this maturation, scaffolding proteins are digested and leave the capsid in a process associated to the DNA packaging (Suhanovsky and Teschke 2015). When the head is full or a terminase signal is encountered (depending on the phage species), the DNA is cleaved, the portal shuts, the terminase subunits dissociate, and a connector complex binds to the portal ring. For the E. coli siphovirus HK97, atomic models of maturation intermediates and of the mature empty capsid have been determined (Fig. 18) (Veesler et al. 2012a). Here, the capsid proteins covalently cross-link to each other during the final expansion step, providing extra stability (Ross et al. 2005). In Salmonella phage ε15 and coliphage K1-5, there are minor capsid proteins, necessary to keep the capsid stable.

Structure and Function of Bacteriophages

53

Fig. 17 General mechanism of Caudovirales head assembly. The portal protein ring (also called connector protein, in white), scaffolding proteins (green), and capsid proteins (black) are shown (a). The portal protein ring serves as the base, onto which the scaffolding proteins (green, b) and then the capsid proteins (black, c) assemble. (d) Proteases (shown as scissors) cleave the scaffolding proteins, and the terminase complex (dark blue) translocates the phage genomic double-stranded DNA (light blue) into the expanded capsid (e). When the capsid is full, head completion proteins (orange) take the place of the terminase complex, and decoration proteins (purple) may bind to the capsid

Fig. 18 Assembly intermediates of the bacteriophage HK97 head. (a) View along the icosahedral threefold axes of the different intermediates. The hexamers of the capsid protein are colored differently for each intermediate, while the pentamers are colored red. (b) Cross section of the intermediates shown above. Note the increase in internal volume and the thinning of the capsid wall during maturation. (c) Ribbon diagrams of the icosahedral asymmetric unit, consisting in each case of one hexamer (colored differently for each intermediate) and one subunit of the pentamer (red). Prohead I, prohead II, intermediate II, intermediate IV, head I, and head II are PDB entries 3QPR, 3E8K, 3DDX, 2FRP, 2FS3, and 2FT1, respectively

They are also called staple proteins (i.e., as metaphorically equivalent to the stapling of sheets of paper together) or instead cementing proteins and play a key role in the stability of the phage capsid against extreme pH or temperature. In other Caudovirales capsids, there are no cementing proteins or cross-linking, but stable

54

M. Sanz-Gaitero et al.

non-covalent interlocking interactions between the capsid proteins lead to a similar result (Morais et al. 2005). To describe head assembly in more detail, we will discuss the example of bacteriophage T4 and note some interesting differences with other phages. To start assembly of the T4 capsid, a circular dodecamer of portal proteins assembles on the inner side of the host cytoplasmic membrane with the help of gp40. Then, prohead core proteins, gp21, gp22, gp67, and gp68; the initiation proteins, IPI, IPII, and IPIII; and gpalt form what is to be an internal scaffold, on the portal complex. The capsid proteins gp23 and gp24 assemble around this core to establish the phage prohead. Gp23 forms the hexameric tiles of the facets, while gp24 forms the pentameric vertices (in many other phages, the same protein is used to assemble the hexameric and the pentameric tiles). The prohead protease gp21 cleaves amino-terminal residues from gp23, gp24, gp67, and gp68 and digests gp21, gp22, the internal proteins, and gpalt into small fragments. These small fragments are then totally digested and leave the capsid (Fokine and Rossmann 2014). Cleavage of the N-terminus of the capsid proteins triggers a rearrangement, flattening the capsid walls. This increases the phage head volume by about 50%. The immature capsid leaves the host membrane, and a complex of terminase proteins and DNA can now bind to the portal protein and form the genome-packaging machine. The small terminase subunit triggers the activity of large terminase protein, which is the ATP-consuming motor that pushes the DNA into the capsid (Yap and Rossmann 2014). The Bacillus subtilis phage ϕ29 lacks the small terminase subunit and instead has a 174-nucleotide pRNA that helps TerL to package the genome (Rao and Feiss 2015). Packaging stops when the portal complex detects a certain pressure. This is the head-full packaging mechanism that phages like T4 or Mu use. For other phages, such as P2, the terminase complex detects a conserved sequence at the beginning of a full genome copy. Packaging is judged complete when the second conserved sequence is detected at the end of the genome. Whatever the mechanism, the large terminase stops it by cutting the DNA (Nemecek et al. 2007). Neck proteins may now bind to the portal vertex. In the case of bacteriophage T4, the neck proteins gp13 and gp14 bind to the portal complex after DNA packaging. They serve as adaptors to bind the portal dodecamer to the gp15 hexamer on the top of the tail (see below). Fibritin proteins (gpwac) are placed around the neck. They make up both the collar and the whiskers. Now, the full head can bind to the pre-assembled tail (Yap and Rossmann 2014).

Caudovirales Head Structure All members of the Caudovirales order have a capsid, or head, which is full of double-stranded DNA. These heads are made up of the same basic building blocks, although some have additional stabilization or decoration proteins (Fig. 19). They vary in size, and the size is correlated with the genome length (Lavigne et al. 2009). A very large known member is bacteriophage G of Bacillus megaterium, with a genome size of nearly 500 kb, several hundred genes, a head domain of 160 nm in

Structure and Function of Bacteriophages

55

Fig. 19 Representation of several different Caudovirales capsids. The name of the virus is shown above each capsid. For coliphage T4 (top view; EMDB entry EMD-6323), Pseudomonas phage ϕKZ (EMDB entry EMD-1392), Staphylococcus phage ϕ812 (EMDB entry EMD-8304), and coliphage P2 (EMDB entry EMD-5406), a surface-rendered capsid is shown, where the distance from the icosahedral center is color-coded, red for closer to the center and blue for more distal, with white for in-between distances. The icosahedral facet nearest to the reader is indicated with a black triangle and some of the capsid protein hexamers with yellow hexagons. The approximate locations of the fivefold, threefold, and twofold symmetry axes are indicated with white numbers. EMBD is the Electron Microscopy Data Bank (https://www.ebi.ac.uk/pdbe/emdb/)

diameter, and a tail over 450 nm long (Ageno et al. 1973). On the other hand, at just over 14 kb and with a capsid diameter of 43 nm, the Rhodococcus phage RRH1 is an exceptionally small phage, with only 20 genes (Petrovski et al. 2012). Many phage heads are icosahedral, but other phage capsids are prolate, i.e., elongated, like those of Bacillus phage ϕ29 and coliphage T4. The ϕ29 head is 54 nm long by 45 nm wide, while the head of T4 has a length of 115 nm and a width of 85 nm (Fokine et al. 2004). All the Caudovirales main capsid proteins have the HK97 fold (Fig. 20). The HK97 fold allows capsid proteins to be flexible, so the capsid can change conformation during maturation. However, it is also sufficiently strong to keep a large amount of DNA in the capsid at a high internal pressure (Suhanovsky and Teschke 2015). The main capsid proteins make up the hexamers on the facets or the pentamers on the vertices, either with the same protein making up both the hexamers and pentamers (e.g., gpN in phage P2) or different but structurally related proteins making up the hexamers and pentamers (e.g., gp23 and gp24, respectively, in phage T4). Sometimes they have added domains and loops or a partially altered topology.

56

M. Sanz-Gaitero et al.

Fig. 20 Caudovirales major capsid proteins: the HK97 fold. Capsid proteins of phages HK97 (PDB entry 2FT1), P22 (PDB entry 5UU5), ϕ29 (PDB entry 1YXN), ε15 (PDB entry 3 J40), and T4 major capsid proteins gp23 and gp24 (PDB entry 5VF3). The N-arm is shown in red, the E-loop in green, the P-loop in yellow, the spine helix in magenta, and the five-stranded β-sheet in orange. Phage ε15 and T4 capsids have cementing proteins, which are shown in dark blue. The carboxyterminal anchor of the T4 Hoc decoration protein is shown in black. In phages P22, ϕ29, and T4, there is an extra domain (cyan) in the capsid protein that may have the same function as the cementing proteins have in other phages

The size and shape (icosahedral or prolate) of the capsid are probably largely determined by the scaffolding structure around which it assembles. In larger capsids, more hexameric tiles of the main capsid protein are incorporated, leading to higher triangulation numbers, up to T = 52 for very large bacteriophages (Hua et al. 2017). Proteins with the HK97 fold have several common features (Fig. 20). A fivestranded β-sheet and two helices form the A-domain, which is at the center of hexamers and pentamers. The E-loop, the P-loop, and the spine helix form the P-domain, which makes up the outer part of hexamers and pentamers. An aminoterminal arm (N-arm) with α-helical and β-sheet content connects with a neighboring capsid protein hexamer or pentamer. In addition, some Caudovirales have cementing proteins (Fig. 20; e.g., Soc in phage T4). In addition to the cementing proteins, or instead of them, the major capsid proteins may have an extra domain in the capsid protein that may also have a stabilization function. All Caudovirales capsids have one special vertex. At one of the 12 vertices, a dodecameric head-tail connector or portal protein is present instead of a pentamer of capsid proteins (Prevelige and Cortines 2018). The dodecameric portal initiates capsid assembly. The genomic DNA is packaged into the assembled capsid through this portal. The portal is essential for tail assembly, and DNA again passes through it upon adsorption. The basic structure of all portal proteins is the same, despite of, in many cases, little sequence homology (Fig. 21; Parent et al. 2018).

Structure and Function of Bacteriophages

57

Fig. 21 Structure of Caudovirales portal proteins. Side (top) and bottom views (bottom, viewed from outside the phage) of portal complexes of phage P22 (PDB entry 3LJ5), SPP1 (PDB entry 2JES), and T4 (PDB entry 3JA7). One monomer is highlighted in yellow. The rough locations of the barrel, crown, wing, stem, and stalk domains are indicated on the left. Right: Single monomer of the T4 portal complex with the amino- and carboxy-termini indicated. It is rotated 90 with respect to the yellow monomer in the side view of the dodecamer next to it

Portal proteins contain mainly α-helices and coils. They have up to five domains: the barrel, the crown, the wing, the stem, and the stalk (Fig. 21; Prevelige and Cortines 2018). Some but not all portals have an α-helical barrel on the inside of the virus (Fig. 21; Olia et al. 2011). The barrel changes conformation upon genome packaging, going from unstructured to helical, acting as a pressure sensor. The portals of phages ϕ29 and T7 do not have barrels. These phages do not perform head-full assembly and so presumably do not need these barrels. Facing the inside of the viral capsid is the crown. The crown is flexibly linked to the wing. The crown and wing display the greatest variability between portals and allow conformation changes during packaging, head-full sensing, and opening for DNA exit. The wing contacts nearby capsid proteins directly and is important for transmitting conformational changes between DNA translocation and capsid structure. The wing is connected to the stem region by a flexible loop. The loops of the 12 portal monomers extend into the central portal channel and prevent leakage of the packaging DNA. The stem is formed by 12 helices, tilted by 30 to 50 with respect to the direction of the channel. The stalk forms the initial channel for DNA packaging. The central channel is about 3 nm in diameter, just enough to allow passage of double-stranded DNA. Upon maturation, conformational changes widen the channel to about 4 nm, presumably to allow smooth DNA delivery. The stalk is on the outer surface of the capsid and interacts with the terminase proteins during DNA translocation and with adaptor proteins to bind the tail in the mature virion.

58

M. Sanz-Gaitero et al.

Fig. 22 Decoration proteins of Caudovirales. (a) Low-resolution cryo-electron microscopy structure of Bacillus phage ϕ29 (gray; EMDB entry EMD-1506). One of the head fibers is indicated with an arrow. (b) Fibrous part of the ϕ29 head fiber trimer colored in red, green, and blue (PDB entry 3QC7). The amino-terminal part forms a trimeric super helix-turn-helix-coiled coil, while the carboxy-terminal part forms a small tip domain to which each monomer contributes an α-helix. The amino- and carboxy-termini of the green chain are indicated. (c) Hoc from coliphage RB49, a close relative to phage T4 (orange; PDB entry 3SHS). The three aminoterminal immunoglobulin-like domains are indicated (D1-3); the fourth domain, which in the intact phage would bind to the capsid, is not resolved. The amino- and carboxy-termini are indicated

Apart from capsid proteins and cement proteins, Caudovirales capsids sometimes have what are called decoration proteins (Fig. 22). Decoration proteins are not necessary for capsid assembly or phage infectivity, but may help increasing phage binding to their bacterial host. They may also help to attach to surfaces where host bacteria are likely to pass, such as the lung or gut epithelia. Head decoration proteins can be used in phage display applications, for example, for displaying antigens when using phages as vaccination vehicles (Tao et al. 2018). In phage T4 and T4-like phages, the Hoc protein (highly immunogenic outer capsid protein) protrudes from the center of the gp23 hexamers (Fig. 19; Fokine et al. 2011). Hoc is anchored to the capsid with its carboxy-terminal end and consists of four consecutive immunoglobulin-like domains (Fig. 22), exposing the aminoterminal domains to the medium. Bacillus phage ϕ29 has 55 head fibers bound to the capsid (Xiang and Rossmann 2011). The ϕ29 prolate capsid (T = 3/Q = 5) consists of the major capsid protein gp8 and the head fiber, which is a trimer of gp8.5. Gp8 has an additional domain (Fig. 20), which provides attachment sites for

Structure and Function of Bacteriophages

59

the head fibers at quasi-threefold symmetry positions. The head fibers have two domains: a base that attaches to the capsid and a protruding fibrous domain. The fibrous domain has a unique helix-turn-helix supercoil fold capped with a small head domain that contains a short triple coiled coil (Fig. 22). In this case, the aminoterminal end of the protein binds to the phage capsid, and the carboxy-termini are exposed to the medium, like for bacteriophage fibers and tailspikes (see below).

Podoviridae Tail Assembly and Structure Bacteriophages of the Podoviridae family have a relatively short, noncontractile tail which is not flexible (Casjens and Molineux 2012). The exact length and shape of these tails vary, presumably in relation to the phage host. A dodecameric adaptor protein binds to the portal, and one or more tail proteins bind to this adaptor protein. These tail proteins are often hexamers. Many podoviruses have six trimeric receptorbinding tailspikes or tail fibers that attach to the tail. When they bind to the phage receptor, this interaction leads to a conformational change in the tail to initiate DNA transfer into the host (Garcia-Doval and van Raaij 2013). In the Podoviridae, after the head is assembled and filled with DNA, tail proteins are added to the portal sequentially, until the tail, and thus the phage, is complete (Fig. 23). In the case of coliphage T7, a dodecamer of gp11 binds first and functions as a gatekeeper complex, to retain the DNA in the capsid (Cuervo et al. 2013). Six copies of gp12 bind to gp11, forming a nozzle. Finally, six trimeric tail fibers (made up of the gp17 protein) are bound to the gatekeeper and retracted toward the capsid (Hu et al. 2013). During infection, the C-termini of the fibers dislodge and bind to the host lipopolysaccharide (Fig. 16; González-García et al. 2015). The T7 tail is about 30 nm long and 17 nm wide, including the portal protein. An isolated tail-portal complex has been studied (Cuervo et al. 2013). It consists of a dodecamer of the portal protein gp8, a dodecamer of the gatekeeper protein gp11, a hexamer of the nozzle protein gp12, and six trimers of the fiber protein gp17. The carboxy-terminal half of the T7 tail fiber forms a threefold symmetric pyramid

Fig. 23 Schematic overview of the construction of the bacteriophage T7 tail as an example of podovirus tail assembly. The DNA-filled head, previously assembled around the dodecameric portal complex (gp8, in white), forms the starting point. Twelve copies of gp11 (purple) join first, forming a circular gatekeeper complex. Subsequently, six gp12 proteins (green) bind and make up the nozzle. Finally, six pre-assembled trimers of gp17 fibers (brown) are incorporated

60

M. Sanz-Gaitero et al.

domain and a globular tip domain (Fig. 24c; Garcia-Doval and van Raaij 2012). Before infection, phage T7 particles have the fiber pointing upward, interacting with the icosahedral capsid (Hu et al. 2013). In the isolated tail, the amino-terminal part of the tail fiber can be clearly seen pointing upward (Fig. 24a; González-García et al. 2015). In this conformation, the nozzle is closed at the tip of the tail, with the six copies of the gp12 protein pointing inward. When the carboxy-terminal distal parts of the fiber contact the host membrane, the proximal amino-terminal ends of the fiber initiate a conformational change, leading to a straightening of the gp12 monomer and an opening of the nozzle (Fig. 24b). During infection, the core complex, consisting of multiple copies of the gp14, gp15, and gp16 proteins, moves from inside the capsid, where they sit just above the portal ring, to form a tube spanning the periplasmic space of the bacterial host (Hu et al. 2013). Presumably, the core complex proteins unfold to a large extent, pass through the gp8-gp11-gp12 complex, and refold in the periplasmic space. The core protein gp16 has been shown to have peptidoglycan digestion activity (Moak and Molineux 2004), so this probably helps the process. The phage genomic DNA then passes through the tail, through the periplasmic tube, and directly into the host cytoplasm. Structural studies on the Prochlorococcus phage P-SPP7 (Liu et al. 2010) suggest it has the same mechanism of infection as coliphage T7. The tail of the Salmonella phage P22 (Fig. 24d) has an organization similar to the tail of coliphage T7 (Bhardwaj et al. 2014). Here, the dodecameric portal protein (gp1) binds to the dodecameric α-helical adaptor protein gp4 and the hexameric nozzle protein gp10. However, the nozzle is not closed, but its channel is occupied by aminoterminal part of the trimeric needle protein gp26. The rest of gp26 protrudes by about 14 nm (Tang et al. 2011). Gp26 is composed of a long α-helical coiled coil, interspersed with β-structure (Fig. 24f; Olia et al. 2007). In phage SF6, gp26 is capped with a globular tip domain (Bhardwaj et al. 2011), but in P22, this tip is absent. As primary receptor-binding proteins and instead of thin, L-shaped, fibers, phage P22 has six stubby trimeric tailspikes, each made up of trimers of gp9 (Fig. 24f). Each tailspike has a small trimeric amino-terminal β-structured domain with which they attach to the phage neck (Seul et al. 2014). The distal carboxy-terminal domain is intertwined and has been shown to be important for the correct trimeric assembly and folding of the tailspikes (Takata et al. 2012). In the central part of the tailspikes, each protomer contains parallel β-helix domains. The β-helix domains function in adhesion to the O-antigen repeating units of the host lipopolysaccharide and cleavage of the O-antigen repeating units. Multiple rounds of cleavage and adhesion allow the phage to approach the host membrane. Presumably, the tip of gp26 then gets pushed against the membrane and leads to a conformational change in the tail, setting off the events necessary for DNA-transfer into the host. Tailspikes with a β-helical domain that cleaves the host O-antigen are a common occurrence in the adsorption devices of Caudovirales, not only in the Podoviridae but also in Myoviridae (Walter et al. 2008), so this method of approaching the membrane is apparently successful and necessary for infecting certain host bacteria. The exact

Structure and Function of Bacteriophages

61

Fig. 24 Structure of tails of podoviruses that infect Gram-negative bacteria. (a) Structure of the coliphage T7 tail before DNA ejection (EMDB entry EMD-1163). Approximate volumes of the portal protein (gp8; green), adaptor protein (gp11; blue), nozzle protein (gp12; orange), and fiber (gp17; beige) are colored. (b) Structure of the coliphage T7 tail after DNA ejection (EMDB entry EMD-2717). The approximate positions of the carboxy-terminal domains of the fiber are indicated. Black lines indicate the positions of an individual gp12 molecule in each of the two structures, which straightens out to allow DNA ejection. (c) Structure of the carboxy-terminal domain of the phage T7 fiber (gp17; PDB entry 4A0U). The three chains of the protein trimer are colored differently, and the amino- and carboxy-termini of the green chain are indicated. (d) Structure of the Salmonella phage P22 tail (EMDB entry EMD-5348). The portal protein (gp1) is shown in beige, the adaptor protein (gp4) in light blue, the nozzle protein (gp10) in purple, the tailspikes (gp9) in magenta, and the tail needle (gp26) in brown. (e) Structure of the dodecameric P22 gp4 adaptor ring (PDB entry 4V4K) in side view (top) and bottom view (bottom). One of the monomers is highlighted in orange. The carboxy-terminal end interacts with the bottom of the dodecameric portal (gp1). (f) Structure of the phage P22 needle (gp26; PDB entry 2POH). The three chains of the protein trimer are colored differently, and the amino- and carboxy-termini of the red chain are indicated. (g) Phage P22 tailspike structure (gp9; PDB entry 2XC1) with the three chains of the protein trimer colored differently. A fragment of O-antigen receptor (PDB entry 1TYX) is superposed in the binding site facing the reader. (h) Phage K1F endosialidase tailspike structure (PDB entry 1V0F) with the three chains of the protein trimer colored differently. Fragments of oligomeric α-2,8-sialic acid are shown

62

M. Sanz-Gaitero et al.

receptor-binding sites and enzymatic mechanisms of different tailspikes vary, but their structural framework is very similar. Some podoviruses have a more complex tail with multiple tailspikes or fibers. An example is coliphage K1-5 (Leiman et al. 2007), in which the single receptorbinding protein is replaced with a protein that binds to two different tailspikes. One of these tailspikes is an endosialidase (Fig. 24h; Schulz et al. 2010b), which trimerises using an intramolecular chaperone (Schulz et al. 2010a), just like the T5 side tail fibers described in the next section. The endosialidase, which has multiple binding sites for sialic acid moieties (Fig. 24h) and an active site in each monomer, allows the phage to tunnel its way through the poly-sialic acid capsule of host bacteria. The other tailspike is more like the P22 tailspike. Phage ϕ29 infects the Gram-positive bacterium B. subtilis, so its DNA does not need to traverse an outer membrane. Phage ϕ29 has a longer tail than many other podoviruses (about 50 nm). Here, the portal is a dodecamer of gp10, occupying the special vertex and forming the upper part of the collar. A tubular protein, dodecamer of gp11, binds to the portal forming the lower collar and the tail tube. A hexamer of gp9 and two copies of gp13 bind to the tail tube, forming the distal knob (Tang et al. 2008). Twelve appendages are attached to the phage collar (Fig. 25). Each appendage folds using an intramolecular chaperone (Xiang et al. 2009), like the endosialidase of coliphage K1-F and the fibers of the siphovirus T5 (see below). The appendages are homo-trimers of gp12 and can be in the up or down position (Farley et al. 2017). They are responsible for the digestion of the teichoic acid layer of the bacterium. At the end of the tail, a knob complex is located, consisting of a hexamer of gp9 (Xu et al. 2016) and probably two molecules of gp13 (Xiang et al. 2008). Gp13 is a peptidoglycan-degrading enzyme that helps the tail knob to reach the bacterial cytoplasmic membrane. Gp13 has an amino-terminal domain with a lysozyme fold and a carboxy-terminal domain with an endopeptidase fold (Fig. 25e). The domains are linked by an oligo-glycine linker which allows flexibility and was not resolved in the crystal structure. Once the membrane has been reached, gp9 is able to form a pore through it to allow DNA transfer directly into the host cytoplasm (Xu et al. 2016). The arrangement of a longer tail and 12 appendages is not limited to podoviruses infecting Gram-positive bacteria. Coliphage N4 also has 12 appendages and an overall similar structure to phage ϕ29 (Choi et al. 2008).

Siphoviridae Tail Assembly In the Siphoviridae (like Myoviridae), tail assembly takes place in parallel to head assembly, starting with the baseplate, i.e., the distal end of the tail (Davidson et al. 2012). Coliphage λ has been the model for siphovirus assembly (Katsura 1990). To start assembly, a trimer of gpJ makes up the distal tip structure of the tail (Xu et al.

Structure and Function of Bacteriophages

63

Fig. 25 Structure of bacteriophage ϕ29 tail. (a) Cryo-electron microscopy map of the ϕ29 bacteriophage (gray; EMDB entry EMD-1420) with the atomic structures of the connector protein gp10 (green), two appendages, each consisting of a trimer of gp12 (violet), and the tail knob, a hexamer of gp9 (cyan). The atomic structure of the appendages only contains the C-terminal region of the trimer. Hence, the N-terminal region is still in gray, as part of the cryo-electron microscopy map. Missing from the tip of the tail is gp13, an enzyme that degrades the peptidoglycan layer and facilitates access of the tail tip to the cytoplasmic membrane. (b) Side (top) and bottom (bottom) views of the knob complex. One monomer is highlighted in yellow. (c) Structure of the carboxyterminal domain of the ϕ29 appendage (PDB entry 3GQ7). The three monomers are colored in green, blue, and red. The amino- and carboxy-termini of the green chain are indicated. (d) Structure of the gp9 knob monomer (PDB entry 5FB5) in the same position as the yellow monomers in B. The protein is colored in a rainbow representation, amino-terminus to carboxy-terminus in blue to red. (e) Structures of the amino-terminal lysozyme domain (PDB entry 3CT0) and the carboxy-terminal endopeptidase domain (PDB entry 3CSQ) of gp13. An N-acetyl glucosamine oligomer bound to the amino-terminal domain is shown in stick representation

64

M. Sanz-Gaitero et al.

Fig. 26 Schematic overview of the assembly of bacteriophage λ as an example of siphovirus assembly. The assembly of the tail is shown in greater detail. The tail tip complex (shown in green, at the bottom of the figure), consisting of a trimer of gpJ and the gpI, gpL, and gpK proteins, joins up with a pre-assembled complex of the tape measure protein gpH (in blue, copy number unknown), covered with the tail assembly chaperone proteins gpG and gpGT (both shown in orange). GpM joins at an unspecified position. Incorporation of the tail tube protein gpV (in red) displaces gpG and gpGT. When the tail is complete, it is capped by gpU (purple) and gpZ also incorporates. To complete phage assembly, the DNA-filled head (colored as in Fig. 17) and the fibers (brown) join

2014). Subsequently, copies of gpI, gpL, and gpK are added (Fig. 26). At the same time, a sixfold helical coiled coil of the tape measure protein gpH gets covered with the tail chaperone proteins gpG and gpGT (gpGT is a read-through product of gpG gene due to a frameshift regulated to give a gpG-to-gpGT ratio of 30:1). GpG is thought to bind to gpH, increasing gpH’s solubility, while gpGT may guide the incorporation of gpV building blocks. The C-terminal end of gpH binds to the top of the tail tip complex, and the tail tube initiator protein gpM binds at an unspecified position. Multiple copies of the tail tube protein gpV assemble onto the tip and around gpH, displacing gpG and gpGT. The tail termination protein gpU binds to the top of the tail, while gpZ also joins at an unknown location. The complete tail can now bind to the assembled phage head. Not all siphoviruses have side tail fibers, but in the case of phage λ, side tail fibers – each consisting of trimers of the side tail fiber protein capped by trimers of the tail fiber assembly protein – also bind to the tail tip. The tail fiber assembly protein is expressed in large amounts and probably also functions as a chaperone for correct assembly of the side tail fibers. The exact number of side tail fibers that bind to each λ virion is uncertain. In the case of coliphage T5, another siphovirus, the three side tail fibers, which are each trimers of the pb1 protein, also assemble with the help of a chaperone. However, this chaperone is intramolecular and a carboxy-terminal extension of the pb1 protein. The intramolecular chaperone helps the fibers to fold, and once folded, a proteolytic site forms and the chaperone cleaves itself off (GarciaDoval et al. 2015).

Structure and Function of Bacteriophages

65

Siphoviruses infecting Gram-positive bacteria probably share many of the assembly features described above. This is confirmed by a model for the assembly of the lactococcal phage TP901-1 (Mahony et al. 2016). Here, the Dit protein is the hub to which the C-terminus of the tape measure protein and the N-terminus of Tal bind. On the top of this platform, tail construction takes place, while on the lower side, baseplate proteins involved in host recognition bind. Like in phage λ, gpG and gpGT chaperone the incorporation of the multiple tail tube protein subunits. The top of the tail is capped by Ttp. Tap, HTC1, and HTC2 are probably involved in head-totail assembly.

Siphoviridae Tail Structure Phages of the Siphoviridae family have tails from just under 100 nm up to more than 400 nm long, depending on the species. These tails may be flexible and are not contractile. The tails are connected to the head portal via small adaptor proteins (Tavares et al. 2012), which generally form dodecameric rings. These adaptor proteins are of different kinds. Some are α-helical and have the same fold as the P22 gp4 protein (Fig. 24e). Examples are gp6 of the coliphage HK97 and gp15 of Bacillus phage SPP1. A second kind, including gpW of phage λ, has a different fold (Fig. 27b), consisting of two α-helices and a β-hairpin (Maxwell et al. 2001). GpW acts as a stopper that prevents exit of the genome (Perucchetti et al. 1988). Phages λ and SPP1 (Fig. 27a) have a second ring of proteins below the adaptor proteins, made up of dodecamers of gpFII and gp16, respectively. These are small β-structured proteins (Fig. 27c; Maxwell et al. 2002). In phage SPP1, this second ring of gp16 dodecamers acts as the stopper (Lhuillier et al. 2009). A loop from each of the 12 gp16 subunits extends into the tail channel, blocking DNA exit (Chaban et al. 2015). The lower ring of connector proteins attaches to a hexameric ring of the tail terminator protein. The structure of the hexameric ring of the tail terminator protein of phage λ, gpU, has been determined (Fig. 27d; Pell et al. 2009a). Each monomer of the tail terminator protein is made up of a five-stranded β-sheet covered with α-helices on the outside. The main part of the tail is made up of a singular tubular structure, the tail tube, which generally has sixfold symmetry (Davidson et al. 2012). The tube is made up of stacked hexameric rings of the tail tube protein (shown in Fig. 27g for the myovirus T4, which has a homologous structure). The tail can be lengthened or shortened by incorporating more or less hexameric rings; the phage λ tail tube contains 32 stacked rings. The tail tube protein fold can be described as a β-sandwich or a folded-over β-sheet (Fig. 27e; Pell et al. 2009b). Like for the tail termination protein, the part facing inside is a β-sheet, and the inner diameter is around 4 nm. The outer diameter of the tail tube is about 9 nm. This diameter may be increased in siphoviruses if there are decoration domains present. For example, each copy of the tail tube protein of phage λ, gpV, has a carboxy-terminal immunoglobulin domain decorating the outside of the tail tube (Fig. 27e; Pell et al. 2010). The tape measure protein, which is located inside the tail tube, probably forms a

66

M. Sanz-Gaitero et al.

Fig. 27 Structural details of siphovirus tail neck and tube. (a) Cryo-electron microscopy map of the Bacillus phage SPP1 neck region (EMDB entry EMD-2993) with models of the portal protein gp6 (green, 2 of 12 total copies), the adaptor protein gp15 (cyan, 2 of 12 total copies), the stopper protein gp16 (magenta, 2 of 12 total copies), the tail terminator protein gp17 (yellow, 1 of 6 total copies), and the tail tube protein gp17.1 (red, 1 of 6 total copies) fitted (PDB entry 5A20). (b) Solution structure of the adaptor protein GpW of coliphage λ (cyan, PDB entry 1HYW). (c) Solution structure of the adaptor protein GpFII of coliphage λ (magenta, PDB entry 1K0H). An asterisk indicates the loop that in the homologous protein gp16 of phage SPP1 points into the inner channel. (d) Phage λ tail terminator protein gpU hexamer (PDB entry 3FZ2) seen from the top (head-binding) side. One of the six monomers is colored orange. (e) Phage λ gpVN tail tube protein domain (PDB entry 2K4Q; top; red) and the immunoglobulin decoration domain gpVC (PDB entry 2 L04; bottom; orange). Amino- and carboxy-termini are indicated. (f) Coliphage T5 tail tube protein gp6, containing a duplicated tail tube domain (PDB entry 5NGJ). The protein is colored in a rainbow spectrum: blue to red from amino- to carboxy-terminus. Tail tube domain 1 (D1) is shown in blue-green, tail tube domain 2 (D2) in green-yellow, and the immunoglobulin-like decoration domain in orange-red. (g) Bacteriophage T4 inner tail tube (PDB entry 5V5F). Three stacked hexamers are shown in gray; one monomer is shown in red

Structure and Function of Bacteriophages

67

hexameric α-helical coiled coil along the length of the tail tube. Bacteriophage T5 has a threefold, rather than sixfold, symmetric tail tube (Arnaud et al. 2017). Its tail tube protein, pb6, is larger than the tail tube proteins of other siphoviruses and contains two β-barrel tail tube domains instead of one, leading to a pseudo-sixfold symmetry (Fig. 27f). The tail tube structure does not change upon DNA ejection (Arnaud et al. 2017), so it is likely that the signal that a suitable host is encountered is transmitted through the tape measure protein. When the tail tip contacts its receptor on the bacterial membrane, it opens and the tape measure protein leaves first. Its release must then lead to a conformational change in the neck and connector region of the tail (Tavares et al. 2012). Some parts of the tape measure protein have homology with peptidase- and peptidoglycan-degrading domains, suggesting that, once ejected, the tape measure protein may refold to facilitate passage of the DNA through the periplasm (Davidson et al. 2012). The tape measure protein may even form a channel across the periplasmic space, like the podovirus core proteins. The end of the siphovirus tail is the part of the structure where most structural variation between phages is observed. Here, a tail tip complex or a baseplate is located (Fig. 28a). These complexes recognize the host bacterium and initiate conformational changes that allow successful infection (Davidson et al. 2012). The tail tip complex is a narrow conical tip, while other phages have a more platelike assembly, i.e., a baseplate. Cell attachment is by tail fibers or spike-shaped receptorbinding proteins, which project from the tail tip or baseplate. The central element of the tail tip is a trimeric hub protein. Each monomer of the trimeric hub protein contains two domains with the same fold as the tail tube protein, a folded-over β-sheet (Fig. 28b), making a pseudo-hexameric ring. The distal tail protein is a hexamer and adapts the hub to the tail. It also contains the folded-over β-sheet and so also leads to a hexameric ring of the same structure (Fig. 28c; Veesler et al. 2010). Coliphages λ and T5 have several trimeric side tail fibers and a central tail fiber (Flayhan et al. 2014). The correct trimerization and folding of the side tail fibers are mediated by chaperone proteins. These chaperone proteins may be intramolecular, like for T5 (Fig. 28e; Garcia-Doval et al. 2015), or they may be separate proteins, like for phage λ. Interestingly, for phage T5, the intramolecular chaperone has also been shown to shield the receptor-binding site. The side tail fibers are responsible for reversible attachment to a common component of the bacterial cell wall, such as the lipopolysaccharide in case of Gram-negative bacteria. This allows for lateral or two-dimensional diffusion until the central tail fiber encounters its receptor, which is usually a protein to which it binds irreversibly (Garcia-Doval and van Raaij 2013). Bacteriophage SPP1, which infects the Gram-positive B. subtilis, lacks side tail fibers, but also binds to a major cell wall component reversibly, in this case to teichoic acids (Baptista et al. 2008). It is not known which protein is responsible for binding to teichoic acids, but its central tail fiber then binds irreversibly to the protein YueB (Vinga et al. 2012).

68

M. Sanz-Gaitero et al.

Fig. 28 Structures of siphovirus tail tips and receptor-binding proteins. (a) Schematic structure of a siphovirus tail tip with the distal tail protein in blue, the trimeric hub protein in yellow, and the trimeric straight tail fiber in red. The side tail fibers are also shown, and the tail tube outline is shown in gray. (b) Trimeric baseplate hub protein of a Listeria prophage (PDB entry 3GS9). One monomer is shown in rainbow color, from blue (amino-terminus) to red (carboxy-terminus); the other two monomers are shown in gray and black. Note the trimeric bottom part and pseudo-hexameric top part. (c) Distal tail protein hexamer of phage SPP1 (PDB entry 2X8K). One of the monomers in the front is shown in blue; the others are gray. The amino-terminus of the protein is indicated. One of the carboxy-terminal galectin domains is indicated with an asterisk. (d) Structure of the distal, carboxyterminal end of the phage T5 side tail fiber (PDB entry 4UW7). The protein chains are colored green, magenta, and cyan. The amino- and carboxy-termini are indicated. (e) Structure of the distal, carboxy-terminal end of the phage T5 side tail fiber with the intramolecular chaperone is still attached (PDB entry 4UW8). Here, the part corresponding to the mature protein is colored black, gray, and white and the intramolecular chaperone red, blue, and yellow. The amino- and carboxytermini are indicated

It has been proposed that siphoviruses with a narrow tail tip bind a protein receptor with very high affinity, while siphoviruses with a more elaborate baseplate bind saccharidic receptors (Veesler et al. 2010). Examples of siphophages with a structurally studied baseplate are the Lactococcus phages TP901-1 and p2 and the Staphylococcus phage ϕ11. The baseplate of phage TP901-1 contains four different proteins (Fig. 29a,b; Veesler et al. 2012b). The center is a hexameric ring of the distal

Structure and Function of Bacteriophages

69

Fig. 29 Structures of siphovirus baseplates and receptor-binding proteins. (a) Lactococcus phage TP901-1 baseplate seen from the side, colored in blue (PDB entry 4 V96). One of the distal tail proteins of the central ring is colored red, the trimeric arm bound to it is colored magenta, and the three receptor-binding protein trimers bound to the arm are in yellow. (b) Lactococcus phage TP901-1 baseplate seen from the bottom colored as in part a. (c) Phage TP901-1 receptor-binding protein trimer (PDB entry 3EJC). The amino- and carboxy-termini of the magenta chain are indicated. (d) Unactivated Lactococcus phage p2 baseplate seen from the side, colored in cyan (PDB entry 2WZP). One of the distal tail proteins of the central ring is colored magenta, one of the central hub proteins is colored red, and one of the receptor-binding protein trimers bound is in yellow. (e) Unactivated Lactococcus phage p2 baseplate seen from the bottom colored as in part d. (f) Calcium-activated Lactococcus phage p2 baseplate seen from the side, colored in green (PDB entry 2X53). One of the distal tail proteins of the central ring is colored magenta, one of the central hub proteins is colored red, and one of the receptor-binding protein trimers bound is in yellow. (g) Calcium-activated Lactococcus phage p2 baseplate seen from the bottom colored as in part f. (h) Lactococcus phage p2 receptor-binding protein trimer with each monomer colored differently (PDB entry 1ZRU). The amino- and carboxy-termini of the green chain are indicated. (i) Receptorbinding protein of Staphylococcus phage ϕ11 (PDB entry 5EFV). The amino- and carboxy-termini of the red chain are indicated, as is the location of the iron ion (Fe)

tail protein, which surrounds a central spike (not shown in the figure). The distal tail protein has the same folded-over β-sheet fold as the tail tube, while the structure of the central spike has not been determined. However, it may well be the equivalent of

70

M. Sanz-Gaitero et al.

the trimeric tail tip hub. From the distal tail protein, six trimeric α-helical arms project sideways. The rest of the arm points downward, and each monomer forms an adaptor domain. To each adaptor domain, three trimeric receptor-binding proteins are attached. The amino-terminal end of the protein forms a short α-helical coiled coil, followed by a short β-helical domain and a carboxy-terminal head domain oriented toward the bacterial host (Fig. 29c; Bebeacua et al. 2010), leading to a total of 54 receptor-binding sites. The receptor-binding proteins point downward toward the host bacterium, ready for adhesion. How receptor binding is related to DNA transfer is less clear, perhaps the strong binding with up to 54 receptor molecules pushes the spike against the cell wall, and this force is sensed by the baseplate, which then opens to let the tape measure protein leave, followed by the viral DNA. The baseplate of phage p2 is composed of three different proteins (Fig. 29d,e; Sciara et al. 2010). The central part of the baseplate is formed by the distal tail protein, which forms a hexameric ring with a central hole. Each monomer in this ring contains the tail tube fold in its N-terminal domain and also has a carboxy-terminal galectin domain. Below this ring is the trimeric hub, with each monomer containing two tail tube domains to adapt to the hexameric ring above. The hub forms a closed dome, blocking passage of the tape measure protein and the phage DNA. From the galectin domain of the distal tail protein, an adapter arm protrudes which interacts with a trimer of the receptor-binding protein. In the unactivated baseplate, the six trimeric receptor-binding proteins point upward, away from the host. Calcium causes a large conformational change in the baseplate, activating it and leading to rotation of receptor-binding domains to point downward (Fig. 29f,g; Sciara et al. 2010). The receptor-binding protein has an amino-terminal β-sandwich shoulder domain, which binds to the adaptor protein. It is followed by a short triple β-helical neck and a carboxy-terminal head domain (Fig. 29h; Spinelli et al. 2006; Tremblay et al. 2006). The structure of the receptor-binding protein was determined bound to glycerol (Fig. 29h), which may be mimicking part of the teichoic acid of the bacterial cell wall. During the conformation change, the monomers of the dome protein separate, allowing passage of the tape measure protein and DNA into the host. The structure of the receptor-binding protein of the Staphylococcus phage ϕ11 revealed a trimer that can be divided, from the amino- to carboxy-terminus, into stem, platform, and tower domains (Fig. 29i; Li et al. 2016; Koç et al. 2016). The stem is formed by several triple α-helical coiled coils. The first two coiled coils are colinear and interrupted by a region where the three protein chains intertwine, and each chain forms a small β-hairpin. In the center of the intertwined region, an iron ion is located, coordinated by six histidine residues, two from each protein chain. After the second coiled coil, a sharp bend leads into the third coiled coil. The stem is followed by a platform of three β-propellers, and the carboxy-terminal part is a tower formed by two β-prism domains. The β-propeller platform and tower are interconnected by another short triple α-helical coiled coil inside the protein. The β-propeller domain is structurally related to carbohydrate degradation proteins and may thus be involved in receptor binding.

Structure and Function of Bacteriophages

71

Myoviridae Tail Assembly As for the Siphoviridae, in the Myoviridae, tail assembly takes place in parallel to head assembly, starting with the baseplate, i.e., the distal end of the tail. Onto the baseplate, the helical inner tail tube and the outer tail sheath are assembled. The length of the tail is controlled by the tape measure protein. When tail assembly is complete, the tail is joined to the full phage head by the connector complex. Long tail fibers, if present, form separately and are joined to the head-tail assembly afterward. Here, we illustrate the assembly of myovirus tails using as examples the relatively simple coliphage Mu and the more complicated phage T4. The general assembly mechanism of these phages is likely to be extensible to myoviruses infecting Grampositive bacteria also. Assembly of the myovirus T4 has been well-studied, and the tail is no exception (Kikuchi and King 1975; Leiman et al. 2010; Arisaka et al. 2016). At the same time, bacteriophage Mu is a relatively simple myovirus and a good general model for

Fig. 30 Myovirus Mu tail assembly. Six wedges, shown in different shades of blue, assemble around a central hub, shown in shades of green, to form a dome-shaped baseplate. Onto the baseplate, first the tail tube assembles in the same way as for the Siphoviridae, followed by the contractile tail sheath. Tail completion proteins then bind and the tail joins to the head. Fibers also incorporate. The names of some of the implicated Mu proteins are shown in black lettering, and the corresponding T4 proteins in gray

72

M. Sanz-Gaitero et al.

contractile tail assembly (Büttner et al. 2016). To form the baseplate, wedge proteins are assembled into a wedge complex. The baseplate is then constructed through assembly of wedges around a central hub complex (Fig. 30). A dimer of Mu protein 47 (Mup47) initiates the formation of the wedge by binding to Mup48 (gp7) and subsequently to Mup46. Six wedge complexes assemble around the trimeric central hub complex made up out of Mup44, Mup45, and Mup43, forming the baseplate. In phage T4, a dimer of gp6 binds to a complex made up of a single copy of gp7, a trimer of gp10, and a dimer of gp8. Six of these wedge complexes come together to form the dome-shaped baseplate, around a hub complex consisting of a trimer of gp27, a trimer of gp5, and a monomer of gp5.4, plus some additional proteins (gp26, gp28, gp51, and gp29). The (gp5)3(gp5.4) complex is the spike. The six gp6 dimers form a tight ring around the trimer of gp27. The baseplate is completed by the addition of six copies of gp53, six trimers of gp9, and six (gp11)3(gp12)3 complexes (trimers of gp12 form the short tail fibers; Leiman et al. 2010; Arisaka et al. 2016). Once the T4 baseplate is complete, a complex consisting of a hexamer of gp48 and a hexamer of gp54 binds to the top of the central hub (Leiman et al. 2010; Arisaka et al. 2016). This complex is the equivalent of Mup43. The (gp48)6(gp54)6 complex primes the assembly of the inner tail tube, which is assembled in the same way as for the Siphoviridae, first assembling a tube consisting of chaperone proteins, which are then replaced by tail tube proteins. In phage T4, the tail tube consists of 138 copies of gp19, forming a six-start helix. The tape measure protein gp29 controls the length of the tail tube, and a hexameric ring of the tail tube terminator protein gp3 stabilizes the completed structure (purple in Fig. 30). Six copies of gp25 then bind to gp53, gp6, gp48, and gp54, to prime the assembly of the tail sheath, which is a six-start helix of gp18. The assembled tail sheath is in a high-energy conformation. A hexamer of gp15 binds to the top of the assembled tail and stabilizes it (orange in Fig. 30). Gp15 then binds to the neck proteins gp13 and gp14 of the capsid, joining the head to the tail. After the head and tail have joined together, in phage T4, the pre-assembled long tail fibers bind to the phage particle. A homo-trimer of gp34 makes up the proximal half fiber (nearest to the baseplate). The distal half fiber is composed of a trimer of gp36 and a trimer of gp37. For the correct folding of gp34 and gp37, but also for the short tail fiber gp12, the chaperone protein gp57 is necessary. In addition, in T4, gp37 also needs gp38 for correct folding (Bartual et al. 2010a). In other phages, like the Salmonella phage S16, gp38 stays bound to gp37 and is the de facto receptorbinding protein. The long tail fiber is completed when the amino-terminal end of the distal half-fiber and carboxy-terminal end of the proximal half-fiber complex are attached through a monomer of gp35 (Hyman and van Raaij 2018). In the case of phage T4, the completed long tail fiber is joined to the tail with the aid of the assembly protein gp63 and the neck fibers (fibritin). The amino-terminal domain of the gp34 homo-trimer binds to the gp9 trimer on the outer ring of the baseplate (Taylor et al. 2016). The long tail fibers are folded up along the tail and capsid, presumably to avoid strong interactions with host membrane fragments and to allow faster diffusion. The neck fibers hold the long tail fibers “up” most of the time, by binding the gp35 knee to the neck (Arisaka et al. 2016). In other myoviruses, like the coliphages Mu and P2, the long tail fibers are simpler and composed of a homo-

Structure and Function of Bacteriophages

73

trimeric protein, which needs a chaperone protein for correct folding (HaggårdLjungquist et al. 1992). This chaperone protein may dissociate after assembly, like in T4, or may stay bound to the distal end of the fiber, like in phage λ or Salmonella phage S16. Many myoviruses, especially those infecting Gram-positive bacteria, do not have tail fibers, but shorter trimeric receptor-binding proteins instead, as will be seen in the structure section below.

Myoviridae Tail Structure Myovirus tails show very variable lengths (between about 100 and 4500 nm), but their width is more conserved (between 18 and 24 nm) (Leiman and Shneider 2012). Sometimes, the phage tail is shorter than its head, like in T4 (Yap and Rossmann 2014); for other phages, it is considerably longer. The length of the tail is probably adapted to the host and may be related to the thickness and/or the toughness of the bacterial cell wall. Like for siphovirus tails, the length of the tail is regulated by the tape measure protein. The width of the tails is more conserved, because the folds of the tail tube and tail sheath proteins are conserved. In myoviruses, potential decoration domains are on the outside of the sheath. Here, we will discuss as an example the tail structure of coliphage T4, which is the most studied myovirus tail, and compare it with the Staphylococcus phage ϕ812 and the Listeria phage A551, as two examples of phages infecting Gram-positive bacteria. The receptor-binding fibers are discussed in a bit more detail. The tail tube structure of T4 and other myoviruses is basically the same as that of a siphovirus tail and consists of stacked hexameric rings of the tail tube protein gp19 (Zheng et al. 2017; Fig. 28g). In the case of the myoviruses, the tail tube protein is between 15 and 19 kDa, and, unlike for many siphoviruses, it is not decorated, due to the presence of the tail sheath around it. The phage T4 tail tube has 23 hexamers of gp19 and is capped on the top by a hexamer of gp3, which probably has the same fold and structural organization as the phage λ tail tube capping protein gpU (Fig. 27d). The phage T4 tail sheath is 24 nm wide and 93 nm long. It has 138 copies of gp18, organized in a six-start helix with a pitch of 4.1 nm and a twist of 17 (Fig. 31a, b; Leiman et al. 2004; Kostyuchenko et al. 2005). The contracted T4 sheath is 9 nm wider and 51 nm shorter (Fig. 31c; Arisaka et al. 2016). It is still a six-start helix, but the pitch has decreased to 1.6 nm and the twist has increased to 33 . During contraction, the gp18 subunits move, as rigid bodies, about 5 nm away from the tail center and tilt about 45 (Aksyuk et al. 2009a). Contraction of the sheath assembly is triggered by the baseplate, starts there, and propagates through the sheath in a wavelike motion (Guerrero-Ferreira et al. 2019). The phage T4 tail sheath is made up of the tail sheath protein gp18. Tail sheath proteins are between 40 and 80 kDa in size, with most around 45 kDa, and have a conserved fold (Leiman and Shneider 2012). The larger tail sheath proteins usually have a decoration domain that is displayed on the outside of the sheath. The T4 gp18 protein has four domains which are inserted into each other like Russian dolls

74

M. Sanz-Gaitero et al.

Fig. 31 Overall structure of the myovirus T4 extended and contracted tail. (a) Extended tail (EMDB entry EMD-1126). Three gp18 domain IV knobs belonging to the same helical strand are highlighted. (b) Extended tail as in panel a but with a reduced contour level to visualize the long tail fibers folded back against the tail. A red asterisk indicates where the carboxy-terminal domain of the fibritin contacts the knee of the long tail fiber. (c) Contracted tail (EMDB entry EMD-5528). Three gp18 domain IV knobs belonging to the same helical strand are highlighted, and the protruding part of the tail tube is shown as a gray rectangle

(Fig. 32a; Aksyuk et al. 2009a; Leiman and Shneider 2012): domain II is inserted into a loop of domain I, domain III is inserted into a loop of domain II, and domain IV, which is not present in tail sheath proteins of many other phages, is inserted into a loop of domain III. Although domain I was absent from the gp18 crystal structure, the structure can be inferred from a structurally homologous protein (PDB entry 3HXL). In the structure of domain I, one β-strand is donated by the amino-terminus of a molecule in the next row. Topological considerations suggest that this might also occur in the intact sheath. Furthermore, in the contractile tail sheath of the type VI secretion system, the carboxy-terminal end of the tail sheath homologue inserts into a molecule of the same row (Kudryashev et al. 2015), and this arrangement is likely conserved universally. This fishnet-like organization might be necessary to keep the sheath stable and help it to not fall apart during contraction (Leiman 2018). The conformation of an individual tail sheath protein does not change upon contraction. Domain I maintains the same interactions with neighboring subunits, but domains II and III change partners and actually increase their interaction surface by about four times (Aksyuk et al. 2009a), explaining why the contracted state is more stable.

Structure and Function of Bacteriophages

75

Fig. 32 Structure of the myovirus T4 tail proteins. (a) Structure of the tail sheath protein gp18 (PDB entry 3J2O; domains II–IV are from PDB entry 3FOA a domain I is modeled from PDB entry 3HXL). The strand exchange is illustrated: domain I contains a strand (blue arrow) from a neighboring gp18 molecule, while another strand that contributes to domain I of a neighbor on the other side is shown as a blue arrow connected with a dotted line. (b) Structure of the gp15 cap seen from the bottom, i.e., the side of the tail tube (PDB entry 4HUD). The six monomers of the hexameric ring are colored differently. (c, d) Relative positions of the tail sheath protein gp18 (blue and cyan) and the tail sheath capping protein gp15 (yellow and orange) in the extended (c; PDB entry 3J2M) tail and the contracted (d; PDB entry 3J2N) tail. (e) Model of the fibritin (gpWac) trimer (PDB entry 3J2O) with high-resolution structure of the amino-terminal and carboxy-terminal domains highlighted in blue and red boxes, respectively (PDB entry 1OX3). The three chains of the trimer are colored differently. The amino- and carboxy-termini for the full-length protein model are indicated

Domain I interacts with the tail tube, while domains II and III are partially exposed to solution, and domain IV is on the outside surface of the sheath. A hexameric ring of the bacteriophage T4 tail terminator protein, gp15, attaches to the top of the phage tail, covering the tail tube capping hexamer gp3 and stabilizing the contractile sheath by interacting with six gp18 molecules and preventing further polymerization of the tail sheath. The hexameric gp15 ring forms the interface for binding the phage head. Each gp15 monomer contains a curled-up eight-stranded antiparallel β-sheet, the center of which faces the inside of the ring

76

M. Sanz-Gaitero et al.

(Fig. 32b). The outside of the ring and the top, where the head-binding site is, are covered with α-helices. From the head side, a dodecamer of the adaptor protein gp13 is attached to the portal and to that a hexamer of gp14. When the head joins the tail, gp14 and gp15 form stable interactions. Once the phage head is attached to the tail, the neck of T4 virions is decorated by the collar and by whiskers. Twelve trimers of the fibritin protein (gpWac) bind to the neck; six of them are folded sideways around the neck and make up the collar, and another six point downward and form the whiskers (Fokine et al. 2013). Fibritin trimers forming the collar and the whiskers alternate. Each gpWac fiber consists of a long segmented, α-helical triple coiled coil with a small carboxy-terminal trimerization domain (Fig. 32e; Boudko et al. 2002; Boudko et al. 2004). The amino-terminal domain is bound to the phage neck, while the carboxy-terminal domain interacts with the long tail fiber knee (Fig. 31b). The whiskers act as chaperones, helping to attach the long tail fibers to the virus during the assembly

Fig. 33 The baseplate of bacteriophage T4. (a) Top view of the pre-attachment dome-shaped baseplate (PDB entry 5IV5). (b) Top view of the pre-attachment baseplate without the central tail tube and spike. (c) Side view of the pre-attachment dome-shaped baseplate. Black lines indicate the estimated location of the start of the long tail fibers. (d) Top view of the post-attachment T4 baseplate (PDB entry 5IV7). (e) Side view of the post-attachment T4 baseplate. The estimated locations of the three front-facing short tail fibers are indicated as red lines. (f) Side view of the T4 tail tube and its spike. A legend relating protein names to their colors is included

Structure and Function of Bacteriophages

77

process. The collar and whiskers are also environment-sensing devices, regulating the retraction or deployment of the long tail fibers under unfavorable or favorable conditions, thus preventing or promoting infection, respectively (Arisaka et al. 2016). The myovirus baseplate is responsible for coordinating correct host recognition with sheath contraction. In bacteriophage T4, the pre-attachment baseplate is domeshaped (Fig. 33a, b, c) and in a metastable high-energy state. It has overall pseudosixfold symmetry with a central trimeric hub. The baseplate central part or hub is surrounded by six wedges, to each of which a receptor-binding long tail fiber is bound. Each wedge also contains a copy of the short tail fiber. At the bottom end of the polymeric gp19 tail tube, a hexameric ring of gp54 is located and below that another hexameric ring of gp48 (Fig. 33f). Both these proteins have the typical folded-over β-sheet tail tube motif. Attached to the gp48 ring is the trimeric gp27 hub. Each of the gp27 monomers has two folded-over β-sheet tail tube domains (Kanamaru et al. 2002), adapting perfectly to the gp48 ring. The central spike is composed of three copies of gp5 and one of gp5.4. Each gp5 monomer has a lysozyme domain on the side, which hydrolyzes the peptidoglycan layer during cell wall penetration. The tip of the tail tube is a triple β-helical tube of gp5, capped by the pointed gp5.4 monomer. This sharp point helps penetrate the membrane (Browning et al. 2012). The baseplate can be divided into several parts (Taylor et al. 2016). The inner baseplate consists of a ring of 12 gp6 molecules (Aksyuk et al. 2009b). Six of these gp6 molecules are in one conformation and lie more on the inside, forming a constricted iris around the top of the hub and the bottom of the tube. Another six molecules are in a different conformation and alternate with the former, lying more toward the outside of the ring. The amino-terminal halves of the gp6 molecules are part of a (gp6)2gp7 heterotrimeric module, together forming an α-helical core bundle. Gp25 and gp53 also connect the core bundle to the central hub. Gp7 extends outward to the peripheral baseplate and connects with gp9 and gp10. A homotrimer of gp9 is the base to which the long tail fibers are connected (Kostyuchenko et al. 1999). Gp10 is a trimeric protein with a distorted X-shape. Each of its four domains (D1–D4) has threefold symmetry. D2 and D3 resemble each other. They interact with the amino-terminal regions of gp12 and gp11, respectively, orienting these proteins perpendicular to each other. The short tail fibers are extended trimers of the gp12 protein (van Raaij et al. 2001; Thomassen et al. 2003). Their amino-termini are bound to gp10 (Leiman et al. 2006). In the pre-attachment baseplate, the short tail fibers are bent and their knee region connects to a trimer of gp11, while their carboxy-terminal head domains interact with the amino-terminal part of a gp12 trimer and with gp10 from a neighboring wedge (Leiman et al. 2000). The D1 and D4 domains of gp10 interact with gp7 in the intermediate baseplate, including a disulfide bridge. There is also an intermolecular gp10-gp10 disulfide bridge. These covalent interactions provide extra stability to the baseplate (Taylor et al. 2016). Gp10, gp11, and gp12 share folds, so it is likely they resulted from each other by gene duplication events.

78

M. Sanz-Gaitero et al.

In T4, when at least three long tail fibers have bound a receptor molecule, a signal is transferred to the baseplate of the phage, which then changes conformation. The binding information transfer is likely related to the angle of attachment of the long tail fiber to the baseplate. In the free phage, this angle is variable and the fibers are flexible up to certain limits. When several fibers are attached to their receptor, external forces on the phage may force these angles to values outside this range, pushing proteins near the long fiber attachment site to a different conformation and triggering a sequential conformation change (Taylor et al. 2016). The baseplate flattens and acquires a six-pointed star shape (Fig. 33d, e). This conformational change is very extensive, involving changes of location and interaction partners for several baseplate proteins. During the transformation, the gp10/gp11/gp12 complex rotates as a unit (Taylor et al. 2016). At the end of this rotation, gp10, and thus the short tail fibers, point straight toward the host cell surface. Gp11 rotates upward, releasing its grip on the knee of the short tail fibers. The interaction of the carboxy-terminal head domain of gp12 with the amino-terminal part of gp12 and with gp10 of the neighboring wedge is also broken. This allows the gp12 short tail fibers to extend fully and reach the cell surface. Only the proximal (amino-terminal) end of the gp12 trimer remains bound to the baseplate, and the distal carboxy-terminal part forms a tight, irreversible interaction with the core region of the host lipopolysaccharide. Gp7 transfers the rotation of the gp10/gp11/gp12 complex to the (gp6)2gp7 heterotrimer. As a result, the diameter of the gp6 iris increases, and gp25 and gp53 dissociate from the tube and the hub. The movement of gp25 away from the tube also initiates sheath contraction. Gp18 subunits from the first ring, which are stacked on gp25, are pushed outward, and the whole sheath contracts as a wave, starting from the baseplate (Guerrero-Ferreira et al. 2019). The widening of the iris and the loosening of gp25 and gp53 allow the passage of a large part of the tail tube through the baseplate and through the bacterial cell wall, driven by sheath contraction. The tailspike, with its lysozyme domains having helped by locally degrading the peptidoglycan, probably falls off during this process and remains in the bacterial periplasm, leaving the end of the tail tube open, so the phage DNA can transfer directly into the host cytoplasm. Myoviruses infecting Gram-positive hosts tend not to have long side tail fibers. The structure of the Listeria phage A511 baseplate is simpler that of phage T4 (Guerrero-Ferreira et al. 2019). It contains the conserved tube-baseplate core complex, consisting of the tube proteins (gp19 and gp54 in T4), the baseplate hub proteins (gp48 and gp27 in T4), the tailspike protein (gp5 in T4), and the four wedge proteins (gp6, gp7, gp25, and gp56 in T4). These proteins are likely to be also conserved in simpler myoviruses that infect Gram-negative bacteria, like the phage P2 and Mu (Guerrero-Ferreira et al. 2019). Simpler baseplates contain only one trimeric protein (the host receptor-binding fiber) that is attached to gp7 (Leiman and Shneider 2012). More complicated phages like T4 developed gp9 and the long tail fibers for reversible host selection on the one hand and irreversible binding via gp10, gp11, and gp12 on the other hand. The low-resolution structures of the pre-attachment and post-attachment states of the Staphylococcus phage ϕ812 have also been determined (Nováček et al. 2016),

Structure and Function of Bacteriophages

79

and this baseplate is likely to contain the same basic framework also. In both A511 and ϕ812 phages, in the pre-attachment state, the receptor-binding proteins are pointing upward, toward the head of the phage. After DNA ejection, the baseplates transform into a double-layered structure, and the receptor-binding proteins change to a downward orientation, i.e., toward the bacterium, in a similar process as observed for the Lactococcus siphovirus p2. The most variable parts of myovirus baseplates are the receptor-binding proteins, presumably because they need to adapt to evolving host receptor molecules. They are also subject to domain exchange by horizontal gene transfer, which allows phages to change their host range (see chapter ▶ “Bacteriophage-Mediated Horizontal Gene Transfer: Transduction”). Receptor-binding proteins may be stubby proteins such as tailspikes, an example is the P22 gp9-like tailspike in the myovirus Det7 (Walter et al. 2008). As for the other Caudovirales members, enzymatic activities may be associated with these proteins, to allow hydrolysis of a bacterial capsule and access to the membrane. In other myoviruses, spindly fibers emanate from the baseplate. Some general features in receptor-binding proteins are conserved within their wide structural variation: they are usually trimeric, are anchored to the virus with their amino-terminal domains, and have carboxy-terminal receptor-binding domains. They are also largely composed of intertwined β-strands, which are likely important for their stability (Mitraki et al. 2006). For myoviruses with fibers, these may consist of single trimeric proteins, like in phages P2 and Mu (Haggård-Ljungquist et al. 1992). These proteins have their specific chaperones, which are required for correct folding and may remain attached to the distal end of the fiber in mature virions. In coliphage T4 and the Salmonella phage S16, the fibers are a complex of four or five different proteins (Fig. 34). In the mature virion, the long tail fibers are retracted upward and anchored to the collar by the knee (Hu et al. 2015). In bacteriophage T4, the structures of the carboxy-terminal parts of gp34 (Granell et al. 2017) and gp37 (Bartual et al. 2010b) are known (Fig. 34). Gp34 contains repeats of a mixed α/β fibrous domain in the aminoterminal two thirds of the protein. The carboxy-terminal third is composed of a triple β-helix domain punctuated by three β-prism domains, the last of which is decorated with long β-hairpins that may be involved in binding gp35. The structure of the receptor-binding tip of gp37 contains an elongated six-stranded antiparallel β-strand needle domain containing seven iron ions coordinated by histidine residues. At the end of the tip, the three chains intertwine to form a small head domain, which contains the putative receptor interaction site. For Salmonella phage S16, the crystal structure of the adhesin gp38 attached to the trimeric β-helical tip of gp37 has been determined (Dunne et al. 2018). The monomeric gp38 contains a small α-helical adaptor domain, a β-barrel domain, and a PGII sandwich with three layers of four polyglycine type II helices. The (gp37)3gp38 structure of phage S16 is conserved in other T4-like phages. The iron ion-containing needle motif can also be detected in gp37 of other T4-like phages, sometimes with eight or even nine putative iron ion sites. Other T4-like phages contain receptor-binding domains with yet to be discovered folds. Interestingly, the needle structure of the receptor-binding tip of the T4 gp37 is also conserved in the tip of the siphovirus λ side tail fibers.

80

M. Sanz-Gaitero et al.

Fig. 34 Long tail fibers. A schematic overview of the long tail fiber of bacteriophage T4 is shown, with gp34 in red, gp35 in green, gp36 in blue, and gp37 in yellow. Gray boxes show parts for which a crystal structure has been determined (PDB entries 4NXH for gp34 and 2XGF for gp37). These structures are shown as ribbon diagrams with their carboxy-termini indicated. At the bottom, the tip of the long tail fibers of the Salmonella phage S16 is shown (PDB entry 6F45). This crystal structure contains the C-terminal end of the trimeric gp37 protein bound to a single copy of gp38

Conclusions and Perspectives It has been proposed that viruses can be divided into a small number of structural lineages (Abrescia et al. 2012). All Caudovirales, plus the Herpesviridae, have the HK97 fold, which, given the abundance of tailed phages, may well be the most common protein fold on Earth. All Caudovirales, the Podoviridae, Siphoviridae, and Myoviridae, must be evolutionary related. They share capsids with the same organization and major capsid and portal protein folds. Adaptor proteins also have similar folds. The Podoviridae are structurally the most simple, so perhaps they existed first, and from them, the Siphoviridae evolved by acquiring the tail tube, followed by the Myoviridae by acquiring the contractile tail sheath. Interestingly, the contractile tail of the Myoviridae occurs in bacterial secretion systems and tailocins (Taylor et al. 2018), and it is possible that some bacteria have adopted a phage tail for their own purposes.

Structure and Function of Bacteriophages

81

The Tectiviridae and Corticoviridae, but also eukaryotic viruses like the Adenoviridae, form the second lineage, characterized by the double jelly roll fold. In these viruses, the jelly rolls are perpendicular to the capsid surface. The Microviridae, as well as many eukaryotic RNA viruses, have a capsid protein with a single jelly roll, which lies parallel to the capsid surface. These viruses form the third lineage. A fourth lineage contains the Cystoviridae but also the eukaryotic Reoviridae. The Leviviridae are a distinct lineage, infecting bacteria only. Recent studies, especially using metagenomics, suggest that additional lineages may exist and that the relative abundance of the known lineages may be different than currently assumed. For example, double jelly roll phages may be much more abundant than assumed (Yutin et al. 2018). Independent of how many different structural lineages and variants of them there are, there will be much more work for structural biologists to do in deciphering the structural diversity of phages, both in their overall structure and the detailed folds of their proteins. For many phage families, details about the assembly and structure are only known for a single or a few members. Future research will show whether the assembly mechanisms are general or whether interesting variations exist. The resolution revolution experienced in cryo-electron microscopy (see chapter ▶ “Detection of Bacteriophages: Electron Microscopy and Visualization”) will allow many detailed phage structures to be determined from purified whole phage particles, obviating the need for expressing and purifying all the structural proteins separately. However, crystallography and NMR spectroscopy will remain important for determining detailed structures of flexible phage proteins and their protein-ligand complexes. New and more precise data on the assembly and the structure of bacteriophages will have important implications for phage applications. Detailed knowledge of their assembly may allow for more efficient production of natural and synthetic phages (see chapter ▶ “Bacteriophage Manufacturing: From Early Twentieth-Century Processes to Current GMP”), as well as the design of phage variants as vaccination vehicles, and for drug delivery for gene therapy. Atomic models of phage structural proteins will allow targeted modification of these natural nanoparticles (see chapter ▶ “Bacteriophages in Nanotechnology: History and Future”). Structures of bacteriophage receptor-binding proteins bound to their receptor or a suitable analogue are relatively rare. This may be because the affinity of individual binding sites is low, and it is difficult to study these complexes in solution or in a crystal. However, bacteriophages, just like other viruses, recognize their host cells with multiple receptor-binding proteins that are often trimeric, and each can bind three receptor molecules simultaneously, leading to a strong avidity effect (LortatJacob et al. 2001). For example, 54 potential receptor-binding sites exist in Lactococcus phage TP901-1 (Veesler et al. 2012b). Detailed structural knowledge of the proteins that phages use for receptor recognition and their complexes with receptor analogues may allow the generation of phage mutants with the desired altered host ranges.

82

M. Sanz-Gaitero et al.

Cross-References ▶ Bacteriophage Discovery and Genomics ▶ Bacteriophage Manufacturing: From Early Twentieth-Century Processes to Current GMP ▶ Bacteriophage-Mediated Horizontal Gene Transfer: Transduction ▶ Bacteriophages in Nanotechnology: History and Future ▶ Bacteriophage Use in Molecular Biology and Biotechnology ▶ Detection of Bacteriophages: Electron Microscopy and Visualization ▶ Genetics and Genomics of Bacteriophages ▶ Phage Infection and Lysis Acknowledgments Structure figures were generated using the PYMOL Molecular Graphics System (Schrödinger LLC) and UCSF CHIMERA (Pettersen et al. 2004). The research in our lab is funded by grants BFU2017-82207-P from the Spanish Ministry of Science, Innovation and Universities, State Agency of Research, co-financed by the European Regional Development Fund of the European Union. We thank Antonio Pichel for help with the sections on the Microviridae and Cystoviridae and for preparing Figs. 5 and 9. We are also grateful to Carmen San Martín (CNB-CSIC) and Carmela García-Doval (University of Zurich) for proofreading, to Don Marvin for advice on inovirus structure, and to Petr Leiman (University of Texas Medical Branch) for advice on figures; any remaining mistakes are the responsibility of the authors.

References Abrescia NG, Cockburn JJ, Grimes JM, Sutton GC, Diprose JM, Butcher SJ, Fuller SD, San Martín C, Burnett RM, Stuart DI, Bamford DH, Bamford JKH (2004) Insights into assembly from structural analysis of bacteriophage PRD1. Nature 432:68–74 Abrescia NG, Grimes JM, Kivelä HM, Assenberg R, Sutton GC, Butcher SJ, Bamford JK, Bamford DH, Stuart DI (2008) Insights into virus evolution and membrane biogenesis from the structure of the marine lipid-containing bacteriophage PM2. Mol Cell 31:749–761 Abrescia NG, Bamford DH, Grimes JM, Stuart DI (2012) Structure unifies the viral universe. Annu Rev Biochem 81:795–822 Ackermann HW, Prangishvili D (2012) Prokaryote viruses studied by electron microscopy. Arch Virol 157:1843–1849 Ageno M, Donelli G, Guglielmi F (1973) Structure and physico-chemical properties of bacteriophage G. II, the shape and symmetry of the capsid. Micron 4:376–403 Aksyuk AA, Leiman PG, Kurochkina LP, Shneider MM, Kostyuchenko VA, Mesyanzhinov VV, Rossmann MG (2009a) The tail sheath structure of bacteriophage T4: a molecular machine for infecting bacteria. EMBO J 28:821–829 Aksyuk AA, Leiman PG, Shneider MM, Mesyanzhinov VV, Rossmann MG (2009b) The structure of gene product 6 of bacteriophage T4, the hinge-pin of the baseplate. Structure 17:800–808 Arisaka F, Yap ML, Janamaru S, Rossmann MG (2016) Molecular assembly and structure of the bacteriophage T4 tail. Biophys Rev 8:385–396 Arnaud CA, Effantin G, Vivès C, Engilberge S, Bacia M, Boulanger P, Girard E, Schoehn G, Breyton C (2017) Bacteriophage T5 tail tube structure suggests a trigger mechanism for Siphoviridae DNA ejection. Nat Commun 8:1953 Bamford DH, Palva ET, Lounatmaa K (1976) Ultrastructure and life cycle of the lipid-containing bacteriophage ϕ6. J Gen Virol 32:249–259

Structure and Function of Bacteriophages

83

Baptista C, Santos MA, São-José C (2008) Phage SPP1 reversible adsorption to Bacillus subtilis cell wall teichoic acids accelerates virus recognition of membrane receptor YueB. J Bacteriol 190:4989–4996 Bartual SG, Garcia-Doval C, Alonso J, Schoehn G, van Raaij MJ (2010a) Two-chaperone assisted soluble expression and purification of the bacteriophage T4 long tail fibre protein gp37. Protein Expr Purif 70:116–121 Bartual SG, Otero JM, Garcia-Doval C, Llamas-Saiz AL, Kahn R, Fox GC, van Raaij MJ (2010b) Structure of the bacteriophage T4 long tail fiber receptor-binding tip. Proc Natl Acad Sci U S A 107:20287–20292 Bebeacua C, Bron P, Lai L, Vegge CS, Brøndsted L, Spinelli S, Campanacci V, Veesler D, van Heel M, Cambillau C (2010) Structure and molecular assignment of lactococcal phage TP901-1 baseplate. J Biol Chem 285:39079–39086 Bernal RA, Hafenstein S, Olson NH, Bowman VD, Chipman PR, Baker TS, Fane BA, Rossmann MG (2003) Structural studies of bacteriophage α3 assembly. J Mol Biol 325:11–24 Bernhardt TG, Wang IN, Struck DK, Young R (2001) A protein antibiotic in the phage Qβ virion: diversity in lysis targets. Science 292:2326–2327 Bhardwaj A, Molineux IJ, Casjens SR, Cingolani G (2011) Atomic structure of bacteriophage Sf6 tail needle knob. J Biol Chem 286:30867–30877 Bhardwaj A, Olia AS, Cingolani G (2014) Architecture of viral genome-delivery molecular machines. Curr Opin Struct Biol 25:1–8 Black LW, Thomas JA (2012) Condensed genome structure. Adv Exp Med Biol 726:469–487 Bollback JP, Huelsenbeck JP (2001) Phylogeny, genome evolution, and host specificity of singlestranded RNA bacteriophage (family Leviviridae). J Mol Evol 52:117–128 Boudko SP, Londer YY, Letarov AV, Sernova NV, Engel J, Mesyanzhinov VV (2002) Domain organization, folding and stability of bacteriophage T4 fibritin, a segmented coiled-coil protein. Eur J Biochem 269:833–841 Boudko SP, Strelkov SV, Engel J, Stetefeld J (2004) Design and crystal structure of bacteriophage T4 mini-fibritin NCCF. J Mol Biol 339:927–935 Brewer GJ (1978) Membrane-localized replication of bacteriophage PM2. Virology 84:242–245 Browning C, Shneider MM, Bowman VD, Schwarzer D, Leiman PG (2012) Phage pierces the host cell membrane with the iron-loaded spike. Structure 20:326–339 Butcher SJ, Manole V, Karhu NJ (2012) Lipid-containing viruses: bacteriophage PRD1 assembly. Adv Exp Med Biol 726:365–377 Büttner CR, Wu Y, Maxwell KL, Davidson AR (2016) Baseplate assembly of phage mu: defining the conserved core components of contractile-tailed phages and related bacterial systems. Proc Natl Acad Sci U S A 113:10174–10179 Caldentey J, Bamford DH (1992) The lytic enzyme of the Pseudomonas phage ϕ6. Purification and biochemical characterization. Biochim Biophys Acta 1159:44–50 Canelo E, Phillips OM, del Roure RN (1985) Relating cistrons and functions in bacteriophage PM2. Virology 140:364–367 Casjens S, King J (1975) Virus assembly. Annu Rev Biochem 44:555–611 Casjens SR, Molineux IJ (2012) Short noncontractile tail machines: adsorption and DNA delivery by podoviruses. Adv Exp Med Biol 726:143–179 Caspar DL, Klug A (1962) Physical principles in the construction of regular viruses. Cold Spring Harb Symp Quant Biol 27:1–24 Chaban Y, Lurz R, Brasiles S, Cornilleau C, Karreman M, Zinn-Justin S, Tavares P, Orlova EV (2015) Structural rearrangements in the phage head-to-tail interface during assembly and infection. Proc Natl Acad Sci U S A 112:7009–7014 Chipman PR, Agbandje-McKenna M, Renaudin J, Baker TS, McKenna R (1998) Structural analysis of the Spiroplasma virus, SpV4: implications for evolutionary variation to obtain host diversity among the Microviridae. Structure 6:135–145

84

M. Sanz-Gaitero et al.

Choi KH, McPartland J, Kaganman I, Bowman VD, Rothman-Denes LB, Rossmann MG (2008) Insight into DNA and protein transport in double-stranded DNA viruses: the structure of bacteriophage N4. J Mol Biol 378:726–736 Cockburn JJ, Abrescia NG, Grimes JM, Sutton GC, Diprose JM, Benevides JM, Thomas GJ, Bamford DH, Bamford JK, Stuart DI (2004) Membrane structure and interactions with protein and DNA in bacteriophage PRD1. Nature 432:122–125 Cuervo A, Pulido-Cid M, Chagoyen M, Arranz R, González-García V, Garcia-Doval C, Castón JR, Valpuesta JM, van Raaij MJ, Martín-Benito K, Carrascosa JL (2013) Structural characterization of the bacteriophage T7 tail. J Biol Chem 288:26290–26299 Dai X, Li Z, Lai M, Shu S, Du Y, Zhou ZH, Sun R (2017) In situ structures of the genome and genome-delivery apparatus in a single-stranded RNA virus. Nature 541:112–116 Davidson AR, Cardarelli L, Pell LG, Radford DR, Maxwell KL (2012) Long noncontractile tail machines of bacteriophages. Adv Exp Med Biol 726:115–142 Doore SM, Fane BA (2016) The Microviridae: diversity, assembly, and experimental evolution. Virology 461:45–55 Dunne M, Denyes JM, Arndt H, Loessner MJ, Leiman PG, Klumpp J (2018) Salmonella phage S16 tail fiber adhesin features a rare polyglycine rich domain for host recognition. Structure 26:1573–1582 El Omari K, Meier C, Kainov D, Sutton G, Grimes JM, Poranen MM, Bamford DH, Tuma R, Stuart DI, Mancini EJ (2013a) Tracking in atomic detail the functional specializations in viral RecA helicases that occur during evolution. Nucleic Acids Res 41:9396–9410 El Omari K, Sutton G, Ravantti JJ, Zhang HW, Walter TS, Grimes JM, Bamford DH, Stuart DI, Mancini EJ (2013b) Plate tectonics of virus shell assembly. Structure 21:1384–1395 Espejo RT, Canelo ES, Sinsheimer RL (1969) DNA of bacteriophage PM2: a closed circular double-stranded molecule. Proc Natl Acad Sci U S A 63:1164–1168 Farley MM, Tu J, Kearns DB, Molineaux IJ, Liu J (2017) Ultrastructural analysis of bacteriophage ϕ29 during infection of Bacillus subtilis. J Struct Biol 197:163–171 Feng JN, Model P, Russel M (1999) A trans-envelope protein complex needed for filamentous phage assembly and export. Mol Microbiol 34:745–755 Fiers W, Contreras R, Duerinck F, Haegeman G, Iserentant D, Merregaert J, Min Jou W, Molemans F, Raeymaekers A, Van den Berghe A, Volckaert G, Ysebaert M (1976) Complete nucleotide sequence of bacteriophage MS2 RNA: primary and secondary structure of the replicase gene. Nature 260:500–507 Flayhan A, Vellieux FM, Lurz R, Maury O, Contreras-Martel C, Girard E, Boulanger P, Breyton C (2014) Crystal structure of pb9, the distal tail protein of bacteriophage T5: a conserved structural motif among all siphophages. J Virol 88:820–828 Fokine A, Rossmann MG (2014) Molecular architecture of tailed double-stranded DNA phages. Bacteriophage 4:e28281 Fokine A, Chipman PR, Leiman PG, Mesyanzhinov VV, Rao VB, Rossmann MG (2004) Molecular architecture of the prolate head of bacteriophage T4. Proc Natl Acad Sci 101:6003–6008 Fokine A, Islam MZ, Zhang Z, Bowman VD, Rao VB, Rossmann MG (2011) Structure of the three N-terminal immunoglobulin domains of the highly immunogenic outer capsid protein from a T4-like bacteriophage. J Virol 85:8141–8148 Fokine A, Zhang Z, Kanamaru S, Bowman VD, Aksyuk AA, Arisaka F, Rao VB, Rossmann MG (2013) The molecular architecture of the bacteriophage T4 neck. J Mol Biol 425:1731–1744 Frilander M, Bamford DH (1995) In vitro packaging of the single-stranded RNA genomic precursors of the segmented double-stranded RNA bacteriophage ϕ6: the three segments modulate each other’s packaging efficiency. J Mol Biol 246:418–428 Garcia-Doval C, van Raaij MJ (2012) Structure of the receptor-binding carboxy-terminal domain of bacteriophage T7 tail fibers. Proc Natl Acad Sci U S A 109:9390–9395 Garcia-Doval C, van Raaij MJ (2013) Bacteriophage receptor recognition and nucleic acid transfer. Subcell Biochem 68:489–518

Structure and Function of Bacteriophages

85

Garcia-Doval C, Castón JR, Luque D, Granell M, Otero JM, Llamas-Saiz AL, Renouard M, Boulanger P, van Raaij MJ (2015) Structure of the receptor-binding carboxy-terminal domain of the bacteriophage T5 L-shaped tail fibre with and without its intra-molecular chaperone. Viruses 7:6424–6440 Golmohammadi R, Valegard K, Fridborg K, Liljas L (1993) The refined structure of bacteriophage MS2 at 2.8 Å resolution. J Mol Biol 234:620–639 Golmohammadi R, Fridborg K, Bundule M, Valegard K, Liljas L (1996) The crystal structure of bacteriophage Qβ at 3.5 Å resolution. Structure 4:543–554 González-García VA, Pulido-Cid M, Garcia-Doval C, Bocanegra R, van Raaij MJ, Martin-Benito J, Cuervo A, Carrascosa JL (2015) Conformational changes leading to T7 DNA delivery upon interaction with the bacterial receptor. J Biol Chem 290:10038–10044 Gorzelnik KV, Cui Z, Reed CA, Jakana J, Young R, Zhang J (2016) Asymmetric cryo-EM structure of the canonical Allolevivirus Qβ reveals a single maturation protein and the genomic ssRNA in situ. Proc Natl Acad Sci U S A 113:11519–11524 Granell M, Namura M, Alvira S, Kanamaru S, van Raaij MJ (2017) Crystal structure of the carboxy-terminal region of the bacteriophage T4 proximal long tail fiber protein gp34. Viruses 9:E168 Guerrero-Ferreira RC, Hupfeld M, Nazarov S, Taylor NM, Shneider MM, Obbineni JM, Loessner MJ, Ishikawa T, Klumpp J, Leiman PG (2019) Structure and transformation of bacteriophage A511 baseplate and tail upon infection of Listeria cells. EMBO J 38:e99455 Haggård-Ljungquist E, Halling C, Calendar R (1992) DNA sequences of the tail fiber genes of bacteriophage P2: evidence for horizontal transfer of tail fiber genes among unrelated bacteriophages. J Bacteriol 174:1462–1477 Henderson R (2015) Overview and future of single particle electron cryomicroscopy. Arch Biochem Biophys 581:19–24 Higashitani A, Higashitani N, Horiuchi K (1997) Minus-strand origin of filamentous phage versus transcriptional promoters in recognition of RNA polymerase. Proc Natl Acad Sci U S A 94:2909–2914 Higman VA (2013) Nuclear magnetic resonance methods for studying soluble, fibrous, and membrane-embedded proteins. In: Rusa JM, Piñeiro A (eds) Proteins in solution and at interfaces: methods and applications in biotechnology and materials science. Wiley, New York, pp 23–48 Holliger P, Riechmann L, Williams RL (1999) Crystal structure of the two N-terminal domains of g3p from filamentous phage fd at 1.9 Å: evidence for conformational lability. J Mol Biol 288:649–657 Hong C, Oksanen HM, Liu X, Jakana J, Bamford DH, Chiu W (2014) A structural model of the genome packaging process in a membrane-containing double stranded DNA virus. PLoS Biol 12:e1002024 Hu GB, Wei H, Rice WJ, Stokes DL, Paul Gottlieb P (2008) Electron cryo-tomographic structure of cystovirus ϕ12. Virology 372:1–9 Hu B, Margolin W, Molineaux IJ, Liu J (2013) The bacteriophage T7 virion undergoes extensive structural remodeling during infection. Science 339:577–579 Hu B, Margolin W, Molineux IJ, Liu J (2015) Structural remodelling of bacteriophage T4 and host membranes during infection initiation. Proc Natl Acad Sci U S A 112:E4919–E4928 Hua J, Huet A, Lopez CA, Toropova K, Pope WH, Duda RL, Hendrix RW, Conway JF (2017) Capsids and genomes of jumbo-sized bacteriophages reveal the evolutionary reach of the HK97 fold. MBio 8:e01579-01517 Huiskonen JT, Kivelä HM, Bamford DH, Butcher SJ (2004) The PM2 virion has a novel organization with an internal membrane and pentameric receptor binding spikes. Nat Struct Mol Biol 11:850–856 Huiskonen JT, Manole V, Butcher SJ (2007) Tale of two spikes in bacteriophage PRD1. Proc Natl Acad Sci U S A 104:6666–6671

86

M. Sanz-Gaitero et al.

Hyman P, van Raaij MJ (2018) Bacteriophage T4 long tail fiber domains. Biophys Rev 10:463–471 Inagaki M, Kawaura T, Wakashima H, Kato M, Nishikawa S, Kashimura N (2003) Different contributions of the outer and inner R-core residues of lipopolysaccharide to the recognition by spike H and G proteins of bacteriophage ϕX174. FEMS Microbiol Lett 226:221–227 Jäälinoja HT, Huiskonen JT, Butcher SJ (2007) Electron cryomicroscopy comparison of the architectures of the enveloped bacteriophages ϕ6 and ϕ8. Structure 15:157–167 Johnson JE, Speir JA (1997) Quasi-equivalent viruses: a paradigm for protein assemblies. J Mol Biol 269:665–675 Kanamaru S, Leiman PG, Kostyuchenko VA, Chipman PR, Mesyanzhinov VV, Arisaka F, Rossmann MG (2002) Structure of the cell-puncturing device of bacteriophage T4. Nature 415:553–557 Karlsson F, Borrebaeck CA, Nilsson N, Malmborg-Hager AC (2003) The mechanism of bacterial infection by filamentous phages involves molecular interactions between TolA and phage protein 3 domains. J Bacteriol 185:2628–2634 Katsura I (1990) Mechanism of length determination in bacteriophage lambda tails. Adv Biophys 26:1–18 Kikuchi Y, King J (1975) Assembly of the tail of bacteriophage T4. J Supramol Struct 3:24–38 Kivelä HM, Männistö RH, Kalkkinen N, Bamford DH (1999) Purification and protein composition of PM2, the first lipid-containing bacterial virus to be isolated. Virology 262:364–374 Kivelä HM, Kalkkinen N, Bamford DH (2002) Bacteriophage PM2 has a protein capsid surrounding a spherical proteinaceous lipid core. J Virol 76:8169–8178 Kivelä HM, Daugelavičius R, Hankkio RH, Bamford JKH, Bamford DH (2004) Penetration of membrane-containing double-stranded-DNA bacteriophage PM2 into Pseudoalteromonas hosts. J Bacteriol 186:5342–5354 Kivelä HM, Abrescia NGA, Bamford JKH, Grimes JM, Stuart DI, Bamford DH (2008) Selenomethionine labeling of large biological macromolecular complexes: probing the structure of marine bacterial virus PM2. J Struct Biol 161:204–210 Koç C, Xia G, Kühner P, Spinelli S, Roussel A, Cambillau C, Stehle T (2016) Structure of the hostrecognition device of Staphylococcus aureus phage ϕ11. Sci Rep 6:27581 Korasick DA, Tanner JJ (2018) Determination of protein oligomeric structure from small-angle X-ray scattering. Protein Sci 27:814–824 Kostyuchenko VA, Navruzbekov GA, Kurochkina LP, Strelkov SV, Mesyanzhinov VV, Rossmann MG (1999) The structure of bacteriophage T4 gene product 9: the trigger for tail contraction. Structure 7:1213–1222 Kostyuchenko VA, Chipman PR, Leiman PG, Arisaka F, Mesyanzhinov VV, Rossmann MG (2005) The tail structure of bacteriophage T4 and its mechanism of contraction. Nat Struct Mol Biol 12:810–813 Krupovic M, Daugelavicius R, Bamford DH (2007) A novel lysis system in PM2, a lipid-containing marine double-stranded DNA bacteriophage. Mol Microbiol 64:1635–1648 Kudryashev M, Wang RY, Brackmann M, Scherer S, Maier T, Baker D, DiMaio F, Stahlberg H, Egelman EH, Basler M (2015) Structure of the type VI secretion system contractile sheath. Cell 160:952–962 Kühlbrandt W (2014) Cryo-EM enters a new era. elife 3:e03678 Lavigne R, Darius P, Summer EJ, Seto D, Mahdevan P, Nilson AS, Ackermann KAM (2009) Classification of Myoviridae bacteriophages using protein sequence similarity. BMC Microbiol 9:224 Leiman PG (2018) Stretching the arms of the type VI secretion sheath protein. EMBO Rep 19:191–193 Leiman PG, Shneider MM (2012) Contractile tail machines of bacteriophages. Adv Exp Med Biol 726:93–114 Leiman PG, Kostyuchenko VA, Shneider MM, Kurochkina LP, Mesyanzhinov VV, Rossmann MG (2000) Structure of bacteriophage T4 gene product 11, the interface between the baseplate and short tail fibers. J Mol Biol 301:975–985

Structure and Function of Bacteriophages

87

Leiman PG, Chipman PR, Kostyuchenko VA, Mesyanzhinov VV, Rossmann MG (2004) Threedimensional rearrangement of proteins in the tail of bacteriophage T4 on infection of its host. Cell 118:419–429 Leiman PG, Shneider MM, Mesyanzhinov VV, Rossmann MG (2006) Evolution of bacteriophage tails: structure of T4 gene product 10. J Mol Biol 358:912–921 Leiman PG, Battisti AJ, Bowman VD, Stumeyer K, Mühlenhoff M, Gerardy-Schahn R, Scholl D, Molineaux IJ (2007) The structures of bacteriophages K1E and K1-5 explain processive degradation of polysaccharide capsules and evolution of new host specificities. J Mol Biol 371:836–849 Leiman PG, Arisaka F, van Raaij MJ, Kostyuchenko VA, Aksyuk AA, Kanamaru S, Rossmann MG (2010) Morphogenesis of the T4 tail and tail fibers. Virol J 7:355 Lhuillier S, Gallopin M, Gilquin B, Brasilès S, Lancelot N, Letellier G, Gilles M, Dethan G, Orlova EV, Couprie J, Tavares P, Zinn-Justin S (2009) Structure of bacteriophage SPP1 head-to-tail connection reveals mechanism for viral DNA gating. Proc Natl Acad Sci U S A 106:8507–8512 Li X, Koç C, Kühner P, Stierhof YD, Krismer B, Enright MC, Penadés JR, Wolz C, Stehle T, Cambillau C, Peschel A, Xia G (2016) An essential role for the baseplate protein gp45 in phage adsorption to Staphylococcus aureus. Sci Rep 6:26455 Liu Y, Eisenberg D (2002) 3D domain swapping: as domains continue to swap. Protein Sci 11:1285–1299 Liu X, Zhang Q, Murata K, Baker ML, Sullivan MB, Fu C, Dougherty MT, Schmid MF, Osburne MS, Chisholm SW, Chiu W (2010) Structural changes in a marine podovirus associated with release of its genome into Prochlorococcus. Nat Struct Mol Biol 17:830–836 Llamas-Saiz AL, van Raaij MJ (2013) X-ray crystallography of biological macromolecules: fundamentals and applications. In: Rusa JM, Piñeiro A (eds) Proteins in solution and at interfaces: methods and applications in biotechnology and materials science. Wiley, New York, pp 3–22 Lortat-Jacob H, Chouin E, Cusack S, van Raaij MJ (2001) Kinetic analysis of adenovirus fiber binding to its receptor reveals an avidity mechanism for trimeric receptor-ligand interactions. J Biol Chem 276:9009–9015 Lubkowski J, Hennecke F, Plückthun A, Wlodawer A (1999) Filamentous phage infection: crystal structure of g3p in complex with its coreceptor, the C-terminal domain of TolA. Structure 7:711–722 Mahony J, Stockdale SR, Collins B, Spinelli S, Douillard FP, Cambillau C, van Sinderen D (2016) Lactococcus lactis phage TP901-1 as a model for Siphoviridae virion assembly. Bacteriophage 6:e1123795 Männistö RH, Kivelä HM, Paulin L, Bamford DH, Bamford JK (1999) The complete genome sequence of PM2, the first lipid-containing bacterial virus to be isolated. Virology 262:355–336 Mäntynen S, Sundberg LR, Poranen MM (2018) Recognition of six additional cystoviruses: Pseudomonas virus phi6 is no longer the sole species of the family Cystoviridae. Arch Virol 163:1117–1124 Marvin DA (1998) Filamentous phage structure, infection and assembly. Curr Opin Struct Biol 8:150–158 Marvin DA (2017) Fibre diffraction studies of biological macromolecules. Prog Biophys Mol Biol 127:43–87 Marvin DA, Welsh LC, Symmons MF, Scott WR, Straus SK (2006) Molecular structure of fd (f1, M13) filamentous bacteriophage refined with respect to X-ray fibre diffraction and solid-state NMR data supports specific models of phage assembly at the bacterial membrane. J Mol Biol 355:294–309 Marvin DA, Symmons MF, Straus SK (2014) Structure and assembly of filamentous bacteriophages. Prog Biophys Mol Biol 114:80–122 Maxwell KL, Yee AA, Booth V, Arrowsmith CH, Gold M, Davidson AR (2001) The solution structure of bacteriophage λ protein W, a small morphogenetic protein possessing a novel fold. J Mol Biol 308:9–14

88

M. Sanz-Gaitero et al.

Maxwell KL, Yee AA, Arrowsmith CH, Gold M, Davidson AR (2002) The solution structure of the bacteriophage λ head-tail joining protein, gpFII. J Mol Biol 318:1395–1404 Merckel MC, Huiskonen JT, Bamford DH, Goldman A, Tuma R (2005) The structure of the bacteriophage PRD1 spike sheds light on the evolution of viral capsid architecture. Mol Cell 18:161–170 Mitraki A, Papanikolopoulou K, van Raaij MJ (2006) Natural triple β-stranded fibrous folds. Adv Protein Chem 73:97–124 Moak M, Molineux IJ (2004) Peptidoglycan hydrolytic activities associated with bacteriophage virions. Mol Microbiol 51:1169–1183 Morais MC, Choi KH, Koti JS, Chipman PR, Anderson DL, Rossmann MG (2005) Conservation of the capsid structure in tailed dsDNA bacteriophages: the pseudoatomic structure of ϕ29. Mol Cell 18:149–159 Moreno-Madrid F, Martín-González N, Llauró A, Ortega-Esteban A, Hernando-Pérez M, Douglas T, Schaap IA, de Pablo PJ (2017) Atomic force microscopy of virus shells. Biochem Soc Trans 45:499–511 Nemecek D, Gilcrease EB, Kang S, Prevelige PE, Casjens S, Thomas GJ (2007) Subunit conformations and assembly states of a DNA-translocating motor: the terminase of bacteriophage P22. J Mol Biol 374:817–836 Nemecek D, Boura E, Wu W, Cheng N, Plevka P, Qiao J, Mindich L, Heymann JB, Hurley JH, Steven AC (2013) Subunit folds and maturation pathway of a dsRNA virus capsid. Structure 21:1374–1383 Ni CZ, Syed R, Kodandapani R, Wickersham J, Peabody DS, Ely KR (1995) Crystal structure of the MS2 coat protein dimer: implications for RNA binding and virus assembly. Structure 3:255–263 Nováček J, Šiborová S, Benešík M, Pantůček R, Doškař J, Plevka P (2016) Structure and genome release of Twort-like Myoviridae phage with a double-layered baseplate. Proc Natl Acad Sci U S A 113:9351–9356 Olia AS, Casjens S, Cingolani G (2007) Structure of phage P22 cell envelope–penetrating needle. Nat Struct Mol Biol 14:1221–1226 Olia AS, Prevelige PE, Johnson JE, Cingolani G (2011) Three-dimensional structure of a viral genome-delivery portal vertex. Nat Struct Mol Biol 18:597–603 Oliveira LM, Ye Z, Katz A, Alimova A, Wei H, Herman GT, Gottlieb P (2018) Component tree analysis of cystovirus ϕ6 nucleocapsid cryo-EM single particle reconstructions. PLoS One 13: e0188858 Parent KN, Schrad JR, Cingolani G (2018) Breaking symmetry in viral icosahedral capsids as seen through the lenses of X-ray crystallography and cryo-electron microscopy. Viruses 10:67 Pell LG, Liu A, Edmonds L, Donaldson LW, Howell PL, Davidson AR (2009a) The X-ray crystal structure of the phage λ tail terminator protein reveals the biologically relevant hexameric ring structure and demonstrates a conserved mechanism of tail termination among diverse long-tailed phages. J Mol Biol 389:938–951 Pell LG, Kanelis V, Donaldson LW, Howell PL, Davidson AR (2009b) The phage λ major tail protein structure reveals a common evolution for long-tailed phages and the type VI bacterial secretion system. Proc Natl Acad Sci U S A 106:4160–4165 Pell LG, Gasmi-Seabrook GM, Morais M, Neudecker P, Kanelis V, Bona D, Donaldson LW, Edwards AM, Howell PL, Davidson AR, Maxwell KL (2010 Oct 29) The solution structure of the C-terminal Ig-like domain of the bacteriophage λ tail tube protein. J Mol Biol 403 (3):468–479 Perucchetti R, Parris W, Becker A, Gold M (1988) Late stages in bacteriophage λ head morphogenesis: in vitro studies on the action of the bacteriophage λ D-gene and W-gene products. Virology 165:103–114 Petrovski S, Dyson ZA, Seviour RJ, Tillet D (2012) Small but sufficient: the Rhodococcus phage RRH1 has the smallest known Siphoviridae genome at 14.2 kilobases. J Virol 86:358–363 Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, Ferrin TE (2004) UCSF Chimera, a visualization system for exploratory research and analysis. J Comput Chem 25:1605–1612

Structure and Function of Bacteriophages

89

Philippe C, Krupovic M, Jaomanjaka F, Claisse O, Petrel M, le Marrec M (2018) Bacteriophage GC1, a novel Tectivirus infecting Gluconobacter cerinus, an acetic acid bacterium associated with wine-making. Viruses 10:39 Plevka P, Tars K, Liljas L (2008) Crystal packing of a bacteriophage MS2 coat protein mutant corresponds to octahedral particles. Protein Sci 17:1731–1739 Poranen MM, Tuma R (2004) Self-assembly of double-stranded RNA bacteriophages. Virus Res 101:93–100 Prasad BV, Schmid MF (2012) Principles of virus structural organization. Adv Exp Med Biol 726:17–47 Prevelige PE, Cortines JR (2018) Phage assembly and the special role of the portal protein. Curr Opin Virol 31:66–73 Prevelige PE, Fane BA (2012) Building the machines: scaffolding proteins functions during bacteriophage morphogenesis. Adv Exp Med Biol 726:325–350 Rakonjac J, Bennett NJ, Spagnuolo J, Gagic D, Russel M (2011) Filamentous bacteriophage: biology, phage display and nanotechnology applications. Curr Issues Mol Biol 13:51–57 Rakonjac J, Russel M, Khanum S, Brooke SJ, Rajič M (2017) Filamentous phage: structure and biology. Adv Exp Med Biol 1053:1–20 Rao VB, Feiss M (2015) Mechanisms of DNA packaging by large double-stranded DNA viruses. Annu Rev Virol 9:351–378 Ross PD, Cheng N, Conway JF, Firek BA, Hendrix RW, Duda RL, Steven AC (2005) Crosslinking renders bacteriophage HK97 capsid maturation irreversible and effects an essential stabilization. EMBO J 24:1352–1363 Rumnieks J, Tars K (2014) Crystal structure of the bacteriophage Qβ coat protein in complex with the RNA operator of the replicase gene. J Mol Biol 426:1039–1049 Russel M, Linderoth NA, Sali A (1997) Filamentous phage assembly: variation on a protein export theme. Gene 192:23–32 Santos-Pérez I, Oksanen HM, Bamford DH, Goñi FM, Reguera D, Abrescia NGA (2017) Membrane-assisted viral DNA ejection. Biochim Biophys Acta 1861:664–672 Schulz EC, Dickmanns A, Urlaub H, Schmitt A, Mühlenhoff M, Stummeyer K, Schwarzer D, Gerardy-Schahn R, Ficner R (2010a) Crystal structure of an intramolecular chaperone mediating triple-β-helix folding. Nat Struct Mol Biol 17:210–215 Schulz EC, Schwarzer D, Frank M, Stummeyer K, Mühlenhoff M, Dickmanns A, Gerardy-SchahnR, Ficner R (2010b) Structural basis for the recognition and cleavage of polysialic acid by the bacteriophage K1F tailspike protein EndoNF. J Mol Biol 397:341–351 Sciara G, Bebeacua C, Bron P, Tremblay D, Ortiz-Lombardia M, Lichière J, van Heel M, Campanacci V, Moineau S, Cambillau C (2010) Structure of lactococcal phage p2 baseplate and its mechanism of activation. Proc Natl Acad Sci U S A 107:6852–6857 Seul A, Müller JJ, Andres D, Stettner E, Heinemann U, Seckler R (2014) Bacteriophage P22 tailspike: structure of the complete protein and function of the interdomain linker. Acta Cryst D70:1336–1345 Spinelli S, Desmyter A, Verrips CT, de Haard HJ, Moineau S (2006) Cambillau C (2006) Lactococcal bacteriophage p2 receptor-binding protein structure suggests a common ancestor gene with bacterial and mammalian viruses. Nat Struct Mol Biol 13:85–89 Stockley PG, White SJ, Dykeman E, Manfield I, Rolfsson O, Patel N, Bingham R, Barker A, Wroblewski E, Chandler-Bostock R, Weiss EU, Ranson NA, Tuma R, Twarock R (2016) Bacteriophage MS2 genomic RNA encodes an assembly instruction manual for its capsid. Bacteriophage 6:e1157666 Su S, Gao YG, Zhang H, Terwilliger TC, Wang AH (1997) Analyses of the stability and function of three surface mutants (R82C, K69H, and L32R) of the gene V protein from Ff phage by X-ray crystallography. Protein Sci 6:771–780 Suhanovsky MM, Teschke CM (2015) Nature’s favorite building block: deciphering folding and capsid assembly of proteins with the HK97-fold. Virology 479-480:487–497 Sun L, Young LN, Zhang X, Boudko SP, Fokine A, Zbornik E, Roznowski AP, Molineux IJ, Rossmann MG, Bentley A, Fane BA (2014) Icosahedral bacteriophage ϕX174 forms a tail for DNA transport during infection. Nature 505:432–435

90

M. Sanz-Gaitero et al.

Sun Z, El Omari K, Sun X, Ilca SL, Kotecha A, Stuart DI, Poranen MM, Huiskonen JT (2017) Double-stranded RNA virus outer shell assembly by bona fide domain-swapping. Nat Commun 8:14814 Takata T, Haase-Pettingell C, King J (2012) The C-terminal cysteine annulus participates in autochaperone function for Salmonella phage P22 tailspike folding and assembly. Bacteriophage 2:36–49 Tang J, Olson N, Jardine PJ, Grimes S, Anderson DL, Baker TS (2008) DNA poised for release in bacteriophage ϕ29. Structure 16:935–943 Tang J, Lander GC, Olia A, Li R, Casjens S, Prevelige P, Cingolani G, Baker TS, Johnson JE (2011) Peering down the barrel of a bacteriophage portal: the genome packaging and release valve in P22. Structure 19:496–502 Tao P, Mahalingam M, Zhu J, Moayeri M, Sha J, Lawrence WS, Leppla SH, Chopra AK, Rao VB (2018) A bacteriophage T4 nanoparticle-based dual vaccine against anthrax and plague. MBio 9:e01926-18 Tavares P, Zinn-Justin S, Orlova EV (2012) Genome gating in tailed bacteriophage capsids. Adv Exp Med Biol 726:585–600 Taylor GL (2010) Introduction to phasing. Acta Cryst D66:325–338 Taylor NM, Prokhorov NS, Guerrero-Ferreira RC, Shneider MM, Browning C, Goldie KN, Stahlberg H, Leiman PG (2016) Structure of the T4 baseplate and its function in triggering sheath contraction. Nature 533:346–352 Taylor NMI, van Raaij MJ, Leiman PG (2018) Contractile injection systems of bacteriophages and related systems. Mol Microbiol 108:6–15 Thiriot DS, Nevzorov AA, Zagyanskiy L, Wu CH, Opella SJ (2004) Structure of the coat protein in Pf1 bacteriophage determined by solid-state NMR spectroscopy. J Mol Biol 341:869–879 Thomassen E, Gielen G, Schütz M, Schoehn G, Abrahams JP, Miller S, van Raaij MJ (2003) The structure of the receptor-binding domain of the bacteriophage T4 short tail fibre reveals a knitted trimeric metal-binding fold. J Mol Biol 331:361–373 Tremblay DM, Tegoni M, Spinelli S, Campanacci V, Blangy S, Huyghe C, Desmyter A, Labrie S, Moineau S, Cambillau C (2006) Receptor-binding protein of Lactococcus lactis phages: identification and characterization of the saccharide receptor-binding site. J Bacteriol 188:2400–2410 van Raaij MJ, Schoehn G, Burda MR, Miller S (2001) Crystal structure of a heat and protease-stable part of the bacteriophage T4 short tail fibre. J Mol Biol 314:1137–1146 Veesler D, Robin G, Lichière J, Auzat I, Tavares P, Bron P, Campanacci V, Cambillau C (2010) Crystal structure of bacteriophage SPP1 distal tail protein (gp19.1): a baseplate hub paradigm in gram-positive infecting phages. J Biol Chem 285:36666–36673 Veesler D, Quispe J, Grigorieff N, Potter CS, Carragher B, Johnson JE (2012a) Maturation in action: CryoEM study of a viral capsid caught during expansion. Structure 20:1384–1390 Veesler D, Spinelli S, Mahony J, Lichiere J, Blangy S, Bricogne G, Legrand P, Ortiz- Lombardia M, Campanacci V, van Sinderen D, Cambillau C (2012b) Structure of the phage TP901-1 1.8 MDa baseplate suggests an alternative host adhesion mechanism. Proc Natl Acad Sci U S A 109:8954–8958 Vinga I, Baptista C, Auzat I, Petipas I, Lurz R, Tavares P, Santos MA, São-José C (2012) Role of bacteriophage SPP1 tail spike protein gp21 on host cell receptor binding and trigger of phage DNA ejection. Mol Microbiol 83:289–303 Walter M, Fiedler C, Grassl R, Biebl M, Rachel R, Hermo-Parrado XL, Llamas-Saiz AL, Seckler R, Miller S, van Raaij MJ (2008) Structure of the receptor-binding protein of bacteriophage Det7: a podoviral tail spike in a myovirus. J Virol 82:2265–2273 Wang YA, Yu X, Overman S, Tsuboi M, Thomas GJ, Egelman EH (2006) The structure of a filamentous bacteriophage. J Mol Biol 361:209–215 Xiang Y, Rossmann MG (2011) Structure of bacteriophage ϕ29 head fibers has a supercoiled triple repeating helix-turn-helix motif. Proc Natl Acad Sci U S A 108:4806–4810

Structure and Function of Bacteriophages

91

Xiang Y, Morais MC, Cohen DN, Bowman VD, Anderson DL, Rossmann MG (2008) Crystal and cryoEM structural studies of a cell wall degrading enzyme in the bacteriophage ϕ29 tail. Proc Natl Acad Sci U S A 105:9552–9557 Xiang Y, Leiman PG, Li L, Grimes S, Anderson DL, Rossmann MG (2009) Crystallographic insights into the autocatalytic assembly mechanism of a bacteriophage tail spike. Mol Cell 34:375–386 Xu L, Benson SD, Butcher SJ, Bamford DH, Burnett RM (2003) The receptor binding protein P2 of PRD1, a virus targeting antibiotic-resistant bacteria, has a novel fold suggesting multiple functions. Structure 11:309–322 Xu J, Hendrix RW, Duda RL (2014) Chaperone-protein interactions that mediate assembly of the bacteriophage λ tail to the correct length. J Mol Biol 426:1004–1018 Xu J, Gui M, Wang D, Xiang Y (2016) The bacteriophage ϕ29 tail possesses a pore-forming loop for cell membrane penetration. Nature 534:544–547 Yap ML, Rossmann MG (2014) Structure and function of bacteriophage T4. Future Microbiol 9:1319–1327 Yutin N, Backstrom D, Ettema TJG, Krupovic M, Koonin EV (2018) Vast diversity of prokaryotic virus genomes encoding double jelly-roll major capsid proteins uncovered by genomic and metagenomic sequence analysis. Virol J 15:67 Zheng W, Wang F, Taylor NMI, Guerrero-Ferreira RC, Leiman PG, Egelman EH (2017) Refined cryo-EM structure of the T4 tail tube: exploring the lowest dose limit. Structure 25:1436–1441 Zimmer SG, Millette RL (1975) DNA-dependent RNA polymerase from Pseudomonas BAL-31. II. Transcription of the allomorphic forms of bacteriophage PM2 DNA. Biochemistry 14:300–307

Adsorption: Phage Acquisition of Bacteria John J. Dennehy and Stephen T. Abedon

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Free Virions to Virocells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Movement of Free Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Virion Diffusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Non-diffusive Movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Encounter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Attachment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adsorption Factors and Reversible Attachment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Irreversible Attachment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adsorption Kinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Importance of Phage Titers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adsorption Rate Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . How Fast Adsorption Is Fast Enough? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adsorption Rate Constant Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genome Translocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

94 94 100 101 102 103 103 104 104 105 107 107 108 110 110 111 113 113 114

Abstract

To infect, a virion must first encounter a bacterium, where “encounter” essentially is a euphemism for “collision.” That is, a virion must physically touch a target bacterium to enable attachment. Indeed, phage acquisition of a bacterium to infect typically will consist of a combination of virion movement (including but not J. J. Dennehy Biology Department, Queens College and The Graduate Center of the City University of New York, New York, NY, USA e-mail: [email protected] S. T. Abedon (*) Department of Microbiology, The Ohio State University, Mansfield, OH, USA © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_2

93

94

J. J. Dennehy and S. T. Abedon

exclusively via diffusion), virion encounter with a target bacterium, virion reversible attachment to that bacterium, subsequent irreversible attachment, and then phage-genome translocation into a bacterium’s cytoplasm. Many of these concepts often are lumped under the heading of phage adsorption, and in this chapter we review these various aspects of phage adsorption/acquisition of bacteria. As companion chapters covering what happens either following or because of phage adsorption, see also chapters ▶ “Phage Infection and Lysis,” ▶ “Bacteriophage Ecology,” and ▶ “Bacteriophage Pharmacology and Immunology.”

Introduction Successful phage therapy is largely dependent on the ability of phages to infect bacteria (chapters ▶ “Phage Infection and Lysis” and ▶ “Bacteriophage Pharmacology and Immunology”). The first step of phage acquisition of a bacteria cell is typically described as adsorption. Phage adsorption, however, may be considered to consist of some combination of virion diffusion, virion encounter with a bacterium, virion extracellular attachment to a potential host cell, and then genome translocation from the virion’s capsid into that cell’s cytoplasm. The latter, at least arguably, is also an aspect of a phage’s infection of a cell rather than strictly or solely an aspect of phage adsorption. Independent of semantics, all of these steps nevertheless are involved in phage acquisition of a bacterium to infect. Together, especially as leading up to the translocation step, these various processes collectively also can be described as aspects of a phage’s extracellular “search” for new bacteria to infect. In this chapter we describe these various steps of “adsorption,” including virion movement, encounter, virion attachment to bacterial surfaces, the kinetics of phage adsorption to bacteria, and then translocation of the phage genome into the bacterial cytoplasm; the latter we also refer to here as uptake. The various steps of phage adsorption as leading into the start of phage infections are summarized in Fig. 1. Definitions of key terms can be found in Table 1.

Free Virions to Virocells There are two steps that are key to successful phage therapy: virion acquisition of bacteria and subsequent phage killing of bacteria. Temporally, the first of these steps is virion acquisition of bacteria, which involves a combination of virion movement, virion encounter with bacteria, and subsequent virion transition from a free phage state to instead an infected cell (Fig. 1). Forterre (2013) contrasts these states as virion vs. “virocell.” Thus, from p. 233 of Forterre, “The virocell. . ., being a cellular organism, corresponds to the ‘living form’ of the virus, whereas virions are in fact the equivalent of seeds or spores for multicellular organisms.” Thus virions, like seeds or spores, are charged with combinations of movement, survival, and initiation of the vegetative stage of the life cycle.

Adsorption: Phage Acquisition of Bacteria

95

Fig. 1 Depiction of the steps of phage adsorption, as ending upon the start of phage-genome translocation. Phage adsorption technically is the irreversible attachment of free virions to host cell surfaces. Adsorption nevertheless is often described in terms of some combination of virion diffusion, encounter between phage virion and bacterium (i.e., collision), some degree of reversible binding which likely involves virion motion across the surface of the bacterium (reversible attachment), a transition to irreversible attachment of the virion to the bacterial surface (which is adsorption proper), and then phage-genome translocation into the bacterial cytoplasm (i.e., genome uptake, which may or may not be strictly included as a component of phage adsorption but gives rise to a virocell)

Notwithstanding Forterre’s perspective, we suggest here that rather than strictly “virions” it in fact is free virions, or free phages, that contrast with virocells (Delbrück 1946), i.e., as virions also can exist within cells, i.e., within virocells, prior to their release. Many authors nevertheless use the term “infection” to describe virion adsorption or attachment. Their assumption, presumably, is that once attachment has occurred then so too has infection. For the sake of not being ambiguous, however, this practice should be avoided. That is, when the terms “infection” and “adsorption” or “attachment” are used interchangeably, then it sometimes can be difficult for readers to appreciate which of these aspects of the phage life cycle are being referred to. Specifically, phage acquisition of a bacterium is an aspect of free virions, while phage infection of a bacterium is an aspect of virocells (chapter ▶ “Phage Infection and Lysis”). The likelihood that free phages may become virocells is a function of phage virion per capita adsorption rates in combination with both phage and bacterial concentrations. Specifically, the higher phage per capita adsorption rates, then the faster phage virions acquire target bacteria and thereby also the faster target bacteria are phage-acquired. Note, though, that these are not identical concepts – rates of adsorption of free phages, that is, vs. adsorption rates of bacteria by phages – but

96

J. J. Dennehy and S. T. Abedon

Table 1 Definitions of Key Terms Adsorption

Adsorption capacity

Adsorption cofactor

Adsorption factor

Adsorption rate

Adsorption rate constant

Attachment

Capsid

Especially irreversible attachment of a virion to a host cell. Many authors use the term “infection” when “adsorption” or “attachment” would be more appropriate, however. For example, “multiplicity of infection” for many phages more precisely should be described as a multiplicity of adsorption or instead a “multiplicity of attachment” Number of phages capable of irreversible attachment to the surface of an individual bacterial cell before, presumably for reasons of crowding, no additional virions can attach Term used in phage biology to describe especially small substances, such as monovalent cations, divalent cations, or the amino acid tryptophan (free tryptophan), that directly interact with virions to result in an adsorption-capable virion state Chemical or physical aspect of environments, such as pH, osmolarity, or temperature, that impact phage virion adsorption rates, i.e., aspects of environments that can reversibly impact especially phage attachment abilities but which traditionally have not been described in phage biology as adsorption cofactors. We thus use the word “factor” here as a broader concept than that of “cofactor,” where cofactors from enzymology are limited to smaller molecules, such as vitamins (in animal physiology) that interact with and provide to proteins chemistries that are other than those associated with amino acid moieties. Thus, pH is an adsorption factor but not an adsorption cofactor, while free tryptophan as an adsorption cofactor would also be included here under the heading of adsorption factor Measure of the duration between the beginning of a phage virion’s extracellular existence and the point of irreversible attachment to especially a host bacterium. This rate is largely a function of a combination of host density and phage adsorption rate constant Measure of the rate of virion acquisition of a host bacterium in terms of probabilities of adsorption per unit time. Adsorption rate constants tend to vary as depending especially on virion diffusion rates, host target size, and virion affinity for host bacterium cell surfaces Specific virion interactions with the surface of a host cell, thus allowing initiation of the infection process. Attachment can be distinguished into reversible attachment and irreversible attachment stages. Reversible attachment is followed either by irreversible attachment or instead by virion diffusion away from the bacterial surface, i.e., disassociation. Irreversible attachment is followed instead by phage-genome uptake into the bacterial cytoplasm Especially that aspect of a virion particle that surrounds (i.e., encapsidates) the virus genome, e.g., the head in a tailed phage, whereas the tail can be viewed instead as an accessory or appendage to the capsid (continued)

Adsorption: Phage Acquisition of Bacteria

97

Table 1 (continued) Chaperonin Contractile tail

Encounter Extracellular search

Free phage

Free virion

Generation time

Growth parameter

Hershey and Chase experiment

Irreversible adsorption/ irreversible attachment

Protein that aids in the folding of other proteins, that is, which “chaparones” another protein’s folding Virion adsorption-mediating appendage that, upon irreversible attachment, contracts in length, thereby forcing an interior tail tube through the bacterial cell envelope to make contact with the bacterial plasma membrane. Contractile tails are associated with members of phage family Myoviridae Collision of a virion with a cell in the course of a virion’s extracellular search, as potentially resulting in attachment Virion movement through the extracellular environment, which ideally for the phage is toward virion attachment to a host bacterium. The search most notably involves virion diffusion but also can involve virion movement by other means, e.g., such as hitchhiking on animals, and also involves recognition of host bacteria cell surface molecules following encounter. The extracellular search thus begins with virion release, leads to encounter, and ends at the point of irreversible virion attachment. Though diffusion is a passive process, other aspects can be more active including reversible phage binding to animal mucus as well as the specific interactions with bacteria associated with attachment Mature phage virion that is found in the extracellular environment and has not yet irreversibly attached to a host bacterium. The concept of free phage is equivalent to that of free virion As equivalent to free phage for bacterial viruses, this is the phage entity involved in the virion extracellular search and also the entity that contrasts with virocells Measure of the interval between “birth” and production of the next generation of progeny. Generally the shorter an organism’s generation time, all else held constant (including no trade-offs), then the greater an organism’s population growth rate Quantifiable aspects of virus life cycles including the adsorption rate constant, latent period length, and burst size, but also virion inactivation rates Method of separating phage genomes from phage virions post-irreversible adsorption that involves rapid mechanical disruption using a “Waring” blender. By labeling DNA and protein with different radioactive elements, the role of DNA as the hereditary material in tested phages was confirmed Specific interactions between free virions and the surfaces of bacterial cells that involve permanent changes in virion morphology and/or which are not disruptable without damaging either the virion or the associated bacterium (see Hershey and Chase experiment). At the point of irreversible attachment, the extracellular search has been completed, and genome uptake will commence (continued)

98

J. J. Dennehy and S. T. Abedon

Table 1 (continued) MOI

Multiplicity

Noncontractile tail

Nucleocapsid

Phage receptor

Phage virion

Receptor

Reversible adsorption/ reversible attachment

Spike protein

Superinfection exclusion

Abbreviation of multiplicity of infection, used here synonymously with “multiplicity of adsorption.” The concept of multiplicity of infection, that is, was developed prior to an understanding that phage infection does not necessarily always follow phage adsorption, e.g., a MOI of 5 does not always mean that the genomes of five phages have necessarily all entered the same bacterium, i.e., as due to superinfection exclusion. See also Multiplicity Short for multiplicity of infection, here referring to the ratio of virions that have attached, i.e., adsorbed to phage-susceptible bacteria, to the number of those bacteria. Typically multiplicity of infection is abbreviated as MOI. Thus, a multiplicity or MOI of 0.1 would mean that one phage has adsorbed for every ten adsorbable bacteria present. Note that a less rigorous and less useful though still commonly seen definition of multiplicity is the ratio of phages added to adsorbable bacteria, and not all authors distinguish between these two meanings of multiplicity, as can be described as MOIinput vs. MOIactual Prominent adsorption-effecting virion appendages that do not shorten in the course of phage adsorption. Noncontractile tails are associated with members of phage families Podoviridae and Siphoviridae Combination of a virus capsid and the nucleic acid it contains. In members of phage family Cystoviridae, the nucleocapsid is found within a lipid envelope and enters the bacterium intact in the course of genome uptake One or more types of specific molecules found on the surface of bacteria with which phage virions interact in the course of adsorption and attachment The virion of a phage, with the phrasing used here to explicitly distinguish from phage virocells, though phage virions prior to release can also be found within virocells. Thus, not all phage virions are necessarily also free virions Bacteria-encoded cell surface-exposed molecules to which phage capsid surface proteins specifically bind in order to adsorb to a host cell. These molecules and structures may include pili, transport and porin proteins, peptidoglycan, lipopolysaccharides, and teichoic acids Specific interactions between free virions and the surfaces of bacterial cells that do not involve permanent changes in virion morphology and which are disruptable without damaging either the virion or the associated bacterium. At the point of reversible attachment, the extracellular search has not yet been completed Adsorption virion appendage consisting of a single protein associated with the lipid of enveloped viruses. Spike proteins are displayed by the virions of phage family Cystoviridae Phage-encoded process of blockage of the uptake of the genomes of phage virions that have adsorbed to already phageinfected bacteria (rather than blocking their adsorption or blocking infection progress post genome uptake) (continued)

Adsorption: Phage Acquisition of Bacteria

99

Table 1 (continued) Tail

Tail fiber

Tail tube

Tape measure protein

Target bacterium Translocation

Turbulent flow

Uptake Virion

Virocell

Elaborate appendage attached to a virion’s capsid that is involved in attachment of a tailed phage to a bacterium’s surface along with transfer of DNA to the bacterium’s cytoplasm (the latter called translocation or uptake). Tails can be long and able to contract (contractile tails as seen with members of phage family, Myoviridae), long and noncontractile (phage family Siphoviridae), and quite short (phage family Podoviridae) Virion appendage that can be found associated with tails and that generally is thought to be involved in phage adsorption processes, beginning with encounter Virion transport organ that facilitates the delivery of the phage genome past the host cell wall and membrane into the cytoplasm proper Phage-encoded protein whose gene length is proportional to the length of Siphoviridae and Myoviridae tails. Tape measure proteins also participate in phage-genome translocation into the host cell cytoplasm A potential phage host, should virion encounter with that bacterium occur Movement of a phage’s genome, post attachment, from outside of a cell into a cell’s cytoplasm. The start of a phage infection arguably occurs in the course of such genome translocation. Genome translocation is synonymous with genome uptake Chaotic movement of fluids that results in effectively randomized motion but motion that is other than that directly associated with diffusion Here synonymous with translocation, i.e., of phage’s genome into a cell’s cytoplasm Non-metabolizing, encapsidated virus state. A mature bacteriophage virion found unattached and outside of a host cell is called a free phage Description of a cell that is undergoing virus infection, indicating that this state represents a metabolic “partnership” between infecting virus and infected cell. Contrast virocell especially with free phage, i.e., with free virions, as virocells can contain intracellular (non-free) virions. Contrast the virocell stage of virus life cycles also with the extracellular search

rather respectively refer to rates of loss of free phages to bacterial adsorption (free phage declines) and rates of loss of uninfected bacteria to phage adsorption (uninfected bacteria declines and/or increases in numbers of virocells). In other words, it is possible for free phages to decline in number rapidly without uninfected bacteria substantially declining in number (particularly when starting concentrations of phages are low and concentrations of bacteria are high, e.g., 102 vs. 108 per ml, respectively) as well as for uninfected bacteria to decline in number rapidly but not free phages (the latter particularly when starting phage concentrations are high, while starting bacterial concentrations instead are low, e.g., 108 vs. 102 per ml, also respectively). In any case, the faster the virions can acquire bacteria, then the more rapidly they can have an impact on those bacteria.

100

J. J. Dennehy and S. T. Abedon

Movement of Free Phages For a phage to be successful, the movement ultimately will bring a virion into association with a bacterium that can serve as a host. This movement for free virions is always random rather than directed but still can be differentiated into that which occurs via diffusion, that which occurs in the course of bulk environmental movement of air or water, that which occurs relative to surfaces (other than explicitly bacterial surfaces), and that which occurs as due to turbulent flow (Fig. 2). Diffusion will operate in conjunction with all of these other processes though can be a less significant component of virion motion to the extent that it is slow in comparison especially relative to bulk environmental movement or turbulent flow. In the case of bulk environmental movement of air or water, virions move in association with their immediate environment or microenvironment rather than

Fig. 2 Different types of movement. Thickness of lines is meant to indicate differences in speed, i.e., with diffusion potentially the slowest of these processes (and certainly the slowest when the other processes are significant contributors to phage extracellular searches). The double-headed arrows or simply double arrows indicate that bacteria may be moving as well, though bacterial movement solely by diffusion will tend to be slow in comparison with virion diffusion due to the larger size of bacterial cells (hence, bacterial diffusion is not indicated in the figure). Bacterial motility is indicated especially by relative movement, though diffusion and turbulent flow also represent forms of relative movement, that is, movement as observed within rather than in association with the bulk movement of environment or microenvironment. Note, though, that while relative movement is depicted as directly toward the free virion or bacterium, in fact it can be directly away or anywhere in between as well, e.g., as too can be movement due to diffusion and turbulent flow in any direction

Adsorption: Phage Acquisition of Bacteria

101

relative to that environment or microenvironment. Movement with environments, especially without turbulent flow, should result in little increase in the likelihood of phage encounter with bacteria, though virions involved nevertheless will be moving from one location to another. Virion diffusion, by contrast, occurs relative to the environments or microenvironments within which free virions are located and thus can serve to narrow the spatial gap between a phage virion and its target bacterium, as too can environmental movement relative to surfaces, e.g., water flowing over or through a bacterial biofilm. Turbulent flow may also narrow this gap. Virion movement – which of course is passive as phages do not have any means of propulsion – can occur relative to targeted bacteria by diffusion, can flow relative to surfaces, or, due to turbulent flow, can end with virion encounter with a bacterium to infect. Indeed, encounter with a bacterium is the same point at which free phages begin to transition to adsorbed phages. As these latter virion-bacterium interactions tend to be highly specific, we describe this overall process as an extracellular search or simply “search” (Fig. 2), and especially so for the sake of consistency with the ecological connotation of organisms “searching” for resources to acquire, even if that search does not always involve active movement on an organism’s part. The duration of virion movement is an important component of this search.

Virion Diffusion In terms of the characterization of phage life cycles, it is primarily virion diffusion that is considered when describing the phage extracellular search, though occasionally other forms of motion are formally considered as well (Koch 1960; Murray and Jackson 1992). Virion diffusion will occur given a lack of attachment to or entanglement with other materials. The faster virions diffuse, then the faster they will encounter host cells, all else held constant. Host cell encounter in turn will, depending on phage-to-cell affinity, potentially end a free phage’s extracellular search. Particle diffusion rates are inversely correlated with particle size. Smaller phage virions therefore are expected to display faster diffusion and also may be less likely to become nonspecifically entangled with environmental materials (Saltzman et al. 1994). Furthermore, shape affects diffusion. Experiments have shown that rod-shaped viruses diffuse in tissues or gels faster than spherical viruses, but the converse is true for fluids (Lee et al. 2013). The rate of virion diffusion thus may be under phage control in the sense that virion morphology is modifiable by natural selection. Equivalently, therapists may be able to employ phages for phage therapy whose virions display inherently different rates of diffusion, depending on phage choice, e.g., such as with virions varying in terms of overall size (such as jumbo phages vs. more normal-sized phages), morphology (such as myoviruses vs. podoviruses), tail length (such as very long siphovirus tails vs. the very short podovirus tails), or presence of appendages such as tail fibers. Diffusibility also potentially may be modified via phage virion biotechnological modification.

102

J. J. Dennehy and S. T. Abedon

Though most phages likely are tailed (order Caudovirales), taxonomically most phage families morphologically are tailless (chapter ▶ “Structure and Function of Bacteriophages”). Thus, while most phage families have either spherical or ovoid shapes, this is less true for members of phage order Caudovirales, with many of the classical laboratory phages, such as phage T4 and phage λ, including most phages used for phage therapy, having an icosahedral “head” and rod-like “tail.” Tail lengths may affect diffusibility and can be modified via mutations in tape measure proteins (Katsura 1987; Abuladze et al. 1994). The retraction of tail fibers while passing through unfavorable adsorption environments (below) could result as well in faster diffusion as may be evidenced by faster sedimentation rates during ultracentrifugation (Kanner and Kozloff 1964). Phage λ notably produces larger plaques when its tail fiber is mutationally lost (Hendrix and Duda 1992), which could stem from lower affinity of virions for host cells, though also could be a consequence of faster virion diffusion through agar (Gallet et al. 2011). Filamentous phage virion length and thereby virion size tracks with corresponds to genome length (Russel and Model 2006), with lengths in principle readily modified (Gailus et al. 1994; Rakonjac and Model 1998).

Non-diffusive Movement Lack of phage binding to nonhost materials can result in virion retention within fluids along with virion movement in association with ongoing currents. This latter, non-diffusive movement can increase the likelihood of phage encounter with a target cell since flow can move phages faster than diffusion alone, though this is relevant to phage adsorption only if virions are moving in such flow relative to rather than with target bacteria. If the movement consists of turbulent flow, such that microenvironments are well mixed, then that can result instead in faster virion movement relative to bacteria (Koch 1960). Though not fluid flow per se, nevertheless bacterial motility, such as seen in the course of bacterial flagella-mediated chemotaxis, can in principle increase the likelihood of phage encounter. For example, many bacteria possess large tail-like appendages called flagella, the rotation of which can propel the bacterium through a medium. Such movement, colloquially known as bacterial “swimming,” may increase the likelihood of phage-bacteria collisions relative to non- with these “swimming” vs. not swimming bacteria (Koch 1960). Murray and Jackson (1992), however, suggest that for phage-bacteria systems vs., e.g., virus interaction with eukaryotic algae, such swimming-mediated relative movement could be less relevant to virion encounter given the small size of bacteria relative to that of larger eukaryotic cells. Still, Koch argues for at least twofold increases in encounter rates due to bacterial taxis. Thus, we can conclude that there is at least a potential cost to bacteria that is associated with their displaying motility in that the greater the fluid volume that passes over their cells, then potentially the more phages that should pass over cells as well.

Adsorption: Phage Acquisition of Bacteria

103

In addition, bacterial motility also can contribute to phage movement as virocells. Specifically, phage infection of a motile bacterium will result in the infecting phage being moved to wherever the bacterium moves. Further, the longer that phages remain infecting, including as prophages, the potentially greater the extent of that movement (Igler and Abedon 2019). Upon eventual virion release from the virocell, a phage may be able to encounter new bacteria to infect that potentially may be less readily reached without such hitchhiking on bacterial motility (chapter ▶ “Bacteriophage Ecology”). Virions also may be transported over longer distances such as through the air (Reche et al. 2018), in association with dust and splashed water, or even move in association with animals serving as mechanical vectors (Sisler 1940; Dennehy et al. 2006). This bulk virion movement, however, is not necessarily relative to the positions of co-located bacteria. Thus, while it may contribute to phage biogeographic distributions (chapter ▶ “Bacteriophage Ecology”), it may not contribute to phage encounter with susceptible bacteria.

Encounter The likelihood of a virion’s extracellular search ending with successful bacterial acquisition, all else held constant, will be greater the larger the total target size for adsorption (Stent 1963). Thus, larger cells should be more vulnerable to phage adsorption than smaller cells as too should clumps of clonally related cells, i.e., cellular arrangements or microcolonies (Abedon 2012). The likelihood of attachment will also be greater given the presence simply of more target cells, which have the effect of collectively increasing overall target sizes for phage adsorption, i.e., by providing more cells for phages to target (as affecting target quantity). In addition, at least to a degree and not strictly an aspect of phage diffusion, the greater density or accessibility of phage receptor molecules found on the surface of target cells (as affecting target quality), then the greater the likelihood that virion encounter with a given cell will lead to virion-receptor interaction on a cell’s surface (Schwartz 1976). It is only once virion encounter has occurred that virion attachment and thereby free virion transition to a virocell can occur.

Attachment Phage attachment to the host occurs after collision between the phage virion and host cell, i.e., encounter. Overall, then, the phage extracellular search begins with virion release that is then followed by virion movement, encounter, and reversible attachment and then ends with irreversible attachment (Fig. 2). The terms attachment or adsorption if used without qualification, however, generally are being used synonymously with irreversible attachment.

104

J. J. Dennehy and S. T. Abedon

Overview While direct physical interaction is necessary to achieve attachment given encounter, encounter still is not sufficient for attachment to result. Instead, there must also be a chemically and physically suitable environment along with an appropriate virion morphological state as well as specific, complementary interactions between the virion attachment proteins and specific receptor molecules found on the target bacterium’s surface (Henning and Hashemolhosseini 1994; Bertin et al. 2011; Casjens and Molineux 2012; Chatterjee and Rothenberg 2012; Garcia-Doval and van Raaij 2013). As noted, however, the attachment of a phage to a host is thought to occur in two stages, corresponding to reversible adsorption and then irreversible adsorption (Storms and Sauvageau 2015). As implied by the name, the reversible first stage is transient, though also is poorly defined, involving weak chemical interactions between phage proteins and specific bacterial surface molecules, such as in Gram-negative bacteria, outer membrane proteins, or lipopolysaccharides. Reversible interactions are followed either by the dissociation of the phage virion from the host (i.e., loss of virion attachment) or, instead, the permanent, irreversible binding of the phage virion to the surface of the host cell (permanence of virion attachment). Irreversible attachment is followed by phage-genome uptake into the host cell cytoplasm, which initiates phage infection of a host bacterium. Note that loss of virion attachment following encounter with a host (dissociation) can be followed by virion encounter again with the same host bacterium or instead virion encounter with a different host. Thus, loss of virion attachment, unlike irreversible attachment, is not a permanent state.

Adsorption Factors and Reversible Attachment Reversible attachment is associated with interactions between a virion’s primary attachment proteins and the primary cell surface receptor but occurs without associated permanent changes in virion morphology. The classic examples of primary attachment proteins are the tips of phage tail fibers such as those associated with gene product (gp) 37 of phage T4 (Montag et al. 1990). These tail fiber tips reversibly interact with specific cell surface molecules, allowing the phage virion to “float” upon the surface of the encountered cell (Goldberg 1983). The phage virion’s constrained movement is thought to allow the phage to position itself properly for subsequent irreversible attachment to secondary attachment receptors also found on the cell surface (Hu et al. 2015). As mentioned previously, irreversible attachment will not necessarily occur with 100% efficiency per encounter (Schwartz 1975), especially if virion affinities for phage receptors are weak. The local adsorption environment can be crucial to determining this affinity such that if it is not chemically or physically suitable to phage adsorption, then attachment will not occur. For successful phage therapy, it is crucial that the environmental conditions used for phage isolation and characterization in the laboratory do not deviate too extremely from those seen in vivo or in situ,

Adsorption: Phage Acquisition of Bacteria

105

particularly to the point where phage affinity for target bacteria is substantially lessened. The main environmental factors affecting phage attachment are what solutes are present in what concentrations, the overall osmolarity, or salinity (Augustine et al. 2013), pH, and temperature (Conley and Wood 1975). Conley and Wood (1975) also made a distinction between adsorption cofactors and everything else impacting phage adsorption rates, and these we describe together as adsorption factors. Natural selection likely acts to enhance phage adsorption rates especially under conditions that are equivalent to where a phage is most likely to find itself in association with a target bacterium. An example is the common dependence by phages on the presence of specific chemical adsorption cofactors. A study of chemical cofactors affecting adsorption – e.g., calcium ions, sodium ions, free tryptophan, etc. (Storms et al. 2010) – has a surprisingly long history, and many of the results pertaining especially to phage T4 can be found in the publications of Conley and Wood (1975) and Kutter et al. (1994), though see also one of our Ph.D. dissertations (Abedon 1990). Arguably the most important of these adsorption cofactors are metal ions. These, depending on the phage, are either monovalent cations, e.g., sodium and potassium ions, or instead divalent cations such as calcium or magnesium ions. For example, phage T4 adsorption is maximized under conditions that best replicate that of their natural habitat, the colonic lumen. These conditions include the presence of optimal monovalent cation concentrations, free tryptophan (which otherwise is rare within environments), roughly neutral pH, and body temperature. These requirements potentially provide a means by which T4 phages can distinguish among environments and attach to bacteria mostly under conditions resembling those of the colonic lumen rather than as associated with, for example, extracolonic freshwaters (Abedon 1990). It is important to be aware of adsorption factors when studying or employing phages within environments that differ chemically or physically from those environments from which phages are isolated or had been propagating. That is, the absence of proper adsorption conditions in situ will lower the probability of successful phage attachment to target bacteria. Phages applied at 4
C may be less effective at attaching to target bacteria if those phages normally infect at 37
C, for example. Phages that can propagate in a non-body environment similarly may not necessarily be bathed in needed adsorption cofactors within bodies. How often these possibilities are actually issues for phage therapy is unknown but nevertheless is worth considering.

Irreversible Attachment Secondary receptors found on bacterial surfaces interact with phage secondary attachment proteins. Binding of the secondary attachment proteins to the secondary receptor is much stronger than the binding of the primary attachment protein to the primary receptor and so constitutes an irreversible attachment rather than a reversible attachment. For example, phage T4 initiates infection of E. coli by reversibly binding

106

J. J. Dennehy and S. T. Abedon

to primary receptor molecules OmpC porins and lipopolysaccharides (LPS) on the cell surface with its long tail fibers. Subsequently, T4 virions undergo associated changes in their virion baseplate conformation, and six short tail fibers (gp12) are exposed and irreversibly bind to secondary receptor, which also are LPS molecules (van Raaij et al. 2001; Hu et al. 2015) (chapter ▶ “Structure and Function of Bacteriophages”). Because of the baseplate changes and strong short tail fiber binding, these secondary interactions are not reversible. At this point of irreversible attachment, the phage therefore is committed to infecting an encountered bacterium, whereas prior to this point it is possible for the virion to still dissociate from an encountered cell. This overall process for phage T4 was reviewed in detail by Goldberg (1983). The probability that reversible attachment occurs, or is sustained for sufficiently long periods, is a function of binding affinities between phage primary attachment proteins and host receptor molecules along with, to a degree, the densities of those molecules. The probability that irreversible attachment occurs is a function of how long reversible attachment is allowed to persist in combination with phage binding affinities to secondary receptor molecules as well as total numbers and/or cell surface densities of secondary receptors. Once irreversible attachment has been initiated, the virion transitions into the process of translocating its genome into the bacterial cytoplasm (genome uptake). Regardless of whether phage-genome translocation is successful, the virion is considered to be removed from the extracellular environment at the point of phage irreversible adsorption to a bacterium, i.e., a free phage has been converted into an attached phage. The classic Hershey and Chase (1952) experiment involved a partial mechanical reversal of the otherwise irreversible attachment step. Phage T2 virions were labeled with either radioactive sulfur (35S) or radioactive phosphorus (32P), where these elements are found only in virion proteins or virion DNA, respectively. Using what is described by Hershey and Chase as a “Waring blender (semimicro size),” the portion of the virion that is found outside of the bacterium is sheared off and then separated from the now phage-infected bacterium. Hershey and Chase found that it was the protein and therefore 35S-labeled virion protein that was lost from what otherwise were still successfully phage-infected bacteria. The 32P-labeled DNA, by contrast, co-sedimented with the infected bacteria upon subsequent centrifugation. Phage adsorption thus results in a difficult-to-disrupt interaction between virion proteins and cell surfaces, i.e., interactions requiring the action of high shear forces (in a blender) to disrupt, and that binding is associated with virion genome delivery to the bacterial cytoplasm. Thus, as shown in these experiments, once irreversible attachment has been completed, only the phage T2 nucleic acid genome is required for successful phage infection, as is consistent with the attachment step resulting as well in DNA uptake (chapter ▶ “Bacteriophage Use in Molecular Biology and Biotechnology”). Historically, this finding also was indicative of the relatively small contribution of virion protein vs. virion DNA to phage intracellular development and therefore was viewed as strongly supporting the then still not fully accepted notion that DNA serves as the hereditary material of organisms generally.

Adsorption: Phage Acquisition of Bacteria

107

Adsorption Kinetics Phage adsorption kinetics describe the rapidity of virion adsorption processes. These kinetics, at least from a phage’s perspective, are generally considered to be a function of a combination of the environmental densities of target bacteria along with the phage growth parameter known as the adsorption rate constant. In this subsection we consider the latter, that is, the phage adsorption rate constant, which is an important determinant of both phage population growth rates and bacterial susceptibility to phages, but we also consider the impacts of differences in bacterial or phage concentrations on phage adsorption rates. The phage adsorption rate constant defines a phage’s intrinsic, per capita, adsorption rate to a specific bacterial type as observed under a given set of environmental conditions. The density of target bacteria, however, is similarly important in determining the rate at which individual phage virions are able to acquire susceptible bacteria to infect, i.e., toward phage population growth. For phage therapy, though, ultimately it is what phage titers can achieve in situ, in combination with phage adsorption rate constants, that determine the rate that bacterial populations can be brought under control.

Importance of Phage Titers The rate that free phages are lost to adsorption is a function of a combination of bacterial concentrations and the phage adsorption rate constant. The rate that phages are lost to adsorption, however, is less important for phage therapy than the rate that phage-uninfected bacteria are lost to phage adsorption, and the two concepts are not identical. There is, in other words, a difference between rates of loss of free virions and rates of gains of virocells. This is seen both in relative terms – fraction of free virions lost vs. fraction of virocells gained (next paragraph) – and because not all free virion losses to adsorption are associated with virocell gains. The latter, for example, is seen when more than one virion adsorbs to a single bacterium, e.g., such that a loss of two virions would be associated with the gain of only a single virocell. Regarding relative changes, consider this thought experiment: The presence of a single phage adsorbing even infinitely fast, i.e., adsorption rate constant ¼ 1, at best will result in the adsorption of only a single bacterium. Therefore, the rate that all targeted bacteria are adsorbed on average in this example would approach zero despite the infinite rapidity of the individual phage adsorption process. This occurs because though the rate that phages adsorb is a function of their adsorption rate constant, along with the concentration of bacteria present, the rate that bacteria become phage adsorbed is a function instead of phage titer, also in combination with the phage adsorption rate constant. Thus, in this thought experiment, phage titer was as low as it could possibly be without being zero, so therefore rates that bacteria were adsorbed were also as low as they could possibly be, also without being zero. These ideas point to the importance of phage titers in phage therapy toward assuring that bacteria are acquired by phages at desirable rates (Abedon 2016, 2020;

108

J. J. Dennehy and S. T. Abedon

Danis-Wlodarczyk et al. 2020). Thus, though it is true that the overall rate that bacteria are adsorbed by phages will go up by increasing either concentrations of bacteria or concentrations of phages, only the latter, i.e., greater phage titers, will have the effect of increasing the rate that bacteria are adsorbed. Indeed, holding numbers of phages constant, the more targeted bacteria that are present within an environment, then the less rapidly on average will those bacteria be acquired by phages (since fewer phages will be present relative to numbers of bacteria present). At the same time, however, the more phages that are present, then the faster those bacteria will be adsorbed. Indeed, if numbers of bacteria are sufficiently low that we can ignore reductions in phage titers due to their adsorption to bacteria, e.g., such as when treating foods with phages where ideally few target bacteria will be expected to be present (chapter ▶ “Food Safety”), then the concentration of bacteria can simply be ignored when calculating rates of losses of uninfected bacteria to phage adsorption. That is, the rate of bacterial losses during such treatments can be approximated simply as a product of phage titer and the phage adsorption rate constant (Abedon 2011, 2014).

Adsorption Rate Constants The phage adsorption rate constant is a measure of a combination of rates of phage virion diffusion, bacterial target sizes, and the likelihood of phage irreversible attachment given virion encounter with a bacterium. Experimentally, adsorption rate constants are measured – holding bacterial densities constant – in terms of either rates of loss of phage virions from environments or instead rates of gain of phageinfected bacteria (Hyman and Abedon 2009). Conceptually, the phage adsorption rate constant is a description of the likelihood of adsorption of a single phage virion to a single, isolated bacterium, with that bacterium found within a given volume of medium and as measured over a given duration of time. Thus, for example, units can be per mL per min, which is often abbreviated as mL min1 or instead as mL1 min1. The faster that phage virions diffuse, the larger bacterial targets, or the greater the likelihood of phage irreversible adsorption given virion encounter with a bacterium, then the greater the phage adsorption rate constant. The limit for the rate of phage adsorption to a given target bacterium will be seen when phage affinity for primary phage receptors is sufficiently high that irreversible adsorption occurs with a probability of 1 given phage encounter with a bacterium. Note, though, that adsorption rate constants, if irreversible attachment likelihood given encounter were to equal 1, would still not come to equal infinity, per the previous thought experiments, as virion diffusion rates and/or bacterial target sizes would not also be infinite. That is, the phage adsorption rate constant is a function of more than just the likelihood of virion irreversible attachment given encounter. As noted, actual adsorption rates are dependent not just on the adsorption rate constant but also on bacterial densities. In fact, these two parameters have similar impacts on phage adsorption rates (Shao and Wang 2008), and both display saturation kinetics in terms of rates of phage adsorption. Thus, the more bacteria that are

Adsorption: Phage Acquisition of Bacteria

109

present or the greater the rate that a given phage will find and then adsorb a given bacterium, then the shorter the durations of a phage’s extracellular search. Nevertheless, as products of bacterial density and phage adsorption rate constants become sufficiently large, durations of extracellular searches will be reduced to near zero. This means that there are limits to the degree that adsorption processes may be further speeded up, such as may be achieved by presenting to phages with ever greater concentrations of bacteria. In other words, increasing phage adsorption rates by increasing phage adsorption rate constants can saturate and particularly will do so to the extent that bacteria are highly prevalent such that a phage which fails to attach upon encounter with a bacterium may quickly encounter another bacterium or even the same bacterium (Abedon 2009). As a thought experiment in this case, imagine a bacterial culture so dense that individual bacteria are all immediately adjacent, i.e., touching. In this case, upon a virion’s release, its encounter with a new bacterium would be essentially instantaneous and thus have reached a maximum. So too, were phage titers throughout an environment so high that all virions were touching, then bacteria would encounter phages instantaneously. Unlike the rates that free phage virions find themselves being lost to adsorption as bacterial densities increase, however, the rates that bacteria will receive adsorbing phages will increase as a linear function of phage densities. In other words, the greater phage titers, then the faster bacteria will be adsorbed as long as free phages do not interfere with each other. Note that in terms of interference, the number adsorbed virions that can fit on a bacterial surface, known as adsorption capacity (Stent 1963), and can be in the 100s. On the other hand, as generally it only requires one phage to infect a single bacterium, at some point adding yet more phages may become unnecessary. This scenario may occur as phage multiplicities come to exceed ten (Abedon 2018) and when phage titers are high, e.g., 109 ml1 for phages with greater adsorption rate constants (such as around 109 mL1 min1) and perhaps 1010 ml1 for more slowly adsorbing phages (such as around 1010 mL1 min1) (Danis-Wlodarczyk et al. 2020). There are at least three additional issues relevant to adsorption rate constants. These are (i) that adsorption rate constant determinations assume that environments are homogeneous, i.e., that phage access to all bacteria is equivalent; (ii) that adsorption rate constants ideally are determined in environments, in vitro, whose properties are equivalent to those found in situ; and (iii) that all phage targets for adsorption or interaction are targeted bacteria, e.g., rather than also benign commensal bacteria or other aspects of environments (Trubl et al. 2020) such as mucus (Barr et al. 2013). If a fraction of bacteria found outside of the laboratory are more difficult for phages to reach, if laboratory determinations of phage adsorption rate constants represent overestimations, or if more adsorption targets are present than those of target bacteria, then more phages may need to be supplied to optimize rates that targeted bacteria are acquired by phages. Alternatively, however, the phage potential to reproduce in situ to higher titers in many cases may compensate for insufficiencies in phage dosing, and thereby compensate for insufficient rates that target bacteria become phage adsorbed (chapter ▶ “Bacteriophage Pharmacology and Immunology”).

110

J. J. Dennehy and S. T. Abedon

How Fast Adsorption Is Fast Enough? The rates that bacteria are adsorbed by phages are expected to increase as a linear function of both phage titers and a phage’s adsorption rate constant. Increasing either value tenfold therefore will have the effect of reducing the time until a bacterium is adsorbed also by tenfold: If it takes 100 min for a bacterium to be adsorbed at a given phage density, then at a tenfold higher phage density it will take 10 min. Note, though, that we can expect diminishing returns. Thus, while going from 100 min to 10 min may be meaningful, going from 10 min to 1 min, i.e., by providing yet tenfold higher phage titers, may be less meaningful, and going from 1 min to 0.1 min (6 s) likely will have even less meaning. Except for phages with extremely low affinities for target bacteria, e.g., 1010 mL1 min1, increasing phage titers from, for example, 109 to 1010 phages/mL will not necessarily result in meaningfully faster bacterial adsorption by phages. We can calculate the half-life of a bacterium in the face of phage adsorption of ln(2)/(kP), where k is the phage adsorption rate constant and P is the phage titer (Abedon 2017). For example, if we set k to 2.5  109 mL min1 (Stent 1963), then the decrease in mean free time when we increase the phage titers from 109 to 1010 phages/mL is 15 s, i.e., 17 s for 109 phages/mL vs. about 2 s for 1010 phages/mL. By contrast, increasing from 107 to 108 phages/mL, the bacterial half-life instead decreases from 30 min to 3 min. That is, with 107 phages/mL, we have expectation that after a half an hour of adsorption, only about half of the bacteria will have been phage infected. Generally, therefore, a phage density in the range of 108 per ml, whether achieved through phage dosing alone or reached instead via in situ phage replication, should be viewed as representing an approximately minimum titer to assure rapid virion adsorption (Hagens and Loessner 2010; Abedon 2014, 2018), i.e. phage adsorption to most host bacteria within a few minutes. This estimation is not valid, however, if phages display poor intrinsic adsorption rates or if phages otherwise are slow to reach target bacteria due to diffusion barriers, in which case even higher phage numbers may be necessary. As noted, such phage titers will be needed whether supplied explicitly by dosing or instead generated in situ via phage population growth (chapter ▶ “Bacteriophage Pharmacology and Immunology”).

Adsorption Rate Constant Determination Virion adsorption rate determinations involve mixing phages with target bacteria and then in some manner distinguishing between those phages that remain unadsorbed (i.e., which are “free”) from those that have adsorbed. An important goal of such determinations is the measurement of phage adsorption rate constants, which are a function of the slope associated with the exponential decline in free phages following their mixture with adsorbable bacteria. Such a slope, realistically, can only be determined by measuring numbers of phages, during their adsorption, at least at three separate time points. That is, end point determinations of phage adsorption

Adsorption: Phage Acquisition of Bacteria

111

rates, as based on only two time points, corresponding to start and finish of determinations, at best are only semiquantitative. In any case, distinguishing free phages from adsorbed phages can be accomplished by removing or killing bacteria such by applying chloroform or instead via centrifugation, at which point remaining free phages are enumerated. Equivalently, free phages may be removed or inactivated by employing virucides, such as antiphage serum. Free virions can be separated from adsorbed virions via centrifugation such as using a microfuge. Pelleted phage-infected bacteria, referred to as infective centers, are then enumerated for phage viability, or instead free phages can be enumerated from supernatant. In addition, it is possible as well to employ conditionally lethal phage mutants toward determining rates of phage loss from environments, which can be inactivated following adsorption to target bacteria but which are viable while infecting other, permissive bacteria under permissive conditions; in this case free phages and adsorbed phages are separated by the inactivation of the latter upon irreversible attachment. Adsorption rate constant determination is discussed further in Hyman and Abedon (2009).

Genome Translocation Bacterial surface structures – glycocalyx, S-layers, peptidoglycan cell walls, and lipid membranes – can present physical barriers to bacteriophage attachment or infection. These barriers must be surmounted for phages to translocate their genomes from their capsids into the bacterial cytoplasm. In general, the processes of phagegenome translocation are not well understood, even for some of the best studied phages. For most phages, phage nucleic acid nevertheless is thought to be introduced into the host cytoplasm via a combination of mechanical and enzymatic action, resulting in a process that is variously described as genome ejection, injection, translocation, internalization, or uptake. All phages possess virion-associated adaptations that facilitate this progression (Grayson and Molineux 2007; Roos et al. 2007; Knobler and Gelbart 2009; Molineux and Panja 2013; Bhardwaj et al. 2014) (chapter ▶ “Structure and Function of Bacteriophages”). It should be emphasized that this process of phage genomes reaching the bacterial cytoplasm is not trivially achieved, hence the multiple steps and adaptations involved. Bacterial surface structures, which are usually comprised of polysaccharides, present the first barrier that must be penetrated toward translocation of the phage genome. These structures can be up to hundreds of nanometers thick (Geyer et al. 1983; Hathaway et al. 2012), which can be ten times the diameter of some bacteriophages. Not surprisingly, therefore, many bacteriophages possess enzymes targeting these structures (Geyer et al. 1983; Tomlinson and Taylor 1985; Hughes et al. 1998a; Pires et al. 2016). Interest in these phages and their enzymes are heightened because of their potential to degrade biofilms (Hughes et al. 1998b; Sutherland et al. 2004; Azeredo and Sutherland 2008; Chan and Abedon 2015). For Gram-negative bacteria, the phage genome must cross three barriers: the outer membrane, the peptidoglycan cell wall, and the inner membrane.

112

J. J. Dennehy and S. T. Abedon

Bacteriophages with contractile tails, i.e., members of phage family Myoviridae (chapter ▶ “Structure and Function of Bacteriophages”), rely on mechanical action to penetrate the outer membrane (Hu et al. 2015). For example, with phage T4, attachment of the short tail fibers to LPS triggers the contraction of the phage T4 tail sheath, which results in penetration of the outer membrane by the phage tail tube (Moody 1973). Once the outer membrane has been pierced, the T4 tail tube then enzymatically degrades, via the action of tail tube-associated hydrolytic enzymes, the peptidoglycan layer. The tail tube then crosses the periplasm and fuses with the bacterial plasma membrane (Hu et al. 2015). Gp29, the phage T4 tape measure protein, may form a transmembrane channel that allows the translocation of the genome into the cytoplasm proper. For phages with noncontractile tails, such as members of phage family Siphoviridae, the details of genome translocation are less well understood (Cumby et al. 2015). It is believed that, similar to phage T4, the tape measure protein forms a tube that traverses the bacterial cell wall and membranes, and the phage DNA enters the cytoplasm via this tube (Bohm et al. 2001). Cumby and colleagues showed that two Escherichia coli proteins, FkpA and PtsG, were required for genome translocation by phage HK97, acting as chaperones (chaperonins) to rearrange the tape measure protein into a tubelike structure (Cumby et al. 2015). Specifically, phage HK97 is anchored to the outer membrane and inner membrane via binding to the LamB receptor and the PtsG protein, respectively, and FkpA in the periplasm induces folding of the tape measure protein into the tubelike structure (Cumby et al. 2015). Given their stubby-tailed morphology, members of phage family Podoviridae are unable to extend their tails to the plasma membrane and therefore must use other means to construct channels spanning the periplasm and cell wall. Genome translocation by the podovirus T7 is effected by three phage proteins, gp14, gp15, and gp16, which are ejected from the virion into the cell (Kemp et al. 2004; Chang et al. 2010; Lupo et al. 2015). Gp14 is believed to form a channel across the outer membrane, while the gp15-gp16 complex spans the periplasm and the inner membrane (Chang et al. 2010; Hu et al. 2013; Lupo et al. 2015). It is also believed that the gp15-gp16 complex is able to ratchet the DNA into the cytoplasm using the cell’s proton motive force as a source of energy (Kemp et al. 2004; Molineux and Panja 2013). The gp15-gp16 complex, however, only translocates ~1 kb of the 40 kb genome. The host and eventually the phage T7 RNA polymerase do the rest of this work, pulling the phage DNA into the cell during transcription. The common theme of the above-described genome translocation mechanisms is that bacteriophage tails are exquisitely evolved phage-genome delivery devices. But what of phages lacking tails? Here we describe the genome translocation mechanisms of a tailless phage, the dsRNA, lipid-enveloped, cystovirus and phage Φ6 (phage family Cystoviridae). Phage Φ6 infection of its host, Pseudomonas phaseolicola, begins with the binding of the capsid spike protein P3 to the bacterial pilus (Romantschuk and Bamford 1985). When the pilus is retracted, which may occur, for example, at a rate of 0.5 μm s1 during bacterial movement (Skerker and

Adsorption: Phage Acquisition of Bacteria

113

Berg 2001), then phage Φ6 comes in contact with the bacterial plasma membrane. The phage Φ6 membrane protein P6 then mediates the fusion of the phage Φ6’s lipid envelope and the bacterial outer membrane, which places the phage nucleocapsid into the bacterial periplasm (Romantschuk and Bamford 1985). Once in the periplasm, P5 endolysin proteins embedded in the phage’s nucleocapsid digest the peptidoglycan cell wall (Caldentey and Bamford 1992), and the nucleocapsid penetrates the plasma membrane via membrane invagination and formation of an intracellular vesicle (Romantschuk et al. 1988; Cvirkaite-Krupovic et al. 2010). Cystovirus entry principles, though superficially similar to other phages in that they involve a combination of physical and chemical action to breach cellular barriers, are exceptional among the bacteriophages in terms of specific details and in fact better resemble the genome translocation mechanics observed with eukaryotic viruses than those of the majority of bacterial viruses.

Conclusion The first steps of a phage’s life cycle are commonly described as adsorption, though adsorption itself involves multiple steps, culminating in free virion irreversible attachment to a bacterial cell. Phage acquisition of bacteria in a genetic sense – conversion of a free virion to a virocell – involves even more steps than just irreversible attachment, i.e., at a minimum requiring genome uptake as well. All of these steps are crucial toward both successful phage infection of bacteria (chapter ▶ “Phage Infection and Lysis”) and successful phage therapy (chapter ▶ “Bacteriophage Pharmacology and Immunology”) and plus play crucial roles as well in phage ecological processes (chapter ▶ “Bacteriophage Ecology”). In this chapter we have provided an overview of these various aspects of phage acquisition of bacteria, focusing on virion movement, virion attachment, additional issues governing especially the likelihood of phage attachment given encounter (adsorption factors), and then subsequent genome translocation.

Cross-References ▶ Bacteriophage Ecology ▶ Bacteriophage Pharmacology and Immunology ▶ Bacteriophage Use in Molecular Biology and Biotechnology ▶ Food Safety ▶ Phage Infection and Lysis ▶ Structure and Function of Bacteriophages Acknowledgments JJD acknowledges financial support from the National Institutes of Health NIGMS through grant number 1R01GM124446-01. We also appreciate discussions with Ian Molineux and Dennis Bamford about phage infection mechanisms.

114

J. J. Dennehy and S. T. Abedon

References Abedon ST (1990) The ecology of bacteriophage T4. University of Arizona Abedon ST (2009) Kinetics of phage-mediated biocontrol of bacteria. Foodborne Pathog Dis 6:807–815 Abedon S (2011) Phage therapy pharmacology: calculating phage dosing. Adv Appl Microbiol 77:1–40 Abedon ST (2012) Spatial vulnerability: bacterial arrangements, microcolonies, and biofilms as responses to low rather than high phage densities. Viruses 4:663–687 Abedon ST (2014) Bacteriophages as drugs: the pharmacology of phage therapy. In: Borysowski J, Miedzybrodzki R, Górski A (eds) Phage therapy: current research and applications. Caister Academic Press, Norfolk, pp 69–100 Abedon ST (2016) Phage therapy dosing: the problem(s) with multiplicity of infection (MOI). Bacteriophage 6:e1220348 Abedon ST (2017) Active bacteriophage biocontrol and therapy on sub-millimeter scales towards removal of unwanted bacteria from foods and microbiomes. AIMS Microbiol 3:649–688 Abedon ST (2018) Phage therapy: various perspectives on how to improve the art. Methods Mol Biol 1734:113–127 Abedon ST (2020) Phage therapy: killing titers, multiplicity of infection, adsorption theory, and passive versus active treatments. In: Kurtboke DI, Aminov R (eds) Advances on the applications of bacteriophages. Nova Science Publishers, Hauppauge Abuladze NK, Gingery M, Tsai J, Eiserling FA (1994) Tail length determination in bacteriophage T4. Virology 199:301–310 Augustine J, Louis L, Varghese SM, Bhat SG, Kishore A (2013) Isolation and partial characterization of ΦSP-1, a Salmonella specific lytic phage from intestinal content of broiler chicken. J Basic Microbiol 53:111–120 Azeredo J, Sutherland IW (2008) The use of phages for the removal of infectious biofilms. Curr Pharm Biotechnol 9:261–266 Barr JJ, Auro R, Furlan M, Whiteson KL, Erb ML, Pogliano J, Stotland A, Wolkowicz R, Cutting AS, Doran KS, Salamon P, Youle M, Rohwer F (2013) Bacteriophage adhering to mucus provide a non-host-derived immunity. Proc Natl Acad Sci U S A 110:10771–10776 Bertin A, de Frutos M, Letellier L (2011) Bacteriophage-host interactions leading to genome internalization. Curr Opin Mirobiol 14:492–496 Bhardwaj A, Olia AS, Cingolani G (2014) Architecture of viral genome-delivery molecular machines. Curr Opin Struct Biol 25:1–8 Bohm J, Lambert O, Frangakis AS, Letellier L, Baumeister W, Rigaud JL (2001) FhuA-mediated phage genome transfer into liposomes: a cryo-electron tomography study. Curr Biol 11:1168– 1175 Caldentey J, Bamford DH (1992) The lytic enzyme of the Pseudomonas phage ϕ6. Purification and biochemical characterization. Biochim Biophys Acta 1159:44–50 Casjens SR, Molineux IJ (2012) Short noncontractile tail machines: adsorption and DNA delivery by podoviruses. In: Rossman MG, Rao VB (eds) Viral molecular machines. Springer, Berlin, pp 143–179 Chan BK, Abedon ST (2015) Bacteriophages and their enzymes in biofilm control. Curr Pharm Des 21:85–99 Chang CY, Kemp P, Molineux IJ (2010) Gp15 and gp16 cooperate in translocating bacteriophage T7 DNA into the infected cell. Virology 398:176–186 Chatterjee S, Rothenberg E (2012) Interaction of bacteriophage λ with its E. coli receptor, LamB. Viruses 4:3162–3178 Conley MP, Wood WB (1975) Bacteriophage T4 whiskers: a rudimentary environment-sensing device. Proc Natl Acad Sci U S A 72:3701–3705 Cumby N, Reimer K, Mengin-Lecreulx D, Davidson AR, Maxwell KL (2015) The phage tail tape measure protein, an inner membrane protein and a periplasmic chaperone play connected roles in the genome injection process of E. coli phage HK97. Mol Microbiol 96:437–447

Adsorption: Phage Acquisition of Bacteria

115

Cvirkaite-Krupovic V, Poranen MM, Bamford DH (2010) Phospholipids act as secondary receptor during the entry of the enveloped, double-stranded RNA bacteriophage ϕ6. J Gen Virol 91:2116–2120 Danis-Wlodarczyk K, Dąbrowska K, Abedon ST (2020) Phage therapy: the pharmacology of antibacterial viruses. In: Coffey A (ed) Exploitation of bacteriophages for biocontrol and therapeutics. Caister Academic Press, Norwich Delbrück M (1946) Bacterial viruses or bacteriophages. Biol Rev 21:30–40 Dennehy JJ, Friedenberg NA, Yang YW, Turner PE (2006) Bacteriophage migration via nematode vectors: host-parasite-consumer interactions in laboratory microcosms. Appl Environ Microbiol 72:1974–1979 Forterre P (2013) The virocell concept and environmental microbiology. ISME J 7:233–236 Gallet R, Kannoly S, Wang IN (2011) Effects of bacteriophage traits on plaque formation. BMC Microbiol 11:181 Garcia-Doval C, van Raaij MJ (2013) Bacteriophage receptor recognition and nucleic acid transfer. Subcell Biochem 68:489–518 Geyer H, Himmelspach K, Kwiatkowski B, Schlecht S, Stirm S (1983) Degradation of bacterial surface carbohydrates by virus-associated enzymes. Pure Appl Chem 55:637–653 Goldberg E (1983) Recognition, attachment, and injection. In: Mathews CK, Kutter EM, Mosig G, Berget PB (eds) Bacteriophage T4. American Society for Microbiology, Washington, DC, pp 32–39 Grayson P, Molineux IJ (2007) Is phage DNA ‘injected’ into cells – biologists and physicists can agree. Curr Opin Mirobiol 10:401–409 Hagens S, Loessner MJ (2010) Bacteriophage for biocontrol of foodborne pathogens: calculations and considerations. Curr Pharm Biotechnol 11:58–68 Hathaway LJ, Brugger SD, Morand B, Bangert M, Rotzetter JU, Hauser C, Graber WA, Gore S, Kadioglu A, Muhlemann K (2012) Capsule type of Streptococcus pneumoniae determines growth phenotype. PLoS Pathog 8:e1002574 Hendrix RW, Duda RL (1992) Bacteriophage lambda PaPa: not the mother of all lambda phages. Science (New York, NY) 258:1145–1148 Henning U, Hashemolhosseini S (1994) Receptor recognition by T-even type coliphages. In: Karam JD, Eiserling FA, Black LW (eds) The molecular biology of bacteriophage T4. ASM Press, Washington, DC, pp 291–298 Hershey AD, Chase M (1952) Independent functions of viral protein and nucleic acid in growth of bacteriophage. J Gen Physiol 36:39–56 Hu B, Margolin W, Molineux IJ, Liu J (2013) The bacteriophage T7 virion undergoes extensive structural remodeling during infection. Science (New York, NY) 339:576–579. https://pubmed. ncbi.nlm.nih.gov/23306440/ Hu B, Margolin W, Molineux IJ, Liu J (2015) Structural remodeling of bacteriophage T4 and host membranes during infection initiation. Proc Natl Acad Sci U S A 112:E4919–E4928 Hughes KA, Sutherland IW, Clark J, Jones MV (1998a) Bacteriophage and associated polysaccharide depolymerases-novel tools for study of bacterial biofilms. J Appl Microbiol 85:583–590 Hughes KA, Sutherland IW, Jones MV (1998b) Biofilm susceptibility to bacteriophage attack: the role of phage-borne polysaccharide depolymerase. Microbiology 144:3039–3047 Hyman P, Abedon ST (2009) Practical methods for determining phage growth parameters. Methods Mol Biol 501:175–202 Igler C, Abedon ST (2019) Commentary: a host-produced quorum-sensing autoinducer controls a phage lysis-lysogeny decision. Front Microbiol 10:1171 Kanner LC, Kozloff LM (1964) The reaction of indole and T2 bacteriophage. Biochemistry 3:215–223 Katsura I (1987) Determination of bacteriophage lambda tail length by a protein ruler. Nature 327:73–75 Kemp P, Gupta M, Molineux IJ (2004) Bacteriophage T7 DNA ejection into cells is initiated by an enzyme-like mechanism. Mol Microbiol 53:1251–1265 Knobler CM, Gelbart WM (2009) Physical chemistry of DNA viruses. Annu Rev Phys Chem 60:378–383

116

J. J. Dennehy and S. T. Abedon

Koch AL (1960) Encounter efficiency of coliphage-bacterium interaction. Biochim Biophys Acta 39:311–318 Kutter E, Kellenberger E, Carlson K, Eddy S, Neitzel J, Messinger L, North J, Guttman B (1994) Effects of bacterial growth conditions and physiology on T4 infection. In: Karam JD, Kutter E, Carlson K, Guttman B (eds) The molecular biology of bacteriophage T4. ASM Press, Washington, DC, pp 406–418 Lee KL, Hubbard LC, Hern S, Yildiz I, Gratzl M, Steinmetz NF (2013) Shape matters: the diffusion rates of TMV rods and CPMV icosahedrons in a spheroid model of extracellular matrix are distinct. Biomater Sci 1:581 Lupo D, Leptihn S, Nagler G, Haase M, Molineux IJ, Kuhn A (2015) The T7 ejection nanomachine components gp15-gp16 form a spiral ring complex that binds DNA and a lipid membrane. Virology 486:263–271 Molineux IJ, Panja D (2013) Popping the cork: mechanisms of phage genome ejection. Nat Rev Microbiol 11:194–204 Montag D, Hashemolhosseini S, Henning U (1990) Receptor-recognizing proteins of T-even type bacteriophages. The receptor-recognizing area of proteins 37 of phages T4 TuIa and TuIb. J Mol Biol 216:327–334 Moody MF (1973) Sheath of bacteriophage T4. 3. Contraction mechanism deduced from partially contracted sheaths. J Mol Biol 80:613–635 Murray AG, Jackson GA (1992) Viral dynamics: a model of the effects of size, shape, motion, and abundance of single-celled planktonic organisms and other particles. Mar Ecol Prog Ser 89:103– 116 Pires DP, Oliveira H, Melo LD, Sillankorva S, Azeredo J (2016) Bacteriophage-encoded depolymerases: their diversity and biotechnological applications. Appl Microbiol Biotechnol 100:2141–2151 Reche I, D’Orta G, Mladenov N, Winget DM, Suttle CA (2018) Deposition rates of viruses and bacteria above the atmospheric boundary layer. ISME J 12:1154–1162 Romantschuk M, Bamford DH (1985) Function of pili in bacteriophage ϕ6 penetration. J Gen Virol 66:2461–2468 Romantschuk M, Olkkonen VM, Bamford DH (1988) The nucleocapsid of bacteriophage ϕ6 penetrates the host cytoplasmic membrane. EMBO J 7:1821–1829 Roos WH, Ivanovska IL, Evilevitch A, Wuite GJ (2007) Viral capsids: mechanical characteristics, genome packaging and delivery mechanisms. Cell Mol Life Sci 64:1484–1497 Russel M, Model P (2006) Filamentous bacteriophages. In: Calendar R, Abedon ST (eds) The bacteriophages. Oxford University Press, Oxford, pp 146–160 Saltzman WM, Radomsky ML, Whaley KJ, Cone RA (1994) Antibody diffusion in human cervical mucus. Biophys J 66:508–515 Schwartz M (1975) Reversible interaction between coliphage lambda and its receptor protein. J Mol Biol 99:185–202 Schwartz M (1976) The adsorption of coliphage lambda to its host: effect of variation in the surface density of the receptor and in phage-receptor affinity. J Mol Biol 103:521–536 Shao Y, Wang I-N (2008) Bacteriophage adsorption rate and optimal lysis time. Genetics 180:471– 482 Sisler FD (1940) In: University of Maryland (ed) The transmission of bacteriophage by mosquitoes, College Park Skerker JM, Berg HC (2001) Direct observation of extension and retraction of type IV pili. Proc Natl Acad Sci U S A 98:6901–6904 Stent GS (1963) Molecular biology of bacterial viruses. WH Freeman and Co, San Francisco Storms ZJ, Sauvageau D (2015) Modeling tailed bacteriophage adsorption: insight into mechanisms. Virology 485:355–362 Storms ZJ, Arsenault E, Sauvageau D, Cooper DG (2010) Bacteriophage adsorption efficiency and its effect on amplification. Bioprocess Biosyst Eng 33:823–831

Adsorption: Phage Acquisition of Bacteria

117

Sutherland IW, Hughes KA, Skillman LC, Tait K (2004) The interaction of phage and biofilms. FEMS Microbiol Lett 232:1–6 Tomlinson S, Taylor PW (1985) Neuraminidase associated with coliphage E that specifically depolymerizes the Escherichia coli K1 capsular polysaccharide. J Virol 55:374–378 Trubl G, Hyman P, Roux S, Abedon ST (2020) Coming-of-age characterization of soil viruses: a user’s guide to virus isolation, detection within metagenomes, and viromics. Soil Sys 4:23 van Raaij MJ, Schoehn G, Burda MR, Miller S (2001) Crystal structure of a heat and protease-stable part of the bacteriophage T4 short tail fibre. J Mol Biol 314:1137–1146

Temperate Phages, Prophages, and Lysogeny Joanna Łoś, Sylwia Zielińska, Anna Krajewska, Zalewska Michalina, Aleksandra Małachowska, Katarzyna Kwaśnicka, and Marcin Łoś

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lysogeny and Lysogenic Cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stably Associating with Host Bacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Benefiting the Lysogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lysis-Lysogeny Decisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Induction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Immunity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lysogenic Conversion and Morons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bacteriophage Lambda . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Lysogeny Decision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lambda and Lambdoid Genomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Prophages P2 and P2-Likes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P2 Virion and Genome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P2 Lysogeny . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Temperate Phages of Staphylococci and Streptococci . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Staphylococcal Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Streptococcal Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

120 121 121 123 124 125 126 126 128 128 129 130 131 131 132 132 134

J. Łoś · S. Zielińska · M. Łoś (*) Department of Bacterial Molecular Genetics, Faculty of Biology, University of Gdansk, Gdansk, Poland Phage Consultants, Gdansk, Poland e-mail: [email protected] A. Krajewska · Z. Michalina Department of Bacterial Molecular Genetics, Faculty of Biology, University of Gdansk, Gdansk, Poland A. Małachowska Department of Genetics and Biosystematics, Faculty of Biology, University of Gdansk, Gdansk, Poland K. Kwaśnicka Department of Molecular Biology, Faculty of Biology, University of Gdansk, Gdansk, Poland © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_3

119

120

J. Łoś et al.

Role of Prophages in Modulation of Bacterial Host Biology and Evolution . . . . . . . . . . . . . . . . . . Toxins and Virulence Factors Encoded by Prophages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

134 135 144 145

Abstract

Prophages form with their host a very special type of interaction called lysogeny. There they are able to help the host, but they are also able to kill their hosts, producing their own virion progeny during the process. The phage-host interaction itself is very complicated with many different mechanisms, some of which served as important bases for our understanding of molecular biology. Because of these mechanisms, some prophages are extensively studied and now are considered to be model organisms. Other prophages have caught our attention because they have managed to turn their hosts into deadly pathogens by delivering payloads of toxin genes along with genes encoding other bacterial virulence factors. Many prophages are also able to cross species borders, facilitate horizontal gene transfer, and otherwise give rise to the creation of bacteria, via lysogenization, with new capacities not necessarily observed before. Due to improvements in sequencing technologies, we are now discovering how widespread and important the interaction of prophages with their hosts is in nature. In this chapter some aspects of their biology, interactions with hosts, and contribution to pathogenesis is described.

Introduction Not all phage infections, even if successful, progress directly to virion maturation and release. Instead, it is possible for certain phages, which are called temperate, to infect bacteria, persist, and replicate, not explicitly as viruses but instead as a capsidfree genetic element. In this state the phage genome instead is described as a prophage and the phage-infected bacterium a lysogen. Prophages and associated lysogens can stably persist over long time spans, but as infections by a virus, nevertheless retain a potential to transition to the production of virion progeny via a process known as prophage induction. Thus, free phages can give rise either to what are known as lysogenic cycles or instead to productive cycles (the latter, e.g., lytic cycles), with the transition between these different states variously described as lysis-lysogeny, lytic-lysogeny, or lysis-lysogenic decisions. While productive cycles give rise to free phages, lysogenic cycles give rise either to continued lysogenic cycles or instead, following induction, to productive cycles. While in the lysogenic state, prophages can contribute to the phenotype of the hosting bacterium, displaying such phenomena as superinfection immunity and lysogenic conversion. This chapter summarizes some basic knowledge about lysogeny and prophages, including the mechanisms they use to assure their survival, properly interact with their hosts, and otherwise augment host capabilities. Previous reviews on this subject include, e.g., those of Łoś et al. (2010), Casjens and Hendrix (2015), and HowardVarona et al. (2017). In this volume see also chapter ▶ “Phage Infection and Lysis”

Temperate Phages, Prophages, and Lysogeny

121

as well as chapter ▶ “Bacteriophage-Mediated Horizontal Gene Transfer: Transduction”. Here we consider the basic, general aspects of lysogeny and prophages as well as aspects associated with specific temperate phages.

Lysogeny and Lysogenic Cycles Lysogeny is a long-lasting partnership between a bacterial cell and an infecting phage. The phage at least temporarily gives up its ability to produce virions, while the bacterium generally tolerates and even may benefit from the phage’s presence. Such bacteria in association with these phages are called lysogens, and phages which are able to display such lysogenic cycles are referred to as temperate. The vast majority of temperate phages are also lytic phages, in terms of their productive cycles, though chronically releasing temperate phages are known as well. Most infections by temperate phages seem to result in virion-productive infections; lysogenic cycles nevertheless will also begin with the adsorption of a temperate phage virion to a susceptible bacterial cell, which upon entrance into lysogeny takes on a form, in terms of its genome, known as a prophage.

Stably Associating with Host Bacteria Temperate phages display a variety of ways to stably maintain their prophage in host bacterial cells. This can involve direct integration of the genetic material of prophage with bacterial chromosome, but also prophages exist in the form of stable, extrachromosomal elements, such as plasmids (Fig. 1). Stable maintenance of prophages in the form of a plasmid is more demanding than chromosomal prophages as every cell division may result in losing of the plasmid by one of the daughter cells. That tendency, however, can be counteracted by existing as multicopy plasmids, but this in turn will create an increased burden for host cells, which may cause the lysogen to lose, e.g., a competitive edge against other bacteria. Some temperate phages have evolved very sophisticated sets of mechanisms to force host cells to stably maintain a prophage copy. One of these prophages, which lysogenize cells in the form of a plasmid, is phage P1, which uses at least two mechanisms to ensure stable prophage maintenance. These mechanisms are described in detail in a separate chapter of this book. Briefly, they consist of a partitioning system, which is able to deliver a single copy of prophage to each daughter cell and a toxin-antitoxin addiction system, which is capable of killing daughter cells which have failed to acquire a copy of the prophage plasmid (Yarmolinsky 2004). Some temperate phages bypass the problems with plasmid inheritance simply by evolving a way to integrate into the bacterial chromosome, and thus “outsourcing” to the host the whole work of prophage genome replication and segregation to daughter cells. The process of integration of the prophage genome into the chromosome is usually carried out by highly specialized integrases, which will insert the prophage genome into the predefined integration sites, as occurs in the case of phage lambda (Hendrix 2002), but the process may be much more random and conducted by

Fig. 1 Developmental pathways of temperate phage integrating into chromosome (left panel) and integrating with a cell as a plasmid (right panel)

122 J. Łoś et al.

Temperate Phages, Prophages, and Lysogeny

123

transposases as in a case of phage Mu (Ranquet et al. 2005). Many temperate phages integrate into bacterial chromosomes sometimes in multiple locations. Some of them, in a multi-chromosome bacterial species, may integrate to a few chromosomes at the same time. An example of this is CTXcla, a cholera toxin encoding prophage which is integrated into both chromosomes of Vibrio cholerae in classical biotype strains (Kim et al. 2014, 2017). Regardless of the form of the prophage within the host cell, there is a general scheme of maintaining lysogeny, which may be considered to form the basis of a temporary partnership. Following infection of the host cell, prophages need to silence a vast majority of their genes, due in part to the majority of them not being needed in the prophage state (e.g., structural capsid genes) but also that some of them may encode products toxic to the host (e.g., prophage lambda kil gene). The silencing of those genes is maintained by the phage repressor or repressors, which are able to suppress activity of promoters needed for release of phages from lysogeny, which is then followed by progress toward further development steps. Very often, repressors are sophisticated molecular sensors capable of recognizing molecular signals, which then release prophages from repression in response to such signals. In the case of many prophages, mostly represented by large groups of lambdoid phages, such a signal is a triggering of SOS response regulons, which cause repressors to self-destruct in the process of autoproteolysis (McCabe et al. 2005). For other prophages, such as those of phage Mu or P2-like prophages, even though the prophage-inducing signal is not known and not synchronized, massive induction has been observed in case of these phages (Ranquet et al. 2005). Release of prophage repression leads to activation of productive cycle-associated promoters and thus progression of the development cycle toward production of progeny virions and destruction of the host cell.

Benefiting the Lysogen The interaction of the prophage with a host cell may be beneficial for both sides, as prophages usually have lysogenic conversion genes and morons onboard. These genes, while not contributing to the development cycle of prophage, nevertheless are active during the whole lysogeny period and provide some beneficial functions for the host cells, allowing them to perform better in certain situations or environment. Moreover, repressors of prophages may also modulate host cell metabolism, directing it into more effective and competitive use of the available resources, often resulting in faster growth of the lysogenized host cells, when compared to their prophage-free counterparts (Chen et al. 2005). These types of interactions will be discussed more in following paragraphs. Regardless of mutual benefits, lysogeny can have deadly output for each side, as both of them – prophage and bacteria – are trying to outsmart each other. Prophage is constructed in such a way to eventually be induced in response to problems which may appear in the cell. As so, it is a kind of molecular time bomb, which while offering temporary benefits, host bacteria will eventually be triggered. Bacterial cells, however, have mechanisms of reducing

124

J. Łoś et al.

unnecessary genetic content, which will, in a random way, in a process of natural selection, try to get rid of or at least disable the dangerous part of prophage while keeping the beneficial part at its own disposal. The driving DNA deletion mechanism is random, but those bacterial cells, which managed to disable prophage and thus turn it into cryptic prophage, while keeping its beneficial elements, may obtain an advantage over their counterparts, which do not contain prophage, or in which the prophage contained is fully functional and thereby able to eventually kill the host cell. The resulting cryptic prophages may be then identified in bacterial chromosomes like, e.g., in E. coli K12 (Campbell 1998). A prophage is a phage genome which has come to occupy the interior of a bacterial cell, all the while retaining the potential to produce new, progeny virions. But so long as induction has not taken place, they do not actually produce those new virions, indeed any virions. Prophages thus are an example of a bacteriophage lifestyle which is not a direct part of a productive cycle. The genetic association of prophages with their hosts at the first glance, however, would seem to be a parasitic interaction, as prophages utilize host resources in order to multiply their genomes. However, it is in the interest of the prophage – sitting inside of these cells – to not so-cripple its host, which could have the effect of allowing other bacteria to outcompete that lysogen. Rather, prophages can fine-tune and equip their hosts with numerous useful tools, i.e., in terms of lysogenic conversion as introduced above. In the case of bacteriophage lambda, but also many other prophages, the metabolism of the cell is modulated by inhibition of the gluconeogenesis pathway (Dykhuizen et al. 1978; Chen et al. 2005), which allow lysogenized bacteria to grow faster and more efficiently and to outcompete their prophage-free counterparts. Lysogenic cycles at first glance may look like they are not very productive for the phage, but in fact, prophages are reproduced and transferred to every daughter cell, and thus, in the long run, can be more productive than simple lytic cycles. Prophages during multiplication of host cells nevertheless may be induced in a small fraction of a lysogen population, but even this small frequency of induction allows phages to keep their presence as virions in environments, sometimes even at a relatively high level. In fact, this strategy may ensure very effective spread in the environment (Rotman et al. 2010) which is due to the potentially much higher number of progeny phage particles. It can be easily explained by the fact that phage, by killing a single cell, may produce up to ~1000 progeny phages, but the same cell lysogenized by phage may produce billions of daughter cells, all of them carrying prophage. Let us say that among them, thousands of cells will undergo spontaneous inductions, releasing phage progeny, which gives an effective productivity from a single infection event at the level of millions of phage particles.

Lysis-Lysogeny Decisions Entering into a lysogenic cycle by temperate bacteriophage is a serious commitment. This decision, made soon after a temperate phage’s adsorption to a host bacterium, binds the fate of the phage with that bacterium, so has to be made carefully, and

Temperate Phages, Prophages, and Lysogeny

125

chosen particularly when benefits of lysogeny seem to be higher than the benefits of immediate use of host resources to build progeny phage particles. Lysis-lysogeny decisions, perhaps not surprisingly, therefore can also be molecularly very complex. The process of choosing the developmental pathway by phage has been best studied in bacteriophage lambda, where it is still not fully understood, and thereby additional pieces of the puzzles continue to be added to the already complex picture. In the case of this phage, the decision is made on the basis of several signals derived from levels of cAMP and ppGpp alarmones, which indicate the physiological state of the cell. In general, the lower if the level of cell resources and/or the poorer the environment, then the more probable will be the choice of prophage state (Słominska et al. 1999). The rationale behind this is that the more resource-starved the cell, then the fewer progeny it can immediately produce, and thus it may be wise to wait for better times via lysogenic cycles, meanwhile improving host cell performance by fine-tuning host metabolism and adding additional functions by the process of lysogenic conversion, which may help the host, e.g., to colonize a new ecological niche. Another signal, which can be sensed by incoming temperate prophage, is the density of bacteriophages in the environment. In the case of bacteriophage lambda, the more phages which infect the cell in short period of time, then the more likely will be the choice of lysogenization instead of lytic development by the phage. This may be explained by the fact that multiple infections by phage during the lysis versus lysogeny decision may indicate very high local phage densities, which in turn suggest that the local environment is already saturated with phage and thus number of uninfected host cells may be very low. In such situations the most rational choice would be to lysogenize the cell and to reserve cell and its progeny resources for potential production of the progeny phages (Avlund et al. 2009). Recently it was discovered that some phages use an active communication system showing how many cells were lysogenized in the environment. It is achieved by the secretion of very short signal peptides, with concentration influencing the decision of lysis versus lysogeny. High enough levels of these peptides in particular promote lysogeny (Erez et al. 2017). Various methods of communication between phages outside the cell and phages already occupying the cell were summarized by Abedon (2017).

Induction Once taking on the form of a prophage, the virus in addition to expressing lysogenic conversion genes and otherwise replicating in tandem with its host also waits for the molecular signal to be induced. Prophages, that is, can be considered to be potentially only temporary elements of a host cell. All prophages, in other words, sooner or later have up to four possible fates: ongoing existence as a functional prophage (though its sequence will not necessarily remain constant, i.e., prophages in principle can evolve while remaining genetically intact), death in the course of death of its host, inactivation as functional prophage (thus become cryptic prophages, but also outright curing, i.e., prophage loss in full) despite ongoing host survival, or induction. Inactivation may be a result of deletions or mutations within prophage genomes caused by the activity of host enzymes or simple error in DNA handling. Induction,

126

J. Łoś et al.

in turn, is prophage ability to enter a productive development cycle and thereby use host resources to produce progeny phages. As noted above, prophages, to prevent development toward a lytic pathway, need to silence genes responsible for this lytic development. This suppression is achieved by repressor proteins, which are constitutively expressed during lysogeny. They not only suppress the activity of lytic cycle-associated promoters but can regulate activity of their own promoter, in order to keep repressor-protein concentrations within certain ranges. Very often repressors are sophisticated molecular sensors capable of recognizing molecular signals and releasing prophages from repression in response to such signals.

Immunity As noted, prophages after lysogenization of the host cell produce repressor proteins, which keep them in a prophage state and prevent activation of lytic promoters. Each lysogenized cell must contain enough repressor to prevent each of their prophages from being induced. Otherwise the cell will host a lytic cycle resulting in lysis and release of new virion progeny. A cell once lysogenized is immune to infection with the same phage, which is called homoimmunity, or any phage containing the same immunity region, which state is called heteroimmunity. An immunity region is a complete set of elements containing repressor gene and operators, which are sequences to which repressor proteins bind. Operators are responsible for proper placing of repressors within or in proximity of the promoter region, and thus they are responsible for silencing or enhancing that promoter activity. If the DNA of a phage using the same repressor is injected into a cell, then their operators will immediately be occupied with repressors typical for that prophage (Yarmolinsky 2004; Ptashne 2004). This will block any further development of that phage toward a lytic cycle, and in majority of cases it will also block proper prophage establishment (lysogenic cycle), as transcription of some early genes responsible for, e.g., prophage integration, will also be suppressed. Phage genome injected into lysogenized host with the same immunity may still be able to form a lysogen, however, but in many cases with very low frequency (Fogg et al. 2010).

Lysogenic Conversion and Morons Lysogenization of bacteria with a prophage very often causes the host bacterium to gain new properties that are beyond those associated simply with lysogenization. Most frequently the modifications are due to expression of lysogenic conversion genes which allow bacteria to gain new properties and potentially colonize new ecological niches. Prophages, directly by lysogenic conversion and indirectly by participating in the spread of genomic islands, including pathogenicity islands,

Temperate Phages, Prophages, and Lysogeny

127

consequently can greatly contribute to the diversification of bacterial strains, including in terms of the emergence of highly virulent pathogens. Lysogenic conversion can be defined as a temperate phage-associated heritable change in the host cell’s genotype and phenotype that is not caused by simply packing host DNA instead of phage genome in the process of generalized transduction and is independent of the effects expected from repression and integration and other phenomena related to the lysogenic state. The latter, not lysogenic conversion-related changes, can include immunity to superinfection or loss of bacterial functions caused by the insertion of phage into host genes (Łoś et al. 2010). Lysogenic conversion can be, but strictly need not have been, caused by morons, which are defined as an additional gene in a prophage genome, often acquired by horizontal gene transfer, without direct function in mediating either a phage’s lysogenic or lytic cycle (Brüssow et al. 2004). The function of some morons is elusive and may not even be easily observed as a lysogenic conversion. On the other hand, some lysogenic conversion effects, e.g., blocking of the gluconeogenesis pathways by phage repressors, are not caused by morons. This is a cause, when genes, which are essential elements of bacteriophage development or prophage maintenance, are responsible for additional effects on host recognized as a lysogenic conversion. Thus, even though there is an overlap in lysogenic conversion and the function of morons, these two terms have to be treated as a separate phenomenon (Table 1). Lysogenic conversion was discovered in 1951 with the observation that diphtheria toxin is encoded on the prophage genome integrated into Corynebacterium diphtheriae (Freeman 1951). Since then, bacteriophage-encoded toxins have been found in a range of both Gram-positive and Gram-negative strains, including Escherichia coli, Shigella spp., Pseudomonas aeruginosa, Vibrio cholerae, Clostridium tetani, Clostridium botulinum, Staphylococcus aureus, and Streptococcus pyogenes (Barksdale and Arden 1974; O’Brien et al. 1984; Huang et al. 1987; Nakayama et al. 1999; Betley and Mekalanos 1985; Weeks and Ferretti 1984; Goshorn and Schlievert 1989). The focus of researchers on pathogens caused a sort of bias, which may suggest that lysogenic conversion is a phenomenon mostly observed in pathogens. It is very likely, however, that the truth is the opposite, with pathogens best studied in comparison to environmental bacteria. Moreover, lysogenic conversion which leads to change in the pathogenicity of host strain is relatively easy to observe. That is not a case in environmental strains, which we often cannot propagate in laboratory or draw too many conclusions from by observation in their natural environments, versus the comparatively straightforward and easy to monitor environments consisting of pathogens causing disease. In this chapter different aspects of prophage biology were discussed. It is important to stress, that all above mentioned features of prophages are never observed as a complete feature set of single prophages. Instead, each one is utilizing a few of them. In subsequent paragraphs different examples of prophages and their biology will be discussed, to show how these mechanisms make a complete and fully functional molecular program allowing prophage to interact properly with its host.

J. Łoś et al.

128

Table 1 Differences and similarities of lysogenic conversion and the action of morons Causative agent Integral part of the phage cycle? Phenotypic change? Causative element disposable for phage?

Lysogenic conversion Additional or native gene or genetic element in phage No

Action of morons Additional gene in phage No

Yes Yes or no. Ability to dispose element may be conditional

Yes or no Yes

Bacteriophage Lambda Bacteriophage lambda was discovered over 60 years ago and became one of the most important model organisms in molecular biology. Basic molecular mechanisms of crucial cellular processes and regulation of development were investigated using bacteriophage lambda as a model. Many similar phages called lambdoid bacteriophages have since been isolated and characterized. These phages have a similar genome organization to that of phage lambda, and they can recombine with lambda to make biologically functional hybrids (Casjens and Hendrix 2015). Grose and Casjens (2014) defined the lambda supercluster as a group of temperate Enterobacteriaceae phages whose encoded functions are syntenic with the phage lambda genome and whose transcription pattern and gene expression cascade are similar to that of lambda. In this supercluster are only Enterobacteriaceae phages. Others which are similar to lambda phage but infect different bacteria have some of their genes, in many cases lysis genes, which are not syntenic with those of the lambda supercluster (Grose and Casjens 2014). There are also phages which differ substantially in nucleotide sequences from lambda but have a lambda-like lifestyle with similar transcription regulatory mechanisms (Hendrix 2002).

The Lysogeny Decision One of the most studied features of lambda phage is the molecular switch, which is responsible for the lysis versus lysogeny decision. Despite the fact, that the mechanism of this switch is generally quite well understood, there are still many unknowns in the fine-tuning of the decision-making driven by it. In general, the process starts after the phage genome has entered the bacterial cell, when the bacteriophage makes a decision whether to produce new progeny phages and lyse the host cell, or instead to form a prophage and lysogenize the cell. Lytic development pathway is a default for the phage. Switching to the lysogeny pathway during phage lambda development depends on a level of accumulation of the phage lambda CII protein, since this protein is a transcriptional activator that stimulates strong cI gene transcription from the lambda phage pE promoter. CI repressor, the product of cI gene, is required to maintain lysogeny state. It

Temperate Phages, Prophages, and Lysogeny

129

forms multimers that bind simultaneously to three operators known as OL and three others known as OR to repress the early promoters. Moreover, synthesis of Int protein that catalyzes insertion of the phage lambda dsDNA genome into the host chromosome is stimulated by CII transcriptional activation of the phage lambda promoter, pI (Węgrzyn and Węgrzyn 2005). Expression of phage lysis genes is also inhibited by activity of the cII-dependent paQ promoter which in turn directs the production of antisense mRNA for the Q protein, necessary for late genes expression (Hoopes and McClure 1985). Quantity of CII is also influenced by cell physiology, environmental factors and number of phage particles infecting the cell. Lysogenization is much more frequent in a starved bacterium. The CII protein is very unstable and degraded by the hostencoded FtsH protease (Shotland et al. 2000). During starvation, production of cyclic AMP is started. This results in inhibition of FtsH activity and increased stability of CII. Also, another nucleotide alarmone – ppGpp – regulates the amount of FtsH protease and modulates cII-mediated activation of pE and pI promoters (Słominska et al. 1999). Additionally, ppGpp negatively regulates pR promoter activity that results in inhibition of the lytic development of phage lambda (Potrykus et al. 2002). In slowly growing bacteria, polyadenylation of mRNA by poly (A) polymerase increases, resulting in lower stability of mRNAs. Expression of cII gene is negatively regulated by oop RNA transcript that after polyadenylation is degraded, and the cII gene therefore is more effectively expressed (Szalewska-Palasz et al. 1998). Another factor controlling lysis-lysogeny decisions is temperature. Formation of CII, which is active in tetrameric form, depends on temperature. At lower temperatures the process of multimerization is more effective. Also, FtsH-dependent degradation of CII tetramers is less effective than monomers (Shotland et al. 2000). Stability of CII protein depends on phage-encoded CIII protein that is also a substrate for FtsH protease (Herman et al. 1997). Expression of the cIII gene is under control of pL promoter that is more active at lower temperatures (Giladi et al. 1995). Moreover, the transcript for cIII forms two alternative structures and only one of them is able to bind to the 30S ribosomal subunit and initial translation. The proportion of these two structures depends on temperature, and at lower temperatures there is more transcripts that are able to bind to the ribosome (Altuvia et al. 1989). Lysogeny is also favored during high multiplicity of infection. When a low number of phages infect the cell, injection of phage DNA takes place on cellular poles. Interestingly, it was found that the majority of FtsH that degrades CII is located in these regions. On the other hand, during high multiplicity of infection, injection of phage DNA takes place on whole cell surface (Edgar et al. 2008).

Lambda and Lambdoid Genomes Phage lambda genome is 48,503 bp long, and upon lysogenization it is inserted into bacterial chromosome between the Escherichia coli gal and bio genes. The integration is reversed during the induction process, and the prophage genome is excised. Usually this process is very precise, but sometimes the prophage is not excised

130

J. Łoś et al.

properly and chunks of host DNA flanking attachment sites may be taken away, replicated together with phage DNA and packed into produced phage capsids, thus allowing for specialized transduction of flanking DNA (Rolfe 1970) (chapter ▶ “Bacteriophage-Mediated Horizontal Gene Transfer: Transduction”). Lambda and lambdoid prophages carry powerful recombination systems, which make it much easier for them to exchange whole, functional gene blocks, and as a result these phages possess genomes which are highly mosaic. Although occasionally their sequence may not be very similar to each other, they share general similarity in genome structure (gene synteny), even though they may belong to different phage families, as lambda belongs to family Siphoviridae, while a majority of Shiga toxin containing lambdoid phages seem to belong to phage family Podoviridae (Smith et al. 2012). Lambdoid prophages very often deliver to the host various morons (i.e., additional DNA fragments originating originally from outside of phage genome) and lysogenic conversion genes (chapter ▶ “Bacteriophage-Mediated Horizontal Gene Transfer: Transduction”), which can dramatically change the lysogenized host properties. The grim reputation of a fraction of lambdoid bacteriophages is caused by their ability to carry Shiga toxin genes, which can turn otherwise quite innocent E. coli strains into deadly pathogens (Smith et al. 2012). Phage lambda, which is a commonly used model in scientific laboratories and seems to be harmless, nevertheless encodes a serum resistance gene, bor, which in may also augment the E. coli host, potentially contributing to its formation into a pathogen (Barondess and Beckwith 1995).

Prophages P2 and P2-Likes Phage P2 is a temperate bacteriophage isolated by Prof. Giuseppe Bertani in 1951 from the Lisbonne and Carère strain of Escherichia coli, together with phage P1 and P3. (Ackermann 1999). Since then a large number of P2-like phages have been isolated, but P2 is still the most well-described member of the family, along with phage 186. The term P2-like phages is used for phages that share some but not all features with P2 phage. Some of the most known members of P2-like phages are the noted 186 along with phages HP1, HK239, and WΦ. Classification of the P2-like family is based on characteristics like serological relatedness, host range, lack of inducibility by ultraviolet irradiation, inability to recombine with phage lambda, a unique class of cohesive DNA ends on chromosomes, and also the possibility to support growth of satellite phage P4. Genome sequencing has shown that P2-like phages are relatively common in nature and can multiply in different γ-proteobacteria. P2 phage itself can infect most strains of E. coli, as well as Serratia marcescens, Salmonella typhimurium, Klebsiella pneumoniae, or Yersinia sp. It was shown that in E. coli reference library, ECOR (Ochman and Selander 1984), about 30% of the strains encloses P2-like prophages (Nilsson et al. 2004). Upon infection p-phage P2 can derepress phage P4 lysogens via the action of the P2 Cox protein (below), which acts as transcriptional activator of the late P4 promoter, Pll, which is required for P4 replication.

Temperate Phages, Prophages, and Lysogeny

131

P2 Virion and Genome The P2 virion consists of an icosahedral head 60 nm in diameter and complex tail of 135 nm with a contractile sheath. During the coinfection with phage P4, it produces a smaller 45 nm head (Dokland et al. 1992). The baseplate at the end of the tail contains six tail fibers and a spike to cling to the bacterium during the adsorption. Free phage P2 adsorbs during the infection to the core region of the lipopolysaccharide of E. coli and injects its DNA into the cytoplasm. The entire genome has been sequenced (Gene Bank accession number AF063097) and the double-stranded DNA molecule consists of 33.6 Kb, with 19-nt-long cohesive ends (Linderoth et al. 1991). The cohesive ends allow for circularization of genome after infection (Bertani and Six 1988), one of them is fastened to the tail, at the head-tail attachment site, which prevents circularization inside the capsid (Chattoraj and Inman 1974). The 42 phage P2 genes can be organized into 3 essential classes. One class contains genes involved in lysogenization, another class of genes necessary for DNA replication, and a third class consisting of genes encoding structural proteins and lysis functions. Moreover, P2 includes a number of open reading frames (ORFs) that may encode functional proteins.

P2 Lysogeny After the start of infection, phage P2 can enter into two different life cycles, either lytic growth or lysogeny. The choice of life cycle depends on the promoter that takes control. The early promoter, Pe, controls lytic growth and the Pc promoter controls the genes involved in lysogenization. The balance between the phage-encoded repressor proteins, C and Cox, is responsible for determining the outcome of the lysis versus lysogeny decision. This control is termed a transcriptional or development switch. The promoter Pc is repressed by Cox (meaning Control of excision), which is the first gene controlled by the promoter, Pe, which prevents the expression of the genes necessary for lysogeny (Saha et al. 1987). During this repression, the phage can enter into its lytic cycle and start the transcription of early genes. When repression of the Pe promoter by C protein occurs, the P2 genome instead is inserted into the host chromosome by site-specific recombination. Since the C repressor is not inactivated by the SOS/RecA system of E. coli, the prophage cannot be induced by ultraviolet irradiation. The mechanism of integration of P2-like phages is quite similar to λ site-specific recombination. P2-like phages are able to integrate into host chromosome with the use of a phage integrase protein, the histone-like protein, IHF, and both phage attP and bacterial attB attachment sites (Frumerie et al. 2005). P2 phage often integrates at a specific site. Recombination occurs between 27-bp-long sequence (core) and host chromosome. The transcriptional switch of temperate phages must be set in order to control lysogenization and lytic cycle after infection. In phage P2, promoters Pc and Pe are

132

J. Łoś et al.

facing each other. The Pc promoter directs transcription of the C repressor which is responsible for downregulation of the Pe promoter, active during lytic growth. On the contrary, the Pe promoter, which commands the lytic growth, directs transcription of the Cox protein resulting repression on the Pc promoter. The Pc transcript also encodes the integrase and the Pe promoter controls the expression of proteins needed for DNA replication. The decision between lytic and lysogenic life cycle appears to be the consequence of the relative concentration of the Cox protein and C repressor. C and Cox repressors are able to downregulate opposite promoters at relatively low concentrations, but also their own promoters, when proteins are present at high concentration, which prevents unnecessary buildup of these proteins (Saha et al. 1987). The common characteristics of P2-like phages is that two face-to-face-oriented promoters control the proper functions of cycles by directing the transcriptions of two transcripts, which partially overlap without any overlap in genes regions. Although the transcriptional switches of the P2-like phages have similar arrangements, they still can vary within the group. Phage HP1 has two early promoters, PR1 and PR2 (Esposito et al. 1997), and depending on promoter used the transcript overlap can be respectively 44 bp or 72 bp (Esposito et al. 1997). In phage 186 the overlap is about 60 bp and P2 transcripts have the shortest overlap of about 35 bp. Phage 186 has an additional gene, cII, that controls establishment of lysogeny but not its maintenance. This gene is located in early operon, encoding activator that acts on promoter PE (Lamont et al. 1993; Neufing et al. 2001). The phage Wϕ also contains two face-to-face promoters, but the repressors of Wϕ bind to two directly repeated operators, which differ in comparison to P2 operators (Liu and Haggård-Ljungquist 1999). Cox proteins of phages P2, P2 Hy dis, and WΦ have been shown to be multifunctional, since they not only act as repressors of Pc. They can also perform as directionality factors for site-specific recombination that inhibit integration and promote excision of phage genomes in or out of the host chromosome (Nilsson and Haggård-Ljungquist 2007).

Temperate Phages of Staphylococci and Streptococci Staphylococcal Phages Morphological Families and Classification The majority of described staphylococcal phages infect Staphylococcus aureus and were first used for the typing of clinical isolates. High interest in finding and characterizing S. aureus phages was caused by the fact that this common human pathogen is responsible for many nosocomial and community-acquired infections. Also, a growing number of antibiotic-resistant strains contributed to extensive studies of S. aureus phages. In this species, temperate bacteriophages appear to be widespread as every strain of S. aureus sequenced so far contained at least one prophage. Several studies report isolation of phages from coagulase-negative

Temperate Phages, Prophages, and Lysogeny

133

species, like Staphylococcus epidermidis, Staphylococcus hominis, or Staphylococcus saprophyticus. However, only small number of these phages have been sequenced, characterized, and studied. It is probably due to the fact that pathogenesis of coagulase-negative Staphylococci relies on factors required for their commensal mode of life instead of toxins that are often encoded by prophage genes, where instead the latter is for S. aureus. A majority of temperate staphylococcal phages belong to the Siphoviridae family in the Caudovirales order. They are composed of icosahedral capsid (morphotype B1, like coliphage λ) or prolate capsid (morphotype B2) and noncontractile tail with a baseplate structure with double-stranded DNA as a genetic material (Brüssow et al. 2004). Early classification of staphylococcal phages, both strictly lytic and temperate, was based on their lytic properties, morphology, and on genome size and organization. Based on classification proposed by Brüssow and Desiere, staphylococcal phages belonging to Siphoviridae were classified as Sfi21-like or Sfi11-like phages by some authors (Brüssow and Desiere 2001). Recently, the classification of staphylococcal Siphoviridae was updated based on phylogenetic relationships (Gutiérrez et al. 2014; Adriaenssens et al. 2018, 2020). Most of the total of 200 bacteriophages with completed genome sequences available in public databases were assigned to six genera. Classification of the remaining phages is an ongoing process. The Triavirus genus includes nearly 30 phages that infect S. aureus. All viruses belonged to morphotype B2. Proteomic analysis indicated presence of groupspecific proteins like A-type polymerase or unique capsid and tail proteins. RinA protein is also present instead of RinB homolog that is present in other phages (Gutiérrez et al. 2014). RinA and its homolog are proteins responsible for phagemediated packaging and transfer of virulence genes (Ferrer et al. 2011). Genus Biseptimavirus includes at least three S. aureus phages with a similar genome size and morphology corresponding with B1 morphotype. These phages share characteristics with the other genera, such as the presence of nucleases, similar to those from the “Triavirus” genus, and a common morphotype with “Phietavirus.” Phietavirus is a genus that includes at least 36 staphylococcal phages. Those phages show B1 morphotype (with an exception of EW phage), and they also show similar sequence of genes in tail morphogenesis module. Among peptidoglycan hydrolase domains, a CHAP domain and a glucosaminidase domain are present, instead of a lytic transglycosylase SLT domain and a peptidase_M23 domain, that were detected in other two genera (Gutiérrez et al. 2014). Six phages, referred by authors as “orphan phages,” remained unclassified as they lack clear homology to members of three proposed genera (Gutiérrez et al. 2014). However, the present method of classification is not recognized by ICTV (International Committee on Taxonomy of Viruses). Two other genera of staphylococcal Siphoviridae: Fibralongavirus, Sextaecvirus include 4 phages each. One staphylococcal siphovirus, namely SP-beta, which infects Staphylococcus epidermidis was asignet to Spbetavirus genus together with Bacillus phage Z.

134

J. Łoś et al.

Streptococcal Phages Morphological Families and Classification Most streptococcal bacteriophages known, as of this writing, were isolated from β-hemolytic Streptococci group A. It is speculated that around 90% of these streptococci may contain temperate phages (Hynes et al. 1995). Phages infecting Streptococcus pneumoniae were also described (Romero et al. 2009), but there are only a few known prophages of other Streptococcus species. Some have been found to carry virulence factors like antibiotic resistance or toxins, while others appear to have no effect on the phenotypes of their hosts. Furthermore, despite various studies, still little is known of molecular biology of streptococcal phages. Phages induced from lactic streptococci were described to have isomeric heads and noncontractile tails (Huggins and Sandine 1977), therefore matching Siphoviridae B1 morphotype. Phages isolated from various strains of Streptococcus pneumoniae showed the same morphotype. Experiments confirmed that genetic material of some known staphylococcal phages is ds-DNA (Romero et al. 2009). Most of the Streptococcus phages have not been classified yet. Based on classification methods proposed by Brüssow and Desiere, and on phylogenetic relationships there are three defined genera that include streptococcal phages: Brussowvirus (phages 2972, 858, ALQ132, O1205, Sfi11), Moineauvirus (phage DT1, DT1.1, DT1.2, DT1.3, DT1.4, DT1.5, phiAbc2, Sfi19 and Sfi21), and Saphexavirus (phage SPQS1) (Brüssow and Desiere 2001; Adriaenssens et al. 2018, 2020). The latter includes several Enterococcus phages in addition to a single Streptococcus phage).

Role of Prophages in Modulation of Bacterial Host Biology and Evolution Formerly, bacteriophages have been mostly perceived as parasites of bacterial cells as well as convenient tools to investigate the genetics of bacteria. However, the newest data of complete bacterial genome sequences has revealed the crucial role of prophages in the diversity of strains among bacterial species. Nowadays we can observed three main trends which have renewed the interest in phage research: (1) phage influence the cycling of organic matter in the oceans (chapter ▶ “Bacteriophage Ecology”), (2) they are potential tools for the treatment of antibioticresistant bacterial pathogens (section “Therapeutic Use of Bacteriophages”), and (3) they have major impact on bacterial short-term evolution (chapter ▶ “Bacteriophage-Mediated Horizontal Gene Transfer: Transduction”) (Canchaya et al. 2003). Previously discussed lysogenic conversion is only one of at least five different ways by which temperate phages affect bacterial fitness; however, they also can (1) serve as anchor points for host genome rearrangements, (2) disrupt host genes, (3) protect lysogens from lytic infection, (4) lyse competing strains following prophage induction, and (5) can introduce new fitness factors (lysogenic conversion, transduction) – (Table 2) (Brüssow et al. 2004). Mechanism of gene propagation, including virulence factor genes such as those encoding toxins, adhesins or

Temperate Phages, Prophages, and Lysogeny

135

aggressins, is critical for the emergence of new pathogenic strains. Recent studies suggest that a large amount of genetic information in natural environment and in bacterial genomes is of phage origin (Muniesa et al. 2011). On the other hand, phage integration leading to loss of function can be observed, e.g., in case of Staphylococcus aureus, in which L54a and φ13 prophages integrate into the chromosome and cause the inactivation of a lipase and a β toxin gene, respectively (Fortier and Sekulovic 2013), or phage Mu integration, which occurs in random places by transposition and may lead to inactivation of genes and modification of their expression (Harshey 2014). Some bacteria have evolved to exploit the presence of prophages for their own purposes. This usually occurs through crippling the prophage by deletion and then evolving prophage remains by the host for the new function. In Pseudomonas aeruginosa, two phage-tailed gene clusters have developed into bacteriocins (Nakayama et al. 2000). The defective Bacillus subtilis prophage PBSX has maintained the capacity to build a size-reduced phage head into which 13 kb fragments of random bacterial DNA are packaged turning it into gene transfer agent (Canchaya et al. 2003). Such use of defective prophages seems to be quite common in nature, as reviewed by Redfield and Soucy (2018). Perception of especially temperate bacteriophages thus has changed. They are now seen more as a “versatile carrier of genetic information within and between bacterial species and as a means of rearranging existing genetic information into unique combinations. Comparative bacterial genomics has revealed the ‘mutualistic’ role of bacteriophages in the evolution of bacterial pathogens” (Boyd and Brüssow 2002)

Toxins and Virulence Factors Encoded by Prophages Arguably the most studied aspect of prophages influencing host biology was their impact on formation of pathogens. Below, a few examples of this influence are discussed.

Staphylococcus aureus Most of virulence factors present in S. aureus were delivered by phages. Toxins and antibiotic resistance have been identified in other Staphylococcus species, although their role in pathogenesis remains unknown. In general, individual phages carry only a single virulence factor gene, although there are exceptions. S. aureus phage phiSa3 and its relatives, for example, may encode up to five virulence factors (Goerke et al. 2006). Virulence factor genes appear not to be strictly associated with one specific phage and most probably are exchanged by horizontal gene transfer and recombination. In most phages, virulent genes are located near phage attachment site (attP) and downstream the lysis module (Fig. 2). It is consequently speculated that they were obtained by phages by aberrant excision events from a bacterial chromosome (Wagner and Waldor 2002) (chapter ▶ “Bacteriophage-Mediated Horizontal Gene Transfer: Transduction”).

J. Łoś et al.

136

Table 2 Bacterial virulence factors dependent on phage (on the basis of Brüssow et al. 2004) Bacterial host C. botulinum C. diphtheriae E. coli

E. coli O157 M. arthritidis N. meningitidis P. aeruginosa P. multocida S. aureus

S. canis S. enterica

S. flexneri

S. mitis S. pyogenes

Protein Neurotoxin Phage C1 Shiga toxins Enterohaemolysin $FC3208 Cytolethal distending toxin OMPb OMP Superoxide dismutase Vir Membrane proteins Cytotoxin Mitogenic factor Enterotoxin Enterotoxin P Enterotoxin A Enterotoxin A Exfoliative toxin Leukocidin Leukocidin Staphylokinase Mitogenic factor Glucosylation Glucosylation Type III effector Type III effector Type III effector Superoxide dismutase Superoxide dismutase Neuraminidase Virulence factor Antivirulence factor O-antigen acetylase Glucosyl transferase Phage coat proteins Toxin type A Toxin type C

Gene C1 tox stx1, stx2 hly2

Phage Phage C1 β-Phage H-19B φFC3208

cdt

Unnamed

bor eib sodC

λ λ-like Sp4, 10

vir Mu-like ctx toxA see, sel sep entA sea eta pvl pvl sak Unnamed rfb gtr sopE sseI (gtgB) sspH1 sodC-I

MAV1 Pnm1 φCTX Unnamed NA φN315 φ13 φMu50A φETA fPVL φPVL φ13 phisc 1 ε34 P22 SopEΦ GIFSY-2 GIFSY-3 GIFSY-2

sodC-III

Fels-1

nanH gtgE grvA

Fels-1 GIFSY-2 GIFSY-2

oac gtrII

Sf6 SfII, SfV, SfX SM1 T12 CS112

pblA, pblB speA speC

(continued)

Temperate Phages, Prophages, and Lysogeny

137

Table 2 (continued) Bacterial host

Protein Superantigens Hyaluronidase Phospholipase DNase/ streptodornase Mitogenic factors

V. cholerae

Cholera toxin

Gene speA1, speA3, speC, speI, speH, speM, speL, speK, ssa hylP sla sdn, sda mf2, mf3, mf4

ctxAB

Phage 8232.1 H4489A 315.4 315.6, 8232.5 370.1, 370.3, 315.3 CTX$

Fig. 2 Organization of staphylococcal phages of Siphoviridae family as presented by Deghorian and Van Melderen (2012). Five functional modules arranged as follows: lysogeny (dark blue box), DNA metabolism (red box), DNA packaging and genes encoding capsid proteins (green box), genes encoding tail proteins (violet box), and cell host lysis genes (light blue box). If present, genes encoding virulence factors are localized downstream lysis genes or inserted between lysogeny and DNA metabolism genes

S. aureus is a leading cause of gastroenteritis, resulting from the absorption of staphylococcal enterotoxins after consumption of contaminated food (Le Loir et al. 2003; Łoś et al. 2010). S. aureus is also the leading cause of mammary gland infections in dairy animals. Specific antibiotic-resistant strains cause epidemics in hospital settings (Brüssow et al. 2004). Many skin infections such as furunculosis, staphylococcal scalded skin syndrome, and wound infections are caused by this bacterium. S. aureus can cause a wide range of diseases, ranging from toxicosis, such as food poisoning, to invasive diseases. Staphylococcus aureus strains encode a large variety of secreted toxins, and these toxins (Table 3) are responsible for most of the clinical symptoms associated with the infections. Among known S. aureus virulence factors carried by phages are staphylokinase, Panton-Valentine leukocidin (PVL), enterotoxin A, and exfoliative toxin A. Staphylokinase is a 136-aa-long protein carried by phage ɸ13. Its binding with host-produced plasminogen results in the formation of active plasmin, a proteolytic enzyme facilitating bacterial penetration into the surrounding tissues (Bokarewa et al. 2006). Panton-Valentine leukocidin (PVL) is a cytotoxin present in the majority of Methicillin-resistant Staphylococcus aureus (MRSA) and encoded in a prophage designated as ɸ-PVL. It produces two toxins, known as LukS-PV and LukF-PV that act together as subunits. PVL causes

138

J. Łoś et al.

leukocyte destruction and necrotic lesions of skin and mucosa. It is responsible for severe and often lethal necrotizing pneumonia (Lina et al. 1999). The staphylococcal exfoliative toxins (ETs) are extracellular proteins that cause splitting of human skin at the epidermal layer during infection in infants and causes blistering skin disease and staphylococcal scalded skin syndrome (SSSS) (Ladhani et al. 1999). These diseases affect mainly infants and children, with severity varying from localized blisters filled with fluid to general exfoliation that can affect entire body surface. Two antigenically distinct toxins possessing identical activity have been isolated from Staphylococcus aureus, ETA and ETB, which are serologically distinct (Kondo et al. 1975). The gene for ETA (eta) is located on the chromosome, whereas that for ETB is located on a large plasmid. Relatively few clinical isolates produce ETA which during early research suggested that eta gene is acquired by horizontal gene transfer. Indeed, exfoliative toxin A (ETA) gene is carried by ϕETA phage. ETA toxin is 242 aa long, has a molecular mass of 26,950 Da, and is heatstable. It consists of two domains, S1 and S2, each consisting of six-strand β-barrels and a C-terminal α-helix. It has serine protease-like properties, binds to the skin protein profilaggrin, and cleaves substrates after acidic residues (Yamaguchi et al. 2000). Temperate phage (ϕETA) that encodes ETA was isolated. ϕETA has a head with a hexagonal outline and a noncontractile and flexible tail. The genome of ΦETA is a circularly permuted linear double-stranded DNA, and the genome size is 43,081 bp. ΦETA converted ETA nonproducing strains into ETA producers. Southern blot analysis of chromosomal DNA from clinical isolates suggested that ΦETA or related phages are responsible for the acquisition of genes in S. aureus. Staphylococcal bacteriophages are also responsible for mobilization of Staphylococcus aureus pathogenicity islands (SaPIs). They are chromosomal DNA segments acquired by horizontal transfer that rely on a helper phage for moving (Tallent et al. 2007). SaPIs are known to carry various virulence factors such as gene encoding toxic shock syndrome toxin or variants of von Willebrand factor-binding protein that provides S. aureus with ability to coagulate host blood plasma (Lindsay et al. 1998). SaPIs can be replicated and mobilized as a response to SOS-induced excision of a helper prophage, by the infection of host cell by phage or by the joint entry of SaPI and a phage. SaPIs may require certain phage in order to be mobilized, although some phages as phage 80α are known to mobilize all known SaPIs (Łoś et al. 2010). Lysogenic conversion of staphylococci associated with expression of virulence factors was first reported in the early 1960s (Blair and Carr 1961; Winkler et al. 1965). Over 40 years ago discovered some S. aureus toxin: a phage could convert nontoxigenic strains to alpha-hemolysin production. The staphylococcal enterotoxin A gene, sea, was mapped near the attachment site of the temperate phage PS42-D (Betley and Mekalanos 1985; Brüssow et al. 2004). Southern hybridizations revealed that the sea genes in staphylococcal strains were associated with a family of phages rather than with one particular phage. Phage PVL encodes a bicomponent cytotoxin, the Panton-Valentine leukocidin. The two toxin genes lukS and lukF were located between the phage lysin gene and the right attachment site (Kaneko et al. 1998). The two toxins assemble into poreforming transmembrane complexes and lyse their target cells, human

Temperate Phages, Prophages, and Lysogeny

139

Table 3 Staphylococcus aureus virulence factors dependent on phage (on the basis of Helbin et al. 2012) Toxin/pathogenicity determinant (gene) Protein responsible for biofilm formation (bap) Von Willebrand factorbinding protein (vWBps) host-specific variant CHIPS, Chemotaxis Inhibiting Protein of S. aureus (chip)

Exfoliative toxin A (eta) Enterotoxin A (sea) Enterotoxin B (seb) Enterotoxin C (sec) Enterotoxin K (selk) Enterotoxin L (sell) Enterotoxin P (selp) Enterotoxin Q (selq) PVL, Panton-Valentine Leukocidin, bisubunit toxin (lukS and lukF)

SCIN, Staphylococcal complement inhibitor (scn)

Staphylokinase (sak)

Toxic shock syndrome Toxin-1, TSST-1 (tstH)

MGE SaPIbov

SaPIbov2, SaPIbov4, SaPIbov5, SaPIeq1, SaPIov2

φ13, φtp310–3, φ252B, φMu3A, φN315, φNM3, φSa3JH1, φSa3mw, φSa3 ms, φSa3JH9, φSa3USA300, φβCUSA300_TCH1516 ΦETA ΦSa3ms, ΦSa3mw, Φ252B, ΦNM3, ΦMu50a SaPI1, SaPI3 SaPIn1, SaPIm1, SaPImw2, SaPIbov1 ΦSa3ms, ΦSa3mw, SaPI1, SaPI3, SaPIbov1, SaPI5 SaPIn1, SaPIm1, SaPImw2, SaPIbov1 ΦN315, ΦMu3A ΦSa3ms, ΦSa3mw, SaPI1, SaPI3, SaPI5 ΦSA2pvl, ΦSLT, ΦPVL, ΦSA2MW, ΦSA2usa

φ13, φN315, φ252B, φNM3, φMu50A, φSa3JH1, φSa3 ms, φSa3mw, φSa3JH9, φMu3A, φtp310–3, φSa3USA300, φβC-USA300_TCH1516 phl3, ph42D, phφC, φN315, φMu50A

SaPI1, SaPI2, SaPIbov1, SaPI3

Description/mechanism/ symptoms Biofilm formation inside of cattle udder, involved in mastitis symptoms A variant of genomic vWBp protein, cattle, or equine plasma coagulation Decreases activity of C5AR1 and FPR1 neutrofile receptors relative to C5a complement system component and formylated peptides toxin responsible for staphylococcal scalded skin syndrome symptoms Staphylococcal enteritis Staphylococcal enteritis Staphylococcal enteritis Staphylococcal enteritis Staphylococcal enteritis Staphylococcal enteritis Staphylococcal enteritis Causes the lysis of mammalian leukocytes, stimulates overproduction of proinflammatory factors leading to tissue necrosis Intervenes in bacterial cells opsonization through the inhibition of complement system C3bBb convertase Disrupts the bacterial cells phagocytosis, a component of plasmin which helps the Staphylococcus cells to penetrate and proteolytically decay the tissues Main toxin responsible for TSS stimulates proinflammatory cytokines production by epithelial cells

140

J. Łoś et al.

polymorphonuclear leukocytes (Finck-Barbancon et al. 1993). Phosphorylation of LukS by protein kinase A was found to be required for the leukocytolytic activity (Muniesa et al. 2003; Brüssow et al. 2004). Very similar toxin genes were found at the same location in a morphologically and molecularly distinct S. aureus phage, SLT. S. aureus prophage PV83 also encodes a leukocidin, this time a lukM-lukF gene combination. The recent sequencing of several S. aureus strains confirmed and extended the observations from the phage-sequencing projects. Two phages are very similar between the two strains: N315 and Mu50A. The two prophages carry known virulence factors: a gene encoding enterotoxin P (the sep gene), a superantigen involved in the symptoms of food poisoning, and a gene encoding staphylokinase (the sak gene), suspected to be involved in the proteolytic destruction of host tissue. In addition, an M-like protein fragment is encoded by a gene preceding sep. The virulence genes flank the phage lysis cassette on both sides. However, the two prophages are not identical. Especially over the lysogeny and early genes, the two prophages differed in numerous small modular exchanges (Brüssow et al. 2004). During infection, bacterial pathogens encounter the serum and phagocytemediated elements of the innate immune system. Staphylococci produce a number of proteins involved in phagocyte evasion, including a recently discovered chemotaxis inhibitory protein (CHIPS) that binds to and attenuates the activity of the neutrophil receptors for complement and formylated peptides. This function is proposed to protect S. aureus from neutrophil-mediated killing an important host defense against staphylococci. The gene encoding CHIPS (chp) has been shown to reside on a functional phage that also transduces the staphylokinase (sak) and enterotoxin A (sea) genes and eliminates hemolysin production, presumably by insertional inactivation Staphylococci also produce the phage-encoded Panton-Valentine leukocidin (PVL), a cytotoxin with direct activity against human phagocytes. Thus, by inhibiting phagocytosis (CHIPS) and by directly attacking phagocytes (PVL), two different phage gene products counteract phagocyte-mediated destruction of their staphylococcal hosts (Wagner et al. 2002). S. aureus isolate, strain MW2, was sequenced (Baba et al. 2002). It differed from strain N315 by numerous insertions, deletions, and gene replacements. MW2 contains two prophages: Sa2 and Sa3. Sa2 resembles S. aureus phage 12 but also carries the lukS and lukF genes in a constellation identical to that phage in SLT. Sa3 closely resembles phage PVL over most of their genomes, but the two phages differed in their content of virulence genes. A comparison of the different S. aureus prophages revealed that the toxin genes are mobile DNA elements of their own and suggested that they are not stably associated with an individual prophage. Horizontal gene transfer has played a fundamental role not only in the evolution of S. aureus prophages but also in that of their hosts (Brüssow et al. 2004).

Clostridium botulinum These bacteria are strictly anaerobic gram-positive bacteria, which are ubiquitous in the environment. Clostridia produce extremely resistant spores which sporulate under anaerobic conditions. C. botulinum strains were originally defined by their

Temperate Phages, Prophages, and Lysogeny

141

ability to produce one of the closely related but antigenically distinct members (A, B, C1, D, E, F, or G) of the botulinum neurotoxin family. Human botulism is caused by the consumption of toxin-contaminated food. In other cases, the bacteria replicate within the human gut or sometimes in infected wounds, where they release the toxin in situ. Botulinum neurotoxins (BoNTXs) produced by Clostridium botulinum are among the most poisonous substances known. Of the seven types of BoNTXs, genes for type C1 and D toxins (BoNTX C1 and D) are carried by bacteriophages. The gene for exoenzyme C3 also resides on these phages (Sakaguchi et al. 2005). Each type of BoNTX is produced as a large polypeptide and converted to a di-chain molecule composed of L and H chains by bacterial or host proteases. The H chain is responsible for the binding of the toxin to the presynaptic membrane and for the translocation of the L chain into the cytosol. The botulinum neurotoxins are expressed as ca. 150-kDa precursors lacking classical signal peptides (Brüssow et al. 2004). Of the seven types of BoNTXs, genes for type C1 (BoNTX C1) and type D (BoNTX D) are carried by bacteriophages, which were discovered in the early 1970s. These phages are categorized into three groups according to their conversion spectra: phages from strains C-Stockholm (C-ST) and C-468, those from strains D-1873 and C-203, and those from strains D-South African and D-4947. These groups differ also in antigenicity, although they share a similar morphology. BoNTX phages were later found to encode exoenzyme C3, an ADP-ribosyltransferase of GTPases. It is also known that the lysogeny of BoNTX phages is unstable (Sakaguchi et al. 2005; Brüssow et al. 2004). The botulinum neurotoxins A, B, and F are encoded in the chromosome, while G is plasmid encoded (Brüssow et al. 2004; Zhou et al. 1993), and C1 and D are encoded by prophages. The C. botulinum lysogens can be cured easily, and cultures of the C1 and D toxin-producing strains release significant amounts of phage (Brüssow et al. 2004).

Vibrio cholerae Epidemics of cholera caused by toxigenic Vibrio cholerae belonging to the O1 or O139 serogroup are a major public health problem in many developing countries. The disease is an acute dehydrating diarrhea caused principally by the potent enterotoxin, cholera toxin (CT), produced by these organisms during pathogenesis (Faruque et al. 2001). Although V. cholerae is a human pathogen, aquatic ecosystems are major habitats of Vibrio species, which include both pathogenic and nonpathogenic strains that vary in their virulence gene content. Of the >100 known Vibrio serogroups, the two toxigenic serogroups “classical” O1 and O139 have been associated with epidemic cholera. The two human-pathogenic V. cholerae serogroups (O1 and O139) have evolved by sequential acquisition of two key fitness factors: the toxin-coregulated pilus (TCP) and cholera toxin (CT) (Hassan et al. 2010; Brüssow et al. 2004). Both are encoded by phages or phage-like elements (Waldor et al. 1997; Faruque et al. 2001; Brüssow et al. 2004). In toxigenic V. cholerae, CT is encoded by a filamentous bacteriophage designated CTXΦ, which exists as a prophage in the bacterial chromosome. CTXΦ phage genome encodes the functions necessary for a site-specific integration system and

142

J. Łoś et al.

thus can integrate into the V. cholerae chromosome at a specific attachment site known as attRS, forming stable lysogens. A typical CTXΦ genome has two regions, core and the RS2. The 4.6 kb core region encodes CT as well as the functions that are required for the virion morphogenesis, whereas the 2.5 kb RS2 region encodes the regulation, replication, and integration functions of the CTXΦ genome. The A and B subunits of CT are encoded by two separate overlapping open reading frames. DNA sequence analysis has shown that the RS2 region consists of three open reading frames (ORFs) including rstR, rstA, and rstB, and two intergenic regions ig1 and ig2 (Waldor et al. 1997; Faruque et al. 2001; Brüssow et al. 2004). CT is expressed in the host intestine as a classical AB toxin. The B subunit of CT binds to enterocytes and transports the catalytic A subunit into the host cell cytoplasm. There, the A subunit triggers signaling cascades leading to rapid chloride and water efflux into the intestinal lumen, causing watery diarrhea, the hallmark of epidemic cholera (Faruque et al. 2001; Brüssow et al. 2004). CT might enhance bacterial survival in the intestine. Studies have confirmed that some naturally occurring nontoxigenic strains of V. cholerae are infected by CTXΦ and converted to toxigenic strains with epidemic potential (Waldor et al. 1997). TCP is critical for intestinal colonization (Merrell et al. 2002; Brüssow et al. 2004). It is a type IV bundle-forming pilus, whose major subunit (TcpA) was identified in a screen for secreted virulence factors which are coregulated with CT. TCP is expressed in the human intestine and belongs to the major antigens in human infections. The genetic element encoding TCP (also termed VPI for “V. cholerae pathogenicity island”) has been described as the genome of a filamentous phage (VPIΦ or TCPΦ), but the phage nature has been disputed recently (Karaolis et al. 1999; Brüssow et al. 2004). Classical AB toxin occurs via type II secretion. The CT moron of CTX can be functional and provides a selective advantage only in vibrios. Only here are the proper regulators and transport systems available. Filamentous phage CTXΦ, which does not encode its own OM (outer membrane) pore, also requires one component of the eps system for its escape from the bacterium (Brüssow et al. 2004). The major pathogenic genes in V. cholerae are clustered in several regions of the V. cholerae chromosome and the structure of these pathogenic gene clusters indicates that these are capable of being propagated horizontally. The TCP pathogenicity island appears to be the initial genetic factor required for the origination of epidemic strains, since the cholera toxin-converting bacteriophage uses TCP as its receptor for infecting V. cholerae cells. Analysis of the structure of the TCP pathogenicity island suggests that this could be of phage origin or may be transferred by transducing phages (Faruque et al. 2001).

Corynebacterium diphtheriae C. diphtheriae is a strictly human-adapted Gram (+) bacterium. It can cause local infections of the tonsils, pharynx, nose, and conjunctiva and systemic intoxications when the released toxin destroys the parenchyma of the heart, liver, kidneys, or adrenal glands. The diphtheria toxin (DT) is the major virulence factor of this pathogen, and the DT gene is carried by a family of closely related bacteriophages. Diphtheria toxin (DT) is a classical AB toxin. The A subunit of DT is an ADP-ribosyltransferase which covalently modifies the elongation factor EF-2,

Temperate Phages, Prophages, and Lysogeny

143

thereby inhibiting chain elongation during protein synthesis (Brüssow et al. 2004; Zasada 2013). The symptoms are caused by diphtheria toxin (DT) encoded by the corynebacteriophage tox gene, the expression of which is downregulated by the chromosomally encoded diphtheria toxin repressor (DtxR) in an iron-dependent manner (Dinu et al. 2014; Zasada 2013). DtxR is a global metabolic regulator and binds to its DNA sequence targets as a homodimer after activation by divalent metal ions. DtxR is required for appropriate iron-dependent regulation of DT expression. Currently, at least 18 DtxR binding sites are known to occur in C. diphtheriae, and they affect the expression of about 40 genes. Studies conducted in the 1950s showed that non-lysogenic C. diphtheriae strains C4 and C7 become toxicogenic after infection with the tox+ corynephage beta but not with the tox-lacking corynephage gamma. C. diphtheriae phages have been poorly investigated. Most toxigenic C. diphtheriae strains contain DNA sequences related to phage beta, but the tox gene was also found to be associated with the distinct phages δ and ω (Dinu et al. 2014; Brüssow et al. 2004; Zasada 2013).

Streptococcus pyogenes S. pyogenes is a protean pathogen, and humans are its only reservoir. One-third of all humans are colonized with S. pyogenes. The bacteria are commonly found in the throat and on the skin. Streptococcus pyogenes is a multiply lysogenized organism whose phage constitutes 10% of the total genome and encodes a wide variety of putative and established virulence factors, including a large class of pyrogenic exotoxins (Broudy and Fischetti 2003). Recent comparative genomic studies have demonstrated that streptococcal bacteriophage represents the major variation (up to 71%) between strains of S. pyogenes and potentially account for the distinct disease pathologies associated with otherwise similar strains. In addition to modulating the virulence of organisms found within a common species of pathogenic bacteria, toxin-encoding phage produced by such pathogens have been shown to toxin convert both environmental and commensal bacteria, generating pathogenic Tox+ microbes. Thus, bacteriophage represents key vectors for the dissemination of bacterial virulence and the conversion of bacteria from nonpathogenic to pathogenic (Broudy and Fischetti 2003; Brüssow et al. 2004). Various Streptococcus virulence factors are phage encoded. One of the best known is erythrogenic toxin A carried by phage T12. T12 is a prototypic temperate phage of group A streptococci that infects Streptococcus pyogenes, converting harmless strains into virulent ones (McShan et al. 1997). T12 carries a gene that encodes erythrogenic toxin A (SPE-A), also known as scarlet fever toxin A. The T12 genome is circular with total length of 36 kb. It is known that T12 integrates into S. pyogenes chromosome by site-specific recombination into the anticodon loop of a gene that encodes serine tRNA. The phage integrase gene (int) and the phage attachment site (attP) are located upstream of the speA gene in the phage genome. The bacterial attachment (attB) site is located at the 30 end of the tRNA gene and has a sequence homologous to the phage attachment site. The coding sequence of the tRNA gene remains intact after integration of the prophage (McShan et al. 1997). The SPE-A toxin is known to damage plasma membranes of blood capillary endothelial cells found under the skin that

144

J. Łoś et al.

results in a red skin rash. Strains that produce SPE-A toxin are responsible for diseases like scarlet fever and streptococcal toxic shock syndrome (STSS). Another well-known phage-encoded virulence factor is hyaluronidase encoded by bacteriophage-carried genes, hylP and hylP2. A single Streptococcus pyogenes strain can carry one of these genes or both. Streptococcal hyaluronidase is used as a spreading factor due to its ability to attack the hyaluronic acid present around host cells as a cementing substance. The hyaluronidase (HylP) carried by phage H4489A contains a series of 10 Gly-X-Y amino acid triplets, closely resembling the repeating sequences found in collagen. Since the bacteriophage hyaluronidase is found in isolates from patients suffering with rheumatic fever, it is possible that this collagenlike repeat could lead to the induction of antibodies which may cross-react with tissue collagen and result in a disease (Hynes et al. 1995). Furthermore, some Streptococcus temperate phages are known to transfer resistance for antibiotics such as resistance to tetracycline, chloramphenicol, macrolides, and streptomycin most probably via generalized transduction (Wagner and Waldor 2002). Also, proteins PblA and PblB encoded by SM1 phage enable Streptococcus mitis to bind to platelets more efficiently (Bensing et al. 2001a). Those proteins are parts of the phage particle. However, they act as S. mitis surface proteins if SM1 is integrated into the bacterial genome. While not directly responsible for pathogenesis by this bacteria species, presence of this prophage in its genome facilitates colonization of a host. The exact mechanism of this phenomenon is still unknown (Bensing et al. 2001b).

Conclusions Prophages are the essential part of the microbial life. They are able to colonize a host using sophisticated mechanisms, keep their presence in the host sometimes using even more sophisticated solutions, and contribute to the life together by fine-tuning host metabolism and by offering to a host a new ability. This synergy allows hosts to gain new abilities, colonize new ecological niches, or be more efficient, more aggressive, or just more versatile in those, which are already occupied by host. Sometimes, a one prophage too far, superbugs are being created. Even though, on the basis of literature surveys, they seem to be one of the dominant examples of phagehost synergy, it is very likely that they just attract our attention due to the urgent need to understand and counteract their ability to cause disease. They most probably constitute very small, but well-visible minority. The lysogenic conversion they provide to the host may show a wide range of different effects. The deadly payload they can deliver to augment host virulence is really impressive. These range from immune system evasion genes like bor in prophage λ (Barondess and Beckwith 1995) through enzymes capable of decomposition of body components facilitating invasion and acquisition of nutrients (e.g., Hynes et al. 1995) to the most-deadly poisons known to mankind (Brüssow et al. 2004; Sakaguchi et al. 2005). Prophages were for a long time very useful model organisms for studying basic molecular biology mechanisms. As our knowledge progressed, it appeared that the

Temperate Phages, Prophages, and Lysogeny

145

mechanisms being studied are very complicated and sublimed, and the first, simplistic models just scratched the surface of the real complexity of these processes. Such processes like lysis versus lysogeny decisions evolved to be complicated and to take into consideration many different aspects of host physiology (e.g., Słominska et al. 1999). Now it appears that prophages can communicate with their own kind and influence the decisions made by phages infecting a host (Erez et al. 2017; Abedon 2017). Once such a mechanism was discovered, it is only a matter of time before there are many more examples of similar ways of communication influencing the most important decision in temperate prophage life and the life of its host. The fact that prophage interaction with the host may be important for both has been known for a long time (e.g., Dykhuizen et al. 1978), but the knowledge of how widespread this interaction is in nature is a relatively new discovery (e.g., Ventura et al. 2003a, b). Now in a metagenomic era, it is certain new discoveries are waiting for us on this subject. Acknowledgments This work was partially financed by the Ministry of Science and Higher Education, Poland (Grant No. 0312/IP1/2011/71) and National Science Centre, Poland awarded on the basis of grant decision no. DEC-2011/01/B/NZ1/04404.

References Abedon ST (2017) Commentary: communication between viruses guides lysis-lysogeny decisions. Front Microbiol 8:983 Ackermann HW (1999) Tailed bacteriophages: the order Caudovirales. Adv Virus Res 51:135–201 Adriaenssens EM, Wittmann J, Kuhn JH, et al (2018) Taxonomy of prokaryotic viruses: 2017 update from the ICTV Bacterial and Archaeal Viruses Subcommittee. Arch Virol 163(4):1125– 1129. https://doi.org/10.1007/s00705-018-3723-z Adriaenssens EM, Sullivan MB, Knezevic P, et al (2020) Taxonomy of prokaryotic viruses: 2018–2019 update from the ICTV Bacterial and Archaeal Viruses Subcommittee. Arch Virol 165(5):1253–1260. https://doi.org/10.1007/s00705-020-04577-8 Altuvia S, Kornitzer D, Teff D, Oppenheim AB (1989) Alternative mRNA structures of the cIII gene of bacteriophage lambda determine the rate of its translation initiation. J Mol Biol 210:265–280 Avlund M, Dodd IB, Semsey S, Sneppen K, Sandeep KS (2009) Why do phage play dice? J Virol 83:11416–11420 Baba T, Takeuchi F, Kuroda M, Yuzawa H, Aoki K, Oguchi A, Nagai Y, Iwama N, Asano K, Naimi T, Kuroda H, Cui L, Yamamoto K, Hiramatsu K (2002) Genome and virulence determinants of high virulence community-acquired MRSA. Lancet 359:1819–1827 Barksdale L, Arden SB (1974) Persisting bacteriophage infections, lysogeny, and phage conversion. Annu Rev Microbiol 28:265–299 Barondess JJ, Beckwith J (1995) bor gene of phage lambda, involved in serum resistance, encodes a widely conserved outer membrane lipoprotein. J Bacteriol 177:1247–1253 Bensing BA, Siboo IR, Sullam PM (2001a) Proteins PblA and PblB of Streptococcus mitis, which promote binding to human platelets, are encoded within a lysogenic bacteriophage. Infect Immun 69(10):6186–6192. https://doi.org/10.1128/IAI.69.10.6186-6192.2001 Bensing BA, Rubens CE, Sullam PM (2001b) genetic loci of streptococcus mitis that mediate binding to human platelets. Infect Immun 69(3):1373–1380. https://doi.org/10.1128/IAI.69.3. 1373-1380.2001

146

J. Łoś et al.

Bertani G (1951) Studies on lysogenesis I. The mode of phage liberation by lysogenic Escherichia coli. J Bacteriol 62:293–300 Bertani E, Six EW (1988) The P2-like phages and their parasite, P4. In: Calendar R (ed) The bacteriophages. Plenum Publishing Corporation, New York/London Betley MJ, Mekalanos JJ (1985) Staphylococcus enterotoxin A is encoded by a phage. Science 229:185–187 Blair JE, Carr M (1961) Lysogeny in staphylococci. J Bacteriol 82:984–993 Bokarewa MI, Jin T, Tarkowski A (2006) Staphylococcus aureus: Staphylokinase. Int J Biochem Cell Biol 38(4):504–509. https://doi.org/10.1016/j.biocel.2005.07.005 Boyd EF, Brüssow H (2002) Common themes among bacteriophage-encoded virulence factors and diversity among the bacteriophages involved. TRENDS Microbiol 10(11):521–529 Broudy TB, Fischetti VA (2003) In vivo lysogenic conversion of Tox Streptococcus pyogenes to Tox+ with lysogenic streptococci or free phage. Infect Immun 71:3782–3786 Brüssow H, Desiere F (2001) Comparative phage genomics and the evolution of Siphoviridae: insights from dairy phages. Mol Microbiol 39:213–223 Brüssow H, Canchaya C, Hardt WD (2004) Phages and the evolution of bacterial pathogens: from genomic rearrangements to lysogenic conversion. Microbiol Mol Biol Rev 68:560–602 Campbell AM (1998) Prophages and cryptic prophages. In: de Bruijn FJ, Lupski JR, Weinstock GM (eds) Bacterial genomes. Springer, Boston Canchaya C, Fournous G, Chibani-Chennoufi S, Dillmann ML, Brüssow H (2003) Phage as agents of lateral gene transfer. Curr Opin Microbiol 6:417–424 Casjens S, Hendrix RW (2015) Bacteriophage lambda: early pioneer and still relevant. Virology 479–480:310–330 Chattoraj DK (2000) Control of plasmid DNA replication by iterons: no longer paradoxical. Mol Microbiol 37:467–476 Chattoraj DK, Inman RB (1974) Location of DNA ends in P2, 186, P4 and lambda bacteriophage heads. J Mol Biol 87(1):11–22 Chen Y, Golding I, Sawai S, Guo L, Cox EC (2005) Population fitness and the regulation of Escherichia coli genes by bacterial viruses. PLoS Biol 3(7):e229 Cruz JW, Rothenbacher FP, Maehigashi T, Lane WS, Dunham CM, Woychik NA (2014) Doc toxin is a kinase that inactivates elongation factor Tu. J Biol Chem 289:7788–7798 Deghorain M, Van Melderen L (2012) The Staphylococci phages family: an overview. Viruses 4(12):3316–3335. https://doi.org/10.3390/v4123316 Dinu S, Damian M, Badell E, Dragomirescu CC, Guiso N (2014) New diphtheria toxin repressor types depicted in a Romanian collection of Corynebacterium diphtheriae isolates. J Basic Microbiol 54:1136–1139 Dokland T, Lindqvist BH, Fuller SD (1992) Image reconstruction from cryo-electron micrographs reveals the morphopoietic mechanism in the P2^P4 bacteriophage system. EMBO J 11:839–846 Dykhuizen D, Campbell JH, Rolfe BG (1978) The influences of a lambda prophage on the growth rate of Escherichia coli. Microbios 23(92):99–113 Edgar R, Rokney A, Feeney M et al (2008) Bacteriophage infection is targeted to cellular poles. Mol Microbiol 68(5):1107–1116. https://doi.org/10.1111/j.1365-2958.2008.06205.x Erez Z, Steinberger-Levy I, Shamir M, Doron S, Stokar-Avihail A, Peleg Y, Melamed S, Leavitt A, Savidor A, Albeck S, Amitai G, Sorek R (2017) Communication between viruses guides lysislysogeny decisions. Nature 541:488–493 Esposito D, Wilson JC, Scocca JJ (1997) Reciprocal regulation of the early promoter region of bacteriophage HP1 by the Cox and Cl proteins. Virology 234(2):267–276 Faruque SM, Rahman MM, Hasan AK, Nair GB, Mekalanos JJ, Sack DA (2001) Diminished diarrheal response to Vibrio cholerae strains carrying the replicative form of the CTX(Phi) genome instead of CTX(Phi) lysogens in adult rabbits. Infect Immun 69(10):6084–6090. https:// doi.org/10.1128/IAI.69.10.6084-6090.2001 Faruque SM, Nair GB (2002) Molecular ecology of toxigenic Vibrio cholerae. Microbiol Immunol 46(2):59–66

Temperate Phages, Prophages, and Lysogeny

147

Ferrer MD, Quiles-Puchalt N, Harwich MD, Tormo-Más MÁ, Campoy S, Barbé J, Lasa I, Novick RP, Christie GE, Penadés, JR (2011) RinA controls phage-mediated packaging and transfer of virulence genes in Gram-positive bacteria. Nucleic Acids Res 39(14):5866–5878. https://doi. org/10.1093/nar/gkr158 Finck-Barbancon V, Duportail G, Meunier O, Colin DA (1993) Pore formation by a two-component leukocidin from Staphylococcus aureus within the membrane of human polymorphonuclear leukocytes. Biochim Biophys Acta 1182:275–282 Fogg PC, Allison HE, Saunders JR, McCarthy AJ (2010) Bacteriophage lambda: a paradigm revisited. J Virol 84:6876–6879 Fortier LC, Sekulovic O (2013) Importance of prophages to evolution and virulence of bacterial pathogens. Virulence 4(5):354–365 Freeman VJ (1951) Studies on the virulence of bacteriophage-infected strains of Corynebacterium diphtheria. J Bacteriol 61:675–688 Frumerie C, Sylwan L, Ahlgren-Berg A, Haggård-Ljungquist E (2005) Cooperative interactions between bacteriophage P2 integrase and its accessory factors IHF and Cox. Virology 332:284–294 Funnell BE, Slavcev RA (2004) Partition systems of bacterial plasmids. In: Funnell BE, Phillips GJ (eds) Plasmid biology. ASM Press, Washington, DC, pp 81–104 Giladi H, Goldenberg D, Koby S, Oppenheim AB (1995) Enhanced activity of the bacteriophage l pL promoter at low temperature. Proc Natl Acad Sci U S A 92:2184–2188 Goerke C, Wirtz C, Flückiger U, Wolz C (2006) Extensive phage dynamics in Staphylococcus aureus contributes to adaptation to the human host during infection. Mol Microbiol 61:1673– 1685. https://doi.org/10.1111/j.1365-2958.2006.05354.x Golais F, Hollý J, Vítkovská J (2013) Coevolution of bacteria and their viruses. Folia Microbiol 58:177–186 Goshorn SC, Schlievert PM (1989) Bacteriophage association of streptococcal pyrogenic exotoxin type C. J Bacteriol 171:3068–3073 Grose JH, Casjens SR (2014) Understanding the enormous diversity of bacteriophages: the tailed phages that infect the bacterial family Enterobacteriaceae. Virology 468–470:421–443 Gruenig MC, Lu D, Won SJ, Dulberger CL, Manlick AJ, Keck JL, Cox MM (2011) Creating directed double-strand breaks with the Ref protein: a novel RecA-dependent nuclease from bacteriophage P1. J Biol Chem 286:8240–8251 Gutiérrez D, Adriaenssens EM, Martínez B et al (2014) Three proposed new bacteriophage genera of staphylococcal phages: “3alikevirus”, “77likevirus” and “Phietalikevirus”. Arch Virol 159:389–398. https://doi.org/10.1007/s00705-013-1833-1 Harshey RM (2014) Transposable phage Mu. Microbiol Spectr 2(5). https://doi.org/10.1128/microbiolspec.MDNA3-0007-2014 Hassan F, Kamruzzaman M, Mekalanos JJ, Faruque SM (2010) Satellite phage TLCΦ enables toxigenic conversion by CTX phage through dif site alteration. Nature 467:982–985 Helbin W, Polakowska K, Mie¸dzobrodzki J (2012) Phage-related virulence factors of Staphylococcus aureus. Postepy Mikrobiologii 51:291–298 Hendrix RW (2002) Bacteriophage λ and its relatives. In: Streips UN, Yasbin RE (eds) Modern microbial genetics, 2nd edn. Wiley-Liss. https://doi.org/10.1002/047122197X.ch5 Herman C, Thevenet D, D’Ari R, Bouloc P (1997) The HflB protease of Escherichia coli degrades its inhibitor λ cIII. J Bacteriol 179:358–363 Hoopes BC, McClure WR (1985) A cII-dependent promoter is located within the Q gene of bacteriophage lambda. Proc Natl Acad Sci U S A 82:3134–3138 Howard-Varona C, Hargreaves KR, Abedon ST, Sullivan MB (2017) Lysogeny in nature: mechanisms, impact and ecology of temperate phages. ISME J 11:1511–1520 Huang A, Friesen J, Brunton JL (1987) Characterization of a bacteriophage that carries the genes for production of Shiga-like toxin 1 in Escherichia coli. J Bacteriol 169:4308–4312 Huggins AR, Sandine WE (1977) Incidence and properties of temperate bacteriophages induced from lactic streptococci. Appl Environ Microb 33:184–191 Hynes WL, Hancock L, Ferretti JJ (1995) Analysis of a second bacteriophage hyaluronidase gene from Streptococcus pyogenes: evidence for a third hyaluronidase involved in extracellular

148

J. Łoś et al.

enzymatic activity. Infect Immun 63(8):3015–3020. https://doi.org/10.1128/IAI.63.8.30153020.1995 Kaneko J, Kimura T, Narita S, Tomita T, Kamio Y (1998) Complete nucleotide sequence and molecular characterization of the temperate staphylococcal bacteriophage PVL carrying PantonValentine leukocidin genes. Gene 215(1):57–67 Karaolis DK, Somara S, Maneval DR Jr, Johnson JA, Kaper JB (1999) A bacteriophage encoding a pathogenicity island, a type-IV pilus and a phage receptor in cholera bacteria. Nature 399:375–379 Kim EJ, Lee D, Moon SH, Lee CH, Kim DW (2014) CTX prophages in Vibrio cholerae O1 strains J. Microbiol Biotechnol 24:725–731 Kim EJ, Yu HJ, Lee JH, Kim JO, Han SH, Yun CH, Chun J, Nair GB, Kim DW (2017) Replication of Vibrio cholerae classical CTX phage. Proc Natl Acad Sci U S A 114:2343–2348 Kondo I, Sakurai S, Sarai Y, Futaki S (1975) Two serotypes of exfoliatin and their distribution in staphylococcal strains isolated from patients with scalded skin syndrome. J Clin Microbiol 1:397–400 Ladhani S, Joannou CL, Lochrie DP, Evans RW, Poston SM (1999) Clinical, microbial, and biochemical aspects of the exfoliative toxins causing staphylococcal scalded-skin syndrome. Clin Microbiol Rev 12:224–242 Lamont I, Richardson H, Carter DR, Egan JB (1993) Genes for the establishment and maintenance of lysogeny by the temperate coliphage 186. J Bacteriol 175:5286–5288 Le Loir Y, Baron F, Gautier M (2003) Staphylococcus aureus and food poisoning. Genet Mol Res 2:63–76 Lehnherr H (2006) Bacteriophage P1. In: Calender R (ed) The bacteriophages. Oxford University Press, New York, pp 350–364 Lina G, Piémont Y, Godail-Gamot F, Bes M, Peter M-O, Gauduchon V, Vandenesch F, Etienne J (1999) Involvement of panton-valentine leukocidin–producing staphylococcus aureus in primary skin infections and pneumonia. Clin Infect Dis 29(5):1128–1132. https://doi.org/10.1086/ 313461 Linderoth NA, Ziermann R, Haggard-Ljungquist E, Christie GE, Calendar R (1991) Nucleotide sequence of the DNA packaging and capsid synthesis genes of bacteriophage P2. Nucleic Acids Res 19(25):7207–7214 Lindsay JA, Ruzin A, Ross HF, Kurepina N, Novick RP (1998) The gene for toxic shock toxin is carried by a family of mobile pathogenicity islands in Staphylococcus aureus. Mol Microbiol 29(2):527–543. https://doi.org/10.1046/j.1365-2958.1998.00947.x Liu T, Haggård-Ljungquist E (1999) The transcriptional switch of bacteriophage Wphi, a P2 related but heteroimmune coliphage. J Virol 73:9816–9826 Łobocka M, Rose DJ, Plunkett G, Rusin M, Samojedny A, Lehnherr H, Yermolinsky MB, Blattner FR (2004) Genome of bacterophage P1. J Bacteriol 186:7032–7068 Łoś M, Kuzio J, McConell MR, Kropinski AM, Węgrzyn G, Christie GE (2010) Lysogenic conversion in bacteria of importance to the food industry. In: Bacteriophages. ASM Press, Washington, DC, pp 157–198 Magnuson R, Yarmolinsky MB (1998) Corepression of the P1 addiction operon by Phd and Doc. J Bacteriol 186:6342–6351 McCabe BC, Pawlowski DR, Koudelka GB (2005) The bacteriophage 434 repressor dimer preferentially undergoes autoproteolysis by an intramolecular mechanism. J Bacteriol 187:5624–5630 McShan WM, Tang Y-F, Ferretti JJ (1997) Bacteriophage T12 of Streptococcus pyogenes integrates into the gene encoding a serine tRNA. Mol Microbiol 23:719–728. https://doi.org/10.1046/ j.1365-2958.1997.2591616.x Merrell DS, Hava DL, Camilli A (2002) Identification of novel factors involved in colonization and acid tolerance of Vibrio cholerae. Mol Microbiol 43:1471–1491 Muniesa M, de Simon M, Prats G, Ferrer D, Panella H, Jofre J (2003) Shiga toxin 2-converting bacteriophages associated with clonal variability in Escherichia coli O157:H7 strains of human origin isolated from a single outbreak. Infect Immun 71:4554–4562

Temperate Phages, Prophages, and Lysogeny

149

Muniesa M, Imamovic L, Jofre J (2011) Bacteriophages and genetic mobilization in sewage and faecally polluted environments. Microb Biotechnol 4(6):725–734 Nakayama K, Kanaya S, Ohnishi M, Terawaki Y, Hayashi T (1999) The complete nucleotide sequence of φCTX, a cytotoxin-converting phage of Pseudomonas aeruginosa: impications for phage evolution and horizontal gene transfer via bacteriophages. Mol Microbiol 31:399–419 Nakayama K, Takashima K, Ishihara H, Shinomiya T, Kageyama M, Kanaya S, Ohnishi M, Murata T, Mori H, Hayashi T (2000) The R-type pyocin of Pseudomonas aeruginosa is related to P2 phage, and the F-type is related to lambda phage. Mol Microbiol 38:213–231 Neufing PJ, Shearwin KE, Egan JB (2001) Establishing lysogenic transcription in the temperate coliphage 186. J Bacteriol 183:2376–2379 Nilsson AS, Haggård-Ljungquist E (2007) Evolution of P2-like phages and their impact on bacterial evolution. Res Microbiol 158:311–317 Nilsson AS, Karlsson JL, Haggård-Ljungquist E (2004) Site-specific recombination links the evolution of P2-like coliphages and pathogenic enterobacteria. Mol Biol Evol 21(1):1–13 O’Brien AD, Newland JW, Miller SF, Holmes RK, Smith HW, Formal SB (1984) Shiga-like toxinconverting phages from Escherichia coli strains that cause hemorrhagic colitis or infantile diarrhea. Science 226:694–696 Ochman H, Selander RK (1984) Standard reference strains of Escherichia coli from natural populations. J Bacteriol 157(2):690–693 Potrykus K, Węgrzyn G, Hernandez VJ (2002) Multiple mechanisms of transcription inhibition by ppGpp at the lambda pR promoter. J Biol Chem 277:43785–43791 Ptashne M (2004) A genetic switch: phage lambda revisited, 3rd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor Ranquet C, Ariane TA, de Jong H, Maenhaut-Michel G, Geiselmann J (2005) Control of bacteriophage Mu lysogenic repression. J Mol Biol 353:186–195 Redfield RJ, Soucy SM (2018) Evolution of bacterial gene transfer agents. Front Microbiol 9:2527 Rolfe B (1970) Lambda phage transduction of the bio A locus of Escherichia coli. Virology 42:643–661 Romero P, Croucher NJ, Hiller NL, Hu FZ, Ehrlich GD, Bentley SD, García E, Mitchell TJ (2009) Comparative genomic analysis of ten Streptococcus pneumoniae temperate bacteriophages. J Bacteriol 191(15):4854–4862. https://doi.org/10.1128/JB.01272-08 Rotman E, Amado L, Kuzminov A (2010) Unauthorized horizontal spread in the laboratory environment: the tactics of Lula, a temperate lambdoid bacteriophage of Escherichia coli. PLoS One 5(6):e11106 Saha S, Haggård-Ljungquist E, Nordström K (1987) The Cox protein of bacteriophage P2 inhibits the formation of the repressor protein and autoregulates the early operon. EMBO J 6(10):3191–3199 Sakaguchi Y, Hayashi T, Kurokawa K, Nakayama K, Oshima K, Fujinaga Y, Ohnishi M, Ohtsubo E, Hattori M, Oguma K (2005) The genome sequence of Clostridium botulinum type C neurotoxin-converting phage and the molecular mechanisms of unstable lysogeny. Proc Natl Acad Sci USA 102:17472–17477 Shotland Y, Koby S, Teff D, Mansur N, Oren DA, Tatematsu K, Tomoyasu T, Kessel M, Bukau B, Ogura T, Oppenheim AB (1997) Proteolysis of the l CII regulatory protein by FtsH (HflB) of Escherichia coli. Mol Microbiol 24:1303–1310 Shotland Y, Shifrin A, Ziv T, Teff D, Koby S, Kobiler O, Oppenheim AB (2000) Proteolysis of bacteriophage λ CII by Escherichia coli FtsH (HflB). J Bacteriol 182:3111–3116 Słominska M, Neubauer P, Wegrzyn G (1999) Regulation of bacteriophage λ development by guanosine 50 -diphosphate-30 -diphosphate. Virology 262:431–441 Smith DL, Rooks DJ, Fogg PCM, Darby AC, Thomson NR, McCarthy AJ, Allison HE (2012) Comparative genomics of Shiga toxin encoding bacteriophages. BMC Genom 13(1):311 Szalewska-Palasz A, Wróbel B, Wegrzyn G (1998) Rapid degradation of polyadenylated oop RNA. FEBS Lett 432:70–72 Tallent SM, Langston TB, Moran RG, Christie GE (2007) Transducing particles of staphylococcus aureus pathogenicity Island SaPI1 are comprised of helper phage-encoded proteins. J Bacteriol 189:7520–7524

150

J. Łoś et al.

Ventura M, Bruttin A, Canchaya C, Brüssow H (2002) Transcription analysis of Streptococcus thermophilus phages in the lysogenic state. Virology 302:21–32 Ventura M, Canchaya C, Pridmore D, Berger B, Brussow H (2003a) Integration and distribution of Lactobacillus johnsonii prophages. J Bacteriol 185:4603–4608 Ventura M, Canchaya C, Kleerebezem M, de Vos WM, Siezen RJ, Brüssow H (2003b) The prophage sequences of Lactobacillus plantarum strain WCFS1. Virology 316:245–255 Wagner PL, Waldor MK (2002) Bacteriophage control of bacterial virulence. Infect Immun 70:3985–3993 Wagner PL, Livny J, Neely MN, Acheson DW, Friedman DI, Waldor MK (2002) Bacteriophage control of Shiga toxin 1 production and release by Escherichia coli. Mol Microbiol 44:957–970 Waldor MK, Rubin EJ, Pearson GDN, Kimsey H, Mekalanos JJ (1997) Regulation, replication, and integration functions of the Vibrio cholerae CTXΦ are encoded by regions RS2. Mol Microbiol 24:917–926 Walker DH, Anderson TF (1970) Morphological variants of coliphage P1. J Virol 5:765–782 Weeks CR, Ferretti JJ (1984) The gene for type A streptococcal exotoxin (erythrogenic toxin) is located in bacteriophage T12. Infect Immun 46:531–536 Węgrzyn G, Węgrzyn A (2005) Genetic switches during bacteriophage l development. Progr Nucleic Acid Res Mol Biol 79:1–48 Winkler KC, de Waart WJ, Grootsen C (1965) Lysogenic conversion of staphylococci to loss of β-toxin. J Gen Microbiol 39:321–333 Yamaguchi T, Hayashi T, Takami H, Nakasone K, Ohnishi M, Nakayama K, Yamada S, Komatsuzawa H, Sugai M (2000) Phage conversion of exfoliative toxin A production in Staphylococcus aureus. Mol Microbiol 38:694–705 Yarmolinsky MB (2004) Bacteriophage P1 in retrospect and in prospect. J Bacteriol 186:7025– 7028 Yarmolinsky M, Hoess R (2015) The legacy of Nat Sternberg: the genesis of Cre-lox technology. Annu Rev Virol 2:25–40 Yarmolinsky M, Sternberg N (1988) Bacteriophage Pl. In: Calendar R (ed) The bacteriophages, vol 1. Plenum Press, New York, pp 291–438 Zasada AA (2013) Występowanie i chorobotwórczość dla człowieka potencjalnie toksynotwórczych maczugowców- Corynebacterium diphtheriae, Corynebacterium ulcerans i Corynebacterium pseudotuberculosis. Post Mikrobiol 52:201–209 Zhou Y, Sugiyama H, Johnson EA (1993) Transfer of neurotoxigenicity from Clostridium butyricum to a nontoxigenic Clostridium botulinum type E-like strain. Appl Environ Microbiol 59:3825–3831

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction Christine L. Schneider

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview of Transduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Historical Perspectives on Horizontal Gene Transfer in Bacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Discovery of Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Discovery of Conjugation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Discovery of Transduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Examples of Transducing Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bacteriophage P1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bacteriophage P22 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lambda (λ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Features Important for Transduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Host DNA Degradation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DNA Packaging Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mobile Genetic Elements Within Phage Genomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Integration Site Specificity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Examples of Transduction Hijackers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gene Transfer Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage-Inducible Chromosomal Islands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods to Detect and Characterize Transducing Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phenotypic Changes of the Bacterial Host . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phenotypic Changes of the Transducing Particle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genotypic Changes of the Bacterial Host . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genotypic Changes of the Transducing Particle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Potential Risks to Human Health Posed by Transduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transduction and the Spread of Antibiotic Resistance Genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transduction and the Implications to Phage Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

152 153 155 156 156 157 159 159 162 163 168 169 170 170 171 172 172 172 173 175 175 176 176 177 179 180 181

C. L. Schneider (*) Carroll University, Waukesha, WI, USA e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_4

151

152

C. L. Schneider

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

Abstract

Horizontal gene transfer (HGT) can create tremendous genetic diversity in organisms, especially those which reproduce asexually. The three classical mechanisms for HGT in prokaryotes are transformation, conjugation, and transduction. Historically, conjugation and transformation were considered to be the major contributors to bacterial HGT, but recent studies indicate that the role of transduction has been underestimated. Bacteriophages are the most abundant biological entities on the planet and contain a vast supply of genetic diversity. Whole genome sequencing of bacterial species has demonstrated that many bacterial virulence factors are encoded by phage or phage-like elements and that often prophage sequences are the major source of variation between bacterial strains. Bacteriophage-mediated HGT occurs through either generalized or specialized transduction. Generalized transducing particles are generated by aberrant packaging of host DNA in place of the viral genome and have the potential to transfer large blocks of bacterial DNA in a single particle. Specialized transducing particles contain a hybrid DNA molecule consisting of viral genes in combination with a small fragment of bacterial DNA. The hybrid DNA molecule associated with these particles is generated by excision from the host genome of a prophage that had been integrated during its lysogenic cycle. This chapter reviews some of the history behind the discovery of transduction, provides an in-depth view of some classic examples of transducing phages, and examines the continued impact of bacteriophage-mediated HGT on current issues such as antibiotic resistance and the safety of phage therapy.

Introduction The estimated 2.9
1029 bacteria that inhabit the Earth (Kallmeyer et al. 2012) are subject to a variety of mechanisms which generate the immense genetic diversity that has become observable via modern-day sequencing techniques. Mutations provide one source of this diversity but certainly not all of it. In fact, mutations can be detrimental to an organism when it lacks the ability to rectify a deleterious mutation, a phenomenon known as Muller’s ratchet (Felsenstein 1974, Muller 1964). Horizontal gene transfer (HGT) is a second source of bacterial diversity. Horizontal gene transfer, also called lateral gene transfer, is the transfer of DNA from one bacterium to another in contrast to the vertical transmission from parent to progeny. The DNA acquired through HGT can serve as a source of non-mutated genes which can then be used to repair deleterious mutations through recombination. More importantly, however, it is a source of vast amounts of novel genetic information that has already

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

153

stood the test of natural selection and as such is considered to be a major driver of bacterial diversity (Thomas and Nielsen 2005). The three classical mechanisms of HGT in bacteria are transformation, conjugation, and transduction. Transformation was the first mechanism of gene transfer to be discovered and occurs naturally in certain bacterial species. The defining characteristic of transformation is that naked DNA is directly acquired in some manner from the bacterium’s environment. In contrast, conjugation involves the transfer of DNA from one bacterial cell to another through a bridge-like connection associated with a conjugative pilus. Conjugation requires cell-to-cell contact between bacteria of different mating types and as such is often regarded as a form of sexual replication. Transduction involves the transfer of DNA from one bacterial cell to another using a bacteriophage (phage) as an intermediate. Transduction occurs when a bacteriophage mistakenly packages bacterial DNA into a phage particle (transducing particle), either in place of, or in addition to its own genome. This chapter considers this role of phages in the horizontal transfer of DNA between bacteria.

Overview of Transduction Transduction can either be generalized, permitting the transfer of any bacterial gene, or specialized, in which only specific bacterial genes have the potential for transfer. Transducing particles are created when the phage accidentally packages host DNA within its capsid either in place of the viral genome (generalized) or with the viral genome (specialized). To understand how these two different forms of transduction occur, we must first examine the normal replication cycle of bacteriophages (see chapter ▶ “Phage Infection and Lysis”). During a typical infection, bacteriophages bind to specific receptors on the surface of a bacterium and inject their DNA into the host cell cytoplasm. Many phages, described as temperate, at this point will choose between productive (typically lytic) and lysogenic cycles (see chapter ▶ “Temperate Phages, Prophages, and Lysogeny”). During the productive cycles, phage progeny will be produced and released from the bacterium, typically lysing the bacterial cell in the process, hence the term lytic cycle. During lysogenic cycles, temperate phages enter into more symbiotic relationships with their bacterial host. During many lysogenic cycles, the phage genome is inserted into the host chromosome using a phage-encoded integrase enzyme. The integrated phage genome, now called a prophage, is then replicated by normal host cell division until certain conditions (e.g., DNA damage) induce the prophage to enter a productive cycle. During induction, the prophage is excised from the host chromosome, synthesis and virion maturation follow, and then progeny are released upon cell lysis. Generalized and specialized transducing particles differ in part as a function of the latter’s reliance on lysogenic cycles. Generalized transducing particles contain only host DNA and occur when the virus accidentally packages that DNA in place of its own DNA during the lytic cycle (Fig. 1a). Both non-temperate and temperate phages

154

C. L. Schneider

a Generalized Transduction

b Specialized Transduction

productive replication host DNA degradation

lysogeny

packaging of viral DNA rare packaging of host DNA

induction rare excision error

productive replication packaging of hybrid DNA

Fig. 1 Generalized and specialized transduction. Generalized transduction (a) occurs when bacterial DNA is accidentally packaged in place of viral DNA during lytic phage replication. Generalized transducing particles therefore contain only host DNA. Specialized transduction (b) is specific to phage that can establish lysogeny and occurs when excision of the prophage during induction is imprecise. Specialized transducing particles therefore contain a hybrid DNA molecule containing some phage DNA attached to bacterial DNA that was originally located adjacent to the prophage integration site

can create generalized transducing particles, and the phenomenon is typically associated with phages that use a headful packaging mechanism as discussed below (for review, see Masters 1996). In contrast, only integrated temperate phages can create specialized transducing particles because the bacterial DNA is acquired during imprecise excision of the prophage DNA during induction (Fig. 1b), where excision is a characteristic only of induced integrated prophages. This aberrant excision

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

155

results in the production of specialized transducing particles, each containing a hybrid piece of DNA with some of the phage genome attached to bacterial DNA, with the latter being originally located adjacent to the phage insertion site (for review, see Weisberg 1996). Following their production, both generalized and specialized transducing particles rely on normal phage infection processes to deliver the mistakenly packaged host DNA to the next bacterium. In order for a bacterium to maintain the new genetic information, the encoding DNA must be incorporated into the host chromosome or be maintained as a selfreplicating element like a plasmid. Incorporation into the host chromosome occurs through recombination, either host-mediated or potentially phage-mediated in the case of specialized transducing particles and site-specific recombination. Hostmediated recombination typically occurs through homologous recombination, and as such the incoming DNA must contain fairly large regions of similarity (25–200 bp), which tends to limit exchange to species that are 70–75% similar (Thomas and Nielsen 2005). These constraints can be bypassed through the use of host-mediated nonhomologous (illegitimate) recombination, and, although most laboratory studies suggest this rarely occurs in wild-type bacterial evolution (Thomas and Nielsen 2005), even rare events in large replicating bacterial populations have the potential to affect bacterial evolution. Autonomously replicating pieces of DNA like plasmids that are transduced whole do not require recombination into existing DNA to be maintained in the new bacterial cell, although a variety of other factors often serve to limit the host range of plasmids, though broad host range plasmids do nevertheless exist (Jain and Srivastava 2013). Ultimately, the success of HGT depends on an ability to transfer new genes from donor to recipient individuals (e.g., such as within virion particles as in transduction) and an ability to stably maintain the acquired genes (such as via recombination into the host chromosome). Additionally, the ability to express newly acquired genes, some of which may provide a selective advantage to the individual bacterium, will play an important role in whether genes acquired by HGT will be maintained by the bacterium.

Historical Perspectives on Horizontal Gene Transfer in Bacteria Horizontal gene transfer (HGT) and mutation play vital roles in creating the vast genetic diversity observed in bacteria. In principal, transfer can and presumably often does occur between members of the same bacterial strain, though in practice such transfers can be difficult to demonstrate after the fact. This is because the signatures of HGT events within bacterial genomes typically are differences in nucleotide sequence, GC content (ratio of DNA nitrogenous base types), or the acquisition of entirely new genes. Although the focus of this chapter is bacteriophage-mediated HGT, the discovery of transduction is so intimately associated with the discovery of transformation and conjugation that we begin with a brief history of the early elucidation of those processes.

156

C. L. Schneider

The Discovery of Transformation The studies reported in 1928 by Fredrick Griffith are foundational and taught in biology classrooms worldwide. Griffith’s unexpected findings on Streptococcus pneumoniae (pneumococcus) formed the basis for studies by many others that followed, which ultimately proved that DNA is the hereditary material. Importantly, Griffith’s studies also provided the first evidence that bacteria have the capacity to acquire genetic information not only from a parental cell (vertical inheritance) but also from other sources (horizontal gene transfer). Pneumonia was a common cause of human mortality in the early 1900s, thus providing the impetus for Griffith’s research. Griffith was studying so-called smooth strains of pneumococcus that were lethal to mice as well as rough strains, which were not. Mice infected with a heat-inactivated smooth strain lived, but surprisingly mice co-infected with a rough strain and a heat-inactivated smooth strain died even though neither treatment alone was lethal. To demonstrate that the phenotypic change was heritable rather than merely a physiological adaptation, he performed the experiment using different antigenic types for the rough and the heat-inactivated smooth strains. His results demonstrated that not only did the rough strain become virulent, but it also assumed the antigenic type of the heat-inactivated strain (Griffith 1928). He accurately deduced that this involved the acquisition of genetic material from a heatinactivated smooth bacterium by a living rough bacterium (Griffith 1928).

The Discovery of Conjugation Joshua Lederberg set out to determine whether bacteria were capable of experiencing genetic recombination through sexual reproduction. Using a K-12 isolate of Escherichia coli, he isolated auxotrophic strains unable to synthesize certain nutrients (biotin and methionine in one strain and threonine and proline in the other). Isolates plated individually on media lacking the necessary nutritional supplements (i.e., minimal media) were unable to grow, demonstrating that these doubly mutated strains could not revert to the wild-type phenotype through the acquisition of reversion mutations. In contrast, when the isolates were mixed together within the same tube before plating on minimal media, then a few colonies were able to grow, demonstrating that the wild-type phenotype had been restored via some type of interaction between the bacteria (Lederberg and Tatum 1946). Further studies by Bernard Davis demonstrated that recombination did not occur if the strains during “mixing” were separated by a bacteria-impermeable membrane that was nonetheless permeable to smaller entities, within a U-shaped tube. This experiment suggested that, unlike transformation, the phenomenon being observed in Lederberg’s experiments required physical contact between the two different strains (Davis 1950). Additional studies by William Hayes and others demonstrated that conjugation only occurs between E. coli strains of different mating types (F+ and F˗) and that transfer of the genetic information is unidirectional, from the F+ strain to the F˗ strain (Wollman et al. 1956). Unlike transformation, which

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

157

involves the acquisition of naked DNA, conjugation involves the transfer of DNA through a structure that protects it from the harsh environment, which is therefore similar, and yet distinct from phage-mediated transduction.

The Discovery of Transduction Continued research in Joshua Lederberg’s lab led to the discovery of transduction within 10 years of their description of conjugation. The discovery of generalized transduction began when Norton Zinder joined Lederberg’s lab as a graduate student and was tasked with extending Lederberg’s findings on conjugation in E. coli to another gram-negative pathogen, Salmonella typhimurium. Following procedures similar to those used by Lederberg, he isolated one auxotrophic bacterial strain, LT-22, that yielded a few prototrophic isolates when crossed with another auxotrophic strain, LT-2. Interestingly, when he performed a U-tube experiment to determine whether gene transfer required physical contact between LT-22 and LT-2, he isolated prototrophic strains in the LT-22 arm of the U-tube suggesting cell-to-cell contact was not required in this instance (Fig. 2) (Zinder and Lederberg 1952). These results suggested that gene transfer was mediated by a diffusible element generated by LT-2. Interestingly, filtrates of pure LT-2 broth cultures grown in the absence of LT-22 did not generate LT-22 prototrophs. Filtrates of broth cultures of LT-2 grown in the presence of LT-22, however, did (Zinder and Lederberg 1952). This suggested that LT-22 stimulated LT-2 to produce a filterable agent that could then be transferred back to LT-22. Since gene transfer was unaffected by treatment of filtrates with DNase, transfer of naked DNA, that is, transformation, could not explain these findings (Zinder and Lederberg 1952). A variety of other experiments were performed to characterize the filterable agent, the results of which demonstrated that LT-22 harbored a prophage. These results were the first to demonstrate that viruses could mediate gene transfer (transduction), and the system they happened upon proved to be the first documented case of generalized transduction. We now understand the molecular reasons behind these findings. Strain LT-22 is lysogenic for the P22 bacteriophage and produces a small amount of infectious P22 phage particles due to spontaneous induction. The resulting phage virions were able to diffuse through the U-tube membrane and infect the LT-2 strain. The majority of these infections resulted in lytic replication of P22 in the LT-2 strain and the production of infectious phage particles, which could return to the LT-22 bacteria through the U-tube. Since LT-22 is lysogenic for P22, these bacterial cells were immune to reinfection (see chapter ▶ “Temperate Phages, Prophages, and Lysogeny”), and thus not lysed, although P22 adsorption and DNA uptake still occurred. A few of the phage particles produced by LT-2, however, were defective and contained bacterial DNA instead of phage DNA (generalized transduction). These generalized transducing particles could then inject their DNA into the LT-22 strain and thus mediate the exchange of genes from the LT-2 strain to the LT-22 strain (Ebel-Tsipis et al. 1972b). Specialized transduction was first demonstrated by Larry Morse, a graduate student also in Joshua Lederberg’s lab. The discovery was dependent on a number

158

C. L. Schneider Pressure or Suction

LT-22 (phe-, trp-, met+, his+)

LT-2 (phe+, trp+, met-, his-)

Cell impermeable membrane (virus permeable)

Plate on minimal media

no growth no prototrophs

growth of prototrophs (phe+, trp+, met+, his+)

Fig. 2 Discovery of transduction. Depiction of the U-shaped tube experiment used by Zinder and Lederberg during their discovery of transduction. The two different strains of Salmonella (LT-2 and LT-22) are placed on opposite sides of the U-tube and are separated by a membrane that is impermeable to bacteria but permeable to the considerably smaller bacteriophage particles. The bacterial strains were auxotrophic for two different amino acids (histidine and methionine for LT-2 and phenylalanine and tryptophan for LT-22) to minimize the generation of prototrophs by a reversion mutation. The P22 phage produced upon spontaneous induction from the lysogenic LT-22 strain was able to cross the membrane and infect the LT-2 bacteria. Replication of P22 in LT-2 generated virulent P22 as well as a few generalized transducing particles (some of which contained the genes for phe and trp synthesis), which were able to pass through the membrane and inject their DNA into the LT-22 bacteria. Stable incorporation of these genes by the LT-22 bacteria permitted the growth of prototrophs on minimal media plates

of prior observations. First was the recent discovery that temperate bacteriophages could lysogenize their bacterial hosts (see chapter ▶ “Temperate Phages, Prophages, and Lysogeny”). Second was the accidental discovery of the lysogenic nature of the E. coli K-12 strain by Esther Lederberg. During the Lederberg group’s research on K-12, they had isolated a mutant strain that was lysed when incubated with the K-12 parental strain. They hypothesized that K-12 was lysogenic for a bacteriophage, which was able to reactivate and lyse the sensitive mutant strain while the parental strain remained immune (Lederberg 1951). Further study of the temperate phage

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

159

within K-12 (λ) demonstrated that, contrary to Joshua Lederberg’s initial hypothesis that the prophage was plasmid in nature (Gottesman and Weisberg 2004), the prophage was integrated into the bacterial host chromosome near the gal locus (Lederberg and Lederberg 1953). This result was independently confirmed by Élie Wollman (1953). These studies set the stage for the Lederberg group to provide the first direct evidence of specialized transduction by demonstrating the transfer of a limited set of gal genes near the site of λ integration from a lysogenic gal+ bacterial strain to a gal sensitive strain (Morse et al. 1956a).

Examples of Transducing Phages Both generalized and specialized transductions are the result of “mistakes” on the part of the phage. To understand the factors that influence these “mistakes,” we must examine the process on a more molecular level. This section will provide a more detailed view of transduction using four well-characterized bacteriophages as examples, namely, P1, λ, and Mu from E. coli and P22 from S. typhimurium. This section is by no means meant to be a thorough review of these bacteriophages. Instead they have been selected to highlight a variety of different mechanisms by which transduction occurs.

Bacteriophage P1 P1 is a temperate phage of E. coli and has been a workhorse for many decades of E. coli genetic research. The P1 generalized transduction system has been useful in constructing mutant E. coli strains and mapping E. coli genes. This section will highlight features of P1 that are relevant to its ability to perform generalized transduction, and the reader is referred to several excellent reviews for a more thorough discussion of P1 (Sternberg and Cohen 1989; Cohen and Sternberg 1989; Sternberg and Hoess 1983; Sternberg and Maurer 1991; Yarmolinsky and Sternberg 1988). P1 is a member of the Myoviridae family and, like most generalized transducing phages, it uses a “headful” mechanism to package DNA (Fig. 3) (Black 1989; Tavares et al. 2012). Headful packaging, first proposed by Streisinger et al. (1967), posits that concatemers containing the phage genome serve as the template for DNA packaging. Packaging begins with the sequence-specific recognition of a packaging signal in the replicated concatemer (pac) by the small subunit of the terminase enzyme (TerS), which is then cut by the large subunit (TerL) generating an end, which can then be inserted into the phage head. DNA packaging continues until the head is full at which point a second, non-sequence-specific double-strand cut is made by TerL. Packaging proceeds unidirectionally from the initial cleavage at the pac site and is processive, meaning that multiple phage particles can be packaged from one concatemer. Since the length of the phage genome is typically less than the amount of DNA needed to fill the phage head, the extra material that is packaged from the

160

C. L. Schneider

Attachment and entry

Circularization via cos or terminal redundancies pac/cos

Replication (concatemer) pac/cos

Recognition, cleavage and packaging by terminase TerS TerL

unit length (cos) packaging

headful packaging

pac

headful 3

pac

headful 2

pac

pac

headful 1

cos

cos

cos

cos

unit unit unit unit length length length length 1 2 3 4

Fig. 3 Phage DNA packaging methods. Following binding of a phage to its receptor and injection of the phage DNA, the phage genome typically circularizes as a consequence of homologous sequences at the end of the genome either as a result of terminal redundancy (headful phage) or cos sites (unit length phage). Circularization protects the phage genome from host-mediated degradation, and it generates the template used for rolling circle replication to generate the concatemer used for phage particle packaging. The TerS protein of the phage terminase complex recognizes the DNA to be packaged by binding to either the pac site or cos site depending on the virus packaging method. The TerL protein makes a sequence-specific cut to initiate packaging into the phage head. For headful packaging, the DNA is loaded into the phage head by the terminase complex until the head is full at which point TerL makes a non-sequence-specific cut and begins loading DNA into the next phage head. For unit length packaging, the DNA is loaded into the phage head until the next cos site is cut by TerL and used to initiate packaging of another phage head. Several phage particles can be packaged from a single concatemeric template for either packaging method

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

161

concatemer will generate genomes that are terminally redundant (Streisinger et al. 1967; Stresinger et al. 1964). Since the P1 genome is roughly 94 kb (Lobocka and Rose 2004) but the phage head capacity is closer to 110 kb, phage P1 has a terminal redundancy of 10–15 kb (Sternberg 1990). For P1 and many other phages, the terminal redundancy generated by headful packaging provides a region of homology to facilitate circularization of the genome upon entry into the host cell cytoplasm either using the host recombination system (RecA) or by phage-encoded recombinases (Segev et al. 1980). Circularization of the genome following injection is a common feature of many phages as it protects the ends of the genome from degradation and provides the template for replication of concatemers or for integration in the case of temperate phages like λ. Like the normal packaging of P1 phage DNA, the packaging of the host genome into transducing particles occurs in a processive manner, most likely originating at a chromosome break or nick rather than a pac-like sequence (Huang and Masters 2014), resulting in fairly random packaging of E. coli genes. Unlike the packaging of phage DNA, which tends to be limited to three to four sequential headfuls (Bachi and Arber 1977), up to five to ten headfuls of host DNA can be packaged from a single initiating event (Sternberg and Coulby 1987), thus permitting the packaging of up to 20% of the E. coli genome (Harriman 1972). P1 generalized transducing particles lack phage DNA, contain a protein covalently linked to the DNA ends (Ikeda and Tomizawa 1965a), and contain host DNA that was synthesized prior to the start of the phage infection (Masters 1996). P1 generalized transducing particles make up ~0.3% of the phage particles produced during lytic replication and appear to be generated by a subpopulation of cells (Ikeda and Tomizawa 1965a; Hanks et al. 1988). For stable transduction to occur, the incoming DNA must either circularize and replicate as a plasmid or be incorporated into the host genome by host- or phageencoded recombinases. Studies by Sandri and Berger followed the fate of P1 transducing DNA upon injection into the next host bacterium. They generated transducing particles in which the DNA was labeled with 32P and purified the transducing particles apart from the infectious particles by CsCl density centrifugation. About 10–15% of the 32P-labeled DNA from transducing particles was incorporated into the host DNA within the first hour (Sandri and Berger 1980). The majority of this label came from degradation of the incoming DNA and recycling of the nucleotides to the host genome upon replication. Only 1–2% of the DNA was stably incorporated into the host genome, with fragments >10 kb being added in a RecA-dependent manner (Sandri and Berger 1980). The majority of the injected DNA persisted as an extrachromosomal fragment for up to 5 h that was transmitted to only one daughter cell upon division resulting in abortive transduction. The structure of the DNA in abortive transductants appeared to be circular and supercoiled, but the ends were not covalently attached but rather held together by a protein (Masters 1996), which serves to protect the DNA from degradation but also makes it resistant to recombination. Interestingly, as seen with other phages, UV treatment of P1 transducing lysates prior to transduction greatly enhanced the numbers of complete transductants (Newman and Masters 1980) likely by introducing gaps in the DNA which made the transducing DNA more recombinogenic (Masters 1996).

162

C. L. Schneider

These results suggest that the generation of a stable transductant by host-mediated homologous recombination must occur within the first hour despite the fact that the potentially transducing DNA persists for much longer within the host cytoplasm. These results are in line with the transduction frequencies of 105–104 per plaqueforming unit (PFU) reported for P1 (Caro and Berg 1971). P1 is an example of a non-insertional temperate phage, meaning the prophage genome is not inserted into the host chromosome during the lysogenic cycle but rather exists as an extrachromosomal circular DNA plasmid-like element (Ikeda and Tomizawa 1968) that is maintained as a low copy number plasmid and actively partitioned into daughter cells (Austin et al. 1981). Therefore, although P1 is temperate, it is unable to generate specialized transducing particles at least in the traditional sense of aberrant excision of the prophage. Interestingly, the P1 genome contains an insertion sequence, IS1, which can facilitate the insertion of DNA flanked by IS elements (Iida et al. 1978). Since the extent of terminal redundancy in P1 is fairly large, P1 can tolerate large insertions without loss of vital viral genes. This is underscored by the discovery of P1 particles carrying multiple antibiotic resistance markers (Iida and Arber 1977; Mise and Arber 1976; Kondo and Mitsuhashi 1964). In this unusual form of transduction, the transducing particles contain a hybrid DNA molecule like those of specialized particles, but the foreign DNA was acquired by insertion into the prophage genome rather than addition to the end of the genome as a result of aberrant excision.

Bacteriophage P22 P22 is a temperate phage of Salmonella typhimurium with many similarities to λ phage of E. coli. P22 has a similar genetic organization to λ and shares homology to λ genes, and viable hybrids can be constructed between P22 and λ (for review, see Susskind and Botstein 1978). As mentioned earlier, the studies in P22 by Zinder and Lederberg provided the first example of generalized transduction (Zinder and Lederberg 1952). This section will highlight features of P22 relevant to transduction, and the reader is referred to the following reviews for more in-depth discussion of P22 (Campbell 1994; Poteete 1988; Susskind and Botstein 1978). P22 is a member of the Podoviridae family and, like P1, it uses a headful mechanism to package DNA. Based on a native P22 genome size of 41.8 kb and a phage capacity of ~44 kb, the terminal redundancy of P22 is only around 4% (Casjens and Hayden 1988). As in P1, the terminal repeats in P22 aid in circularization of the phage genome once it enters the host cell cytoplasm either through the use of host RecA recombination machinery or through the P22-encoded recombinase gene, erf (Botstein and Matz 1970). During lytic replication, concatemers are produced by rolling circle replication of the circularized P22 genome (as reviewed in Poteete 1988), which are packaged in a unidirectional and processive manner for three to ten headfuls (Tye et al. 1974). Packaging begins with cleavage within the pac site (Wu et al. 2002) and is followed by cleavage as determined by the phage head capacity.

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

163

In contrast to phage P1, the packaging of host DNA into generalized transducing particles by phage P22 appears to be the result of initiation at a few specific pac-like sites within the host chromosome. The results of studies examining differences in phage densities (Ebel-Tsipis et al. 1972a), variations in transduction rates for individual host genes (Schmieger 1970), and co-transduction rates (Chelala and Margolin 1974) all support the model that P22 generalized transduction initiates from a few specific pac-like sites and is processive for at least ten headfuls covering 10% of the host genome. Additional support for this mechanism came from the characterization of high-transducing (HT) P22 mutants (Schmieger 1972). All of these HT mutants map to gene 3, which encodes the small terminase subunit (TerS) and contains the pac site. Interestingly, most mutations appear to affect the function of the TerS protein rather than the pac site (Casjens et al. 1992). One of the mutants, HT12/4, appears to direct the packaging machinery to two new sites in the P22 genome suggesting the high-transducing frequency might be the result of reduced recognition of phage DNA and/or increased recognition of host DNA sites (Casjens et al. 1992). The fate of host DNA contained within generalized transducing P22 particles is similar to that of P1, with only 2% of generalized transducing phages producing stable transductants as measured using radioactive nucleotide incorporation (EbelTsipis et al. 1972b). The formation of stable transductants typically involves integration of the incoming DNA into the host chromosome via the host RecA recombination system (Ebel-Tsipis et al. 1972b). The structure adopted by the DNA from abortive P22 transductants might also be similar to P1 with the gene product of P22 gene 16 (gp16) possibly binding to the DNA ends to protect them from degradation, which also prevents RecA-mediated recombination by shielding them from the host RecBCD exonuclease (Benson and Roth 1997). Unlike P1, P22 undergoes site-directed insertion into the host chromosome during lysogeny (for review, see Poteete 1988). This feature permits phage P22 to also mediate specialized transduction of host genes that lie near the attB site as a result of aberrant excision during reactivation (Kwoh and Kemper 1978). The prophage integration and excision mechanism used by P22 are very similar to that used by phage λ, as discussed in more detail below.

Lambda (l) Research characterizing λ, a temperate phage of E. coli K-12, has contributed immensely to the field of molecular biology and the understanding of gene regulation. Restriction enzymes commonly used today to create recombinant DNA were first identified as a host defense system to λ infection (Bertani and Weigle 1953; Arber 1974; Weigle and Bertani 1953). Research characterizing the mechanism of lysogeny in λ also played a critical role in the development of the elegant theory of gene regulation of the lac operon by Jacob and Monod (Jacob and Monod 1961). This section focuses on the features of phage λ germane to its ability to perform specialized transduction, and the reader is referred to several excellent reviews

164

C. L. Schneider

(Casjens and Hendrix 2015; Court et al. 2007; Gottesman and Weisberg 2004) as well as the Lambda II monograph (Hendrix et al. 1983) for a more detailed history of the discovery of phage λ as well as a detailed description of λ biology. Lambda is a member of the Siphoviridae family, and following adsorption, the ~48 kb linear, double-stranded DNA genome enters the host cell cytoplasm and circularizes via the terminally redundant cohesive 50 overhangs (cos) found at either end of the genome (Hershey et al. 1963; Hershey and Burgi 1965). The circularized genome provides the template for either lytic replication or lysogenization (see chapter ▶ “Temperate Phages, Prophages, and Lysogeny”). Under conditions favoring lytic replication, the circularized genome undergoes theta replication (bidirectional) for several rounds that is then followed by a switch to rolling circle replication to generate the concatemers, which serve as the packaging template for virus assembly (Takahaski 1975). The phage-encoded terminase enzyme cuts at the first cos site to initiate packaging, and in contrast to the non-sequence-specific cuts used in headful packaging, subsequent cuts are made at each cos site (Fig. 3) (for review, see Catalano et al. 1995). Both insertions and deletions can occur within the λ genome, but efficient packaging occurs only for genome sizes ranging from 78% to 105% of the reference λ genome (Feiss et al. 1977). Under conditions that favor lysogeny, the circularized genome is integrated into the host chromosome by integrase, a site-specific tyrosine recombinase encoded by the λ int gene. Integration catalyzed by λ integrase occurs by reciprocal recombination between a 15 base-pair phage attachment site (attP) and a homologous bacterial attachment site (attB) and requires host-encoded protein integration host factor (IHF) and factor for inversion stimulation (Fis) (Ball and Johnson 1991). The resulting prophage is integrated into the host chromosome between the gal and bio genes and flanked by attachment sites (attL and attR) (Fig. 4a) (for review, see Echols and Guarneros 1983; Nash 1981). Phage λ can remain in a lysogenic state for thousands of bacterial generations or more, but certain conditions, like activation of the SOS response to UV-induced DNA damage, can promote excision of the prophage and activation of lytic replication. Excision of the prophage requires all of the proteins necessary for integration as well as the product of the λ Xis gene (for review, see Echols and Guarneros 1983). Specialized transducing particles are produced when prophage excision is imprecise and host genes adjacent to the integrated prophage are accidentally included in the excised fragment (Fig. 1). Specialized transducing particles are therefore only produced upon induction of a lysogen and not upon lytic infection of a sensitive bacterium (Morse et al. 1956a). In λ, the gal and bio genes, which are adjacent to the attL and attR sites, respectively, can be packaged into specialized transducing particles as long as the excised fragment still contains a cos site for efficient packaging. The excised fragment must also still conform to the size limits for efficient phage packaging, and as such, specialized transducing phage often lacks viral genes located at the prophage end furthest from the incorporated host DNA. Phages that carry the bio gene (λbio) therefore lack phage genes required for lysogeny (int), while phages that carry the gal genes (λdgal- d reflects the defect in lytic replication) lack phage genes required for lytic functions (tail genes) (Fig. 4a)

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

a

165

Generation of transducing λ dgal and λ bio (LFT) cos attP

tail gal

λ int bio

E. coli

attB Integration of λ at attB site (Int, IHF, Fis)

Excision (Induction) (Xis, Int, Fis) gal

int

tail cos

attL

bio

rare illegitimate excision (Fis) cos

cos λ dgal

attL gal

E. coli λ lysogen

attR

int

λ bio

attR tail

bio

Generation of λ dilysogen (HFT)

b

cos λ dgal

attL gal gal

int int tail bio cos attL attR

E. coli λ lysogen

integration of λ dgal at attL site int

gal

gal cos

attL

attL

int

tail cos

~ 50%

gal

E. coli dilysogen

attR ~ 50%

cos λ dgal

bio

cos

attL int

tail

attP

λ int

Fig. 4 Generation of low-frequency transducing (LFT) and high-frequency transducing (HFT) λ lysates. LFT lysates (a) are the result of imprecise excision of a λ prophage following induction. During the establishment of lysogeny, the λ genome is integrated into the E. coli genome by recombination between the phage att site (attP) and the bacterial att site (attB). This reaction is catalyzed by the phage-encoded integrase enzyme (Int) but also requires two host proteins: integration host factor (IHF) and factor for inversion stimulation (Fis). The integrated prophage is flanked on the left by the gal operon (galactose metabolism) with the λ integrase gene nearest this

166

C. L. Schneider

(for review, see Weisberg 1987). Specialized transducing phages typically require a helper phage to provide the missing phage gene products in trans in order to be competent in both lytic and lysogenic cycles (Echols and Court 1971). The inaccurate excision that generates specialized transducing phages is different from that of normal Int-/Xis-mediated excision and consequently is a relatively rare event. Excision in this case occurs through a form of illegitimate recombination, which requires the gene product of Fis but not Int or Xis and is independent of RecA (Shanado et al. 1997). The frequency of specialized transducing particles by an aberrant excision is typically only one transducing particle per 106–107 total phage following UV-activated induction (Morse et al. 1956a). Phage populations generated in this fashion are referred to as low-frequency transducing (LFT) lysates. UV and other DNA-damaging agents promote the formation of transducing particles by promoting induction by targeting the λ lysogeny regulator protein CI for degradation (for review, see Roberts and Devoret 1983) (see chapter ▶ “Phage Infection and Lysis”) and by increasing illegitimate recombination (Ikeda et al. 1995). Highfrequency transducing (HFT) lysates containing up to 50% transducing particles can however be generated, the mechanism for which will be discussed below (Morse et al. 1956a). As is the case for generalized transducing DNA, the specialized transducing DNA must either be incorporated into the host chromosome or be maintained as a selfreplicating plasmid in order for stable transduction to occur. Stable incorporation of DNA from specialized transducing particles into the next host genome can occur by (1) replacement of the homologous region of the host DNA with the transducing DNA by two host-mediated homologous recombination events, (2) by integration of the transducing phage at the homologous site via host-mediated recombination with a circularized phage, or (3) by integration of a circularized phage at the attB site by the phage-encoded integrase (Fig. 5) (provided they contain the int gene) (Wolf 1980). Transductants generated by the first mechanism would not be lysogens while transductants generated by the second and third mechanism would be lysogens. For λdgal, roughly one-third of the gal+ transductants produced by inducing a gal+ lysogen and infecting a gal strain were stable and non-lysogenic (mechanism 1), while two-thirds were unstable heterogenotes (gal+/gal) (mechanism 2 or 3)

ä Fig. 4 (continued) end and on the right by the bio operon (biotin synthesis) with the λ tail fiber genes nearest this end. Induction typically results in a reversal of this integration reaction requiring the phage-encoded excisionase (Xis) and integrase proteins as well as the host protein Fis. Following induction, rare illegitimate excision occurs leading to either λdgal transducing particles that contain host DNA from the gal locus and lack λ tail genes or λbio particles that contain DNA from the bio operon and lack the int gene. This illegitimate excision event requires Fis but not Xis or Int. In contrast, HFT lysates (b) are generated when either a λdgal or λbio transducing particle or a wild-type λ generate a di-lysogen. This can occur when a transducing particle infects a lysogen or by co-infection of a cell by both transducing and wild-type particles. Integration in this case occurs by recombination between the λdgal attL (or attR for λbio) and the prophage attL (or attR) site. Induction of a di-lysogen produces lysates with about half of the particles being transducing particles and the other half being wild-type λ particles

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

167

a Host-mediated replacement by double recombination A+

B+

D-

A-

B-

C-

A+

B+

C-

b Host-mediated addition by single recombination b attP

attB

A-

attB

c

B+

a

A+

B+

A-

B-

C-

attP

b

a

A+

B-

C-

Phage-mediated addition by single recombination a

B+ b

A+ attP

attB

attL

A+

B+

a

b

A-

B-

attR

C-

A-

B-

C-

Fig. 5 Incorporation of transducing DNA into the host genome by various types of recombination. DNA from either generalized or specialized transducing particles can replace the homologous section of DNA in the new host genome by host recombinase-mediated double recombination (a). DNA from specialized transducing particles can also be added to the host genome by a single host-mediated recombination between a circularized phage genome and the homologous section of DNA within the host genome (b). In specialized transducing phages that contain a functional integrase gene, the transducing DNA can be added to the host genome by an integration reaction between the phage attP site on a circularized phage genome and the host attB site (c)

168

C. L. Schneider

(Morse et al. 1956b). Since λdgal phages contain int, presumably they would have the potential for phage integrase-mediated integration at the attB site. Given the high efficiency of integrase-mediated recombination, transducing phages containing int (e.g., λdgal) would potentially be expected to transduce with higher frequency than transducing phages that lack int (e.g., λbio). However, the attL site present on gal transducing phages is not identical to the λ attP site and is integrated at the attB site with such reduced efficiency that typically a helper phage is required for efficient lysogenization by λdgal (Sato and Campbell 1970; Weisberg and Gottesman 1969). Interestingly, induction of heterogenotes resulted in the production of HFT lysates with transduction efficiencies several orders of magnitude higher than LFT lysates. Further analysis of the gal+ heterogenotes demonstrated that many are di-lysogens in which the transducing phage has been integrated at the attL site to the left of a wild-type λ prophage (Fig. 4b) (Campbell 1957). Induction of the di-lysogen results in the production of λdgal and wild-type λ phage particles in equal proportions as a result of the equal likelihood of excision via attL x attL (λdgal) or via attL x attR (wild-type λ). HFT lysates have transduction efficiencies ranging from 101 to 103 depending on MOI and also whether the recipient is lysogenic or sensitive to infection by λ, with the highest transduction frequencies occurring at high MOIs in sensitive recipients (Campbell 1957).

Mu In 1963, when Larry Taylor accidentally isolated a temperate phage with a propensity to induce mutations in E. coli, he named it Mu for mutator (Taylor 1963). The realization that Mu replicated as a transposable element led to a wealth of research on mobile DNA using Mu as a model system. This research paved the way toward rapid understanding of the mechanism of HIV integration (which is similar to Mu integration) as well as the development of drugs to target the HIV integrase enzyme (Harshey 2012). This section will focus on the features of Mu required for transduction, and the reader is referred to the following reviews for a more in-depth discussion of Mu (Howe and Bade 1975; Harshey 2014; Harshey 1988; Bukhari 1976). Mu is a member of the Myoviridae family that can infect a variety of different bacterial species (Sandulache et al. 1984). Phage Mu has the unusual feature that each virion contains a linear DNA molecule with a core viral genome flanked on each end by variable host DNA segments acquired during lytic replication. Following adsorption, the double-stranded linear DNA containing the ~37 kb Mu genome, flanked by 50–150 bp of host sequences on the left and ~2 kb on the right, enters the host cell cytoplasm. The termini are protected from degradation by a viral protein (N) which facilitates the noncovalent circularization of the genome in preparation for integration (Harshey and Bukhari 1983; Puspurs et al. 1983). During the infection phase, the Mu genome is inserted into the bacterial genome through a non-replicative transposition reaction catalyzed by the DDE transposase enzyme MuA (for review, see Harshey 2014). The flanking host DNA on the incoming Mu genome is not

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

169

integrated into the new host chromosome, and recent data suggest the removal of the flanking DNA is partially mediated by the host RecBCD exonuclease with final trimming from an as yet unidentified enzyme (Choi et al. 2014). Unlike phage λ, Mu integration can occur in many locations in a somewhat random fashion (Bukhari 1976), and lytic replication is not induced by treatment with UV- or DNA-damaging agents (Harshey 1988). The replication strategy of Mu is quite distinct from other phages but similar to transposons. Following inactivation of the repressor protein that maintains lysogeny, the DDE transposase (MuA) and target DNA activator B (MuB) are expressed, and these facilitate the replicative transposition of the Mu genome resulting in the insertion of 50–100 copies of Mu per host genome (Harshey 1988). Each integrated prophage serves as the template for the headful packaging of presumably a single genome into a Mu virion. Packaging initiates through recognition of a pac sequence near the leftmost region of the genome with cleavage occurring within the flanking host DNA (Harel et al. 1990; Groenen and van de Putte 1985). Packaging continues until the head is full, leading to the inclusion of variable lengths of the host genome on the rightmost end depending on the length of the phage genome (Bukhari and Taylor 1975). Each virion will therefore contain different host flanking sequences as a result of the differing locations of each prophage within the host chromosome. Mu is typically characterized as a generalized transducing phage with transduction frequencies ranging from 106 to 109 per PFU (Howe and Bade 1975). Although some of this transduction may be due to the transfer and subsequent incorporation of host sequences attached to the right end of the Mu genome, a few findings suggest this is not likely in native Mu. First, as mentioned previously, the flanking host DNA is degraded upon integration of the Mu genome during the initial infection and thus would not be available for incorporation through host-mediated homologous recombination. Second, co-transduction frequencies are similar to what is expected based on the size of DNA packaged in a Mu virion (Howe 1973), suggesting that generalized transduction likely occurs through the accidental packaging of host DNA in place of viral DNA (Masters 1996). Whether this occurs as a result of recognition of separate pac-like sequences in the host genome or processive packaging following initiation at a true Mu prophage is unclear (Schroeder et al. 1974; Masters 1996). It is worth noting that Mu derivatives with greatly decreased genome sizes (e.g., mini-Mu) have been engineered for use as high-frequency transducing agents (frequencies of 105–106 per PFU) (Wang et al. 1987). Following induction in the presence of a helper phage, the mini-Mu element and up to 30 kb of adjacent host DNA can be packaged into transducing virions. Again, this type of specialized transduction has not been demonstrated for wild type (Howe 1973).

Phage Features Important for Transduction Not all phages transduce, although given appropriate conditions, several phages, like T4, that typically do not transduce, can generate transducing particles (Wilson et al. 1979). In the previous section, we examined some of the molecular mechanisms by

170

C. L. Schneider

which both generalized and specialized transductions occur using specific example phages. This section focuses on the general features of phages important for generalized and/or specialized transduction, namely, the degradation of host DNA, DNA packaging method and specificity, the presence of other mobile genetic elements within the phage genome, and the specificity of the integration site of temperate phages.

Host DNA Degradation In order for host DNA to be packaged into generalized transducing particles, the host DNA must be intact at the time of packaging. Therefore, non-temperate phages like the T-even and T-odd phages of E. coli, which typically degrade the host DNA to generate the building blocks of their own DNA, do not transduce under normal conditions (Masters 2004; Waddell et al. 2009; Łobocka et al. 2014). Mutants of T4 that are unable to degrade host DNA produce generalized transducing particles using a headful packaging method at a frequency of 108 per PFU (Young et al. 1982). Wild-type bacteriophage T1, when grown under assay conditions that protect transduced cells from lysis, produced a low frequency of generalized transducing particles early during infection, presumably before host DNA degradation (Drexler and Christensen 1979). Furthermore, mutants of T1 that are unable to degrade host DNA exhibit a high rate of generalized transduction with frequencies ranging from 103 to 104 per PFU (Roberts and Drexler 1981a; Roberts and Drexler 1981b). Temperate phages, which do not typically degrade host DNA, are therefore more likely to transduce, as exemplified by the fact that all of the classical transducing phage examples discussed above are temperate phages.

DNA Packaging Method Another important feature of generalized transducing phages is the method they use to package DNA. Phages that use a headful packaging method, namely, those with circularly permuted genomes, transduce, while phages whose genomes have invariable ends typically do not (Klumpp et al. 2008; Penades et al. 2015). By comparing these different packaging methods, we can see the reasons behind this finding. In headful packaging, the pac site is recognized by TerS, and the DNA is cleaved by TerL to initiate insertion into the phage head. The second cut is not dependent on a specific sequence but rather occurs once the phage head is full (Fig. 3). Therefore, provided the host DNA can be recognized and initiate packaging, the processive nature of packaging will result in the production of several generalized transducing particles from a single initiation event. In contrast, phages that package a single nonpermuted genome by making specific cuts at the cos site found at each end of the genome (e.g., λ, T3, and T7) are less likely to perform generalized transduction because the likelihood that two cos-like sites within the bacterial chromosome will be an appropriate distance from each other to facilitate efficient packaging is very

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

171

low (Fig. 3). This is exemplified by the discovery that under certain conditions, phage λ can generate generalized transducing particles but the packaging is not completed within the cell because of the lack of a cos site at the terminus (Sternberg 1986; Sternberg and Weisberg 1975). Once the particles are released from the lysed cell, the DNA that is protruding from the phage head, however, can be removed by treating with DNase, and infectious particles can be generated by the addition of tail proteins (Sternberg and Weisberg 1975). Since the only factor regulating the specificity of headful packaging is the initial recognition and potentially cleavage of the DNA by the terminase complex, the specificity of initiating packaging plays an important role in the frequency of transducing particles produced as well as the range of bacterial genes that can be transduced. Certain phages, like P1, appear to package host DNA that lacks recognizable pac sites (Hanks et al. 1988), and packaging efficiency of different regions of the host genome is proportional to their relative representation within the cell (Masters 1977; Masters 1996). These results suggest that host DNA packaging initiates from DNA ends rather than via the recognition of pac-like sequences (Huang and Masters 2014). Given this flexibility in packaging initiation, it is not surprising that P1 is fairly efficient at the generalized transduction of a wide variety of host genes. In contrast, as mentioned before, phage P22-mediated generalized transduction occurs by recognition of a few pac-like sites within the host chromosome. Thus, the variety of genes that can be transduced by P22 is more limited given its increased specificity in initiating DNA packaging. The hightransducing P22 mutant HT12/4a mentioned earlier that leads to transduction initiation from two new host sites exemplifies the link between generalized transduction and the specificity of the recognition of DNA by TerS. Presumably the mutation within gene 3, which codes for TerS, increases recognition of these two new host sites and/or decreases recognition of the viral pac site (Casjens et al. 1992).

Mobile Genetic Elements Within Phage Genomes Given the demonstrated recombinant prowess of phages as evidenced by their highly mosaic genomes (Hendrix 2003), the presence of regions of homology between a temperate phage and its host might promote transduction through the addition of host genes within the phage genome, provided the phage has sufficient terminal redundancy to accommodate the additional DNA. Insertion elements (IS) and transposons (Tn) are known to facilitate the transfer of DNA and have been found in P1- and P7-like phages (Iida et al. 1978; Iida et al. 1981). Several examples of P1- or P7-like phages capable of transducing antibiotic resistance genes appear to contain IS or Tn elements flanking the antibiotic resistance genes (Smith 1972; Kondo and Mitsuhashi 1964; Billard-Pomares et al. 2014). In fact, there is evidence to suggest that, at least in P1, phage-encoded recombinases might promote IS excision (Lu et al. 1989). This unusual form of transduction would appear to provide a very effective means for HGT by co-opting phage recombinases.

172

C. L. Schneider

Integration Site Specificity Specialized transduction in the classical sense results from the inaccurate excision of a temperate prophage upon induction. Presumably once the excision event occurs, any phage released upon lysis of that cell will have the potential to transduce another cell. Since the host genes that can be transduced must lie adjacent to the insertion site, specialized transduction is limited by the specificity of the prophage integration reaction. This limitation can be overcome in phage λ by deletion of the attB site, which increases the rate of integration at secondary insertion sites within the E. coli chromosome (Shimada et al. 1975; Shimada et al. 1973; Shimada et al. 1972). In fact, many researchers have exploited this feature to study genes within E. coli that are found adjacent to secondary insertion sites (Boulter and Lee 1975; Chung and Greenberg 1973; Katz 1970; Lindahl et al. 1970). Although the integration reaction at secondary sites is less efficient than integration at the wild-type attB site, there is evidence to suggest that lambdoid phages have the capacity to improve this efficiency by adapting their int genes to a secondary attB site (Rutkai et al. 2006).

Examples of Transduction Hijackers Phage particles provide a diffusible, targeted genetic delivery package that is stable in many environments, so it is perhaps not too surprising that this efficient carrier service has been hijacked to promote efficient HGT within certain bacterial species. The following section is intended to introduce the reader to two mechanisms that utilize phage particles to mediate HGT, each in a manner that is similar yet distinct from transduction.

Gene Transfer Agents Gene transfer agents (GTAs) are phage-like particles that appear to be dedicated solely to facilitating HGT and are especially prevalent in α-proteobacteria, although they have been found in other bacterial species, as well as archaea (for review, see Lang et al. 2012). Gene transfer agents were originally identified in Rhodobacter capsulatus based on their ability to efficiently transfer antibiotic resistance genes in a manner similar to but distinct from transduction (Marrs 1974). Gene transfer agent particles are similar to generalized transducing particles in that they contain only host DNA; they differ however in some important ways. While generalized transducing particles represent a minority of the particles produced during a productive infection (< 1%), all of the GTA particles produced by a cell contain a random, linear fragment of host DNA that is too small to encode the genes necessary to produce GTAs (Lang et al. 2012; Yen et al. 1979). As such, GTAs cannot transfer the ability to produce GTAs to another cell, and the production of GTAs is not a result of phage or GTA infection (Lang et al. 2012). Instead, the genes encoding GTAs reside within the chromosome of the GTA-producing cell and have gene order and

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

173

sequence similarity to prophages, highlighting the evolutionary connection between phages and GTAs (Yen et al. 1979; Lang et al. 2012). The mechanism by which GTAs select and package host DNA fragments has not yet been determined, although it is postulated to be similar to headful packaging. In GTAs, the gene encoding TerL appears to be a conserved feature while the gene for TerS is typically absent (McDaniel et al. 2012; Paul 2008). Since TerS typically recognizes the DNA to be packaged, the lack of a gene encoding TerS in many GTA gene clusters might account for the lack of DNA specificity seen in GTAs (Paul 2008). In R. capsulatus, all genes appear to be packaged with equal efficiency, suggesting a random mechanism for the selection of DNA, although genes encoding the RcGTA particle were packaged less frequently (Hynes et al. 2012). Presumably the host chromosome serves as the template for packaging, with packaging proceeding in a headful manner following initial binding and cleavage by TerL. This, however, has yet to be determined experimentally (Hynes et al. 2012). Gene transfer agent particles are typically smaller than generalized transducing particles and transfer correspondingly smaller host DNA fragments (~5–15 kb) (Lang et al. 2012), which then can replace the corresponding region of DNA in the next host chromosome in a RecA-dependent manner (Genthner and Wall 1984). Like prophages, GTA gene expression can be induced by certain environmental conditions, although the conditions that lead to expression differ between prophages and GTAs. Unlike many prophages, most GTAs are not induced by DNA-damaging agents like mitomycin C but appear to be expressed during unfavorable environmental conditions (for review, see Lang et al. 2012; Paul 2008). Interestingly, in R. capsulatus, GTA particle expression appears to be controlled by the same response regulator (CtrA) that regulates transformation (Brimacombe et al. 2014), and the transfer of DNA to a recipient cell requires genes involved in the import of DNA during transformation (Brimacombe et al. 2015). Thus, it seems that at least R. capsulatus GTAs might use a novel system for HGT that combines features of phage-mediated generalized transduction and natural transformation.

Phage-Inducible Chromosomal Islands Phage-inducible chromosomal islands (PICIs) are a family of mobile genetic elements found within the chromosomes of gram-positive bacteria which hijack capsids from helper phages to promote their own dissemination (for review, see Novick and Ram 2016; Penades et al. 2015; Penades and Christie 2015). The best characterized members of the PICI family are the pathogenicity islands of Staphylococcus aureus (SaPIs), which encode several clinically relevant toxins such as toxic shock toxin and enterotoxin B (Penades et al. 2015). In addition to being able to hijack helper phage particles, SaPIs are able to promote generalized transduction even in the absence of SaPI induction. The SaPIs are mobile genetic elements observable as ~15 kb islands within the S. aureus chromosome with notable similarities to prophages. The first SaPI to be characterized, SaPI1, contains the gene for toxic shock syndrome toxin (TSST-1) as

174

C. L. Schneider

well as several homologues of phage genes including int, xis, and terS but, notably as discussed below, not terL or the structural genes to generate capsids (Lindsay et al. 1998; Ruzin et al. 2001). Similar to prophages, SaPIs are maintained in a dormant, integrated state by a repressor protein (Ubeda et al. 2008). In contrast to many prophages, however, DNA-damaging agents that activate the SOS response do not induce expression of the SaPI-encoded genes. Instead, antirepressor proteins encoded by helper phages inactivate the SaPI repressor protein thereby inducing SaPI expression (Ubeda et al. 2008). Similar to prophages, the induction of SaPIs leads to their excision followed by replication to generate linear concatemers, which are typically packaged by the headful mechanism (for review, see Novick and Ram 2016). As mentioned above, the genomes of SaPIs lack homologues to phage TerL genes as well as the structural genes needed to generate phage particles, and as such, they must parasitize helper phage components to facilitate their transfer. Helper phage and the SaPIs each encode their own TerS proteins, which can individually complex with the phage-encoded TerL to initiate packaging from either the phage pac site or the SaPI pac site, respectively (for review, see Novick and Ram 2016). SaPIs promote the selective packaging of their own DNA by encoding proteins which bind to phage TerS to block phage DNA packaging (Ram et al. 2012) and by promoting the production of smaller capsids, which accommodate the smaller SaPI but not the larger phage genome (Quiles-Puchalt et al. 2014b; Ram et al. 2014; Ruzin et al. 2001; Tormo et al. 2008). A few SaPIs lack both terS and an intact SaPI pac site and thus rely entirely on the helper phage packaging system. Interestingly, a few of these SaPIs have both phage pac sites and cos sites, thus permitting their packaging by helper phages of either type (Quiles-Puchalt et al. 2014a). The ability of SaPIs to promote their own dissemination by hijacking materials from helper phages likely plays an important role in the spread of SaPI-encoded virulence genes (for review, see Novick (2003); Lindsay and Holden 2006). Nevertheless, SaPIs, like most headful packaging phages, also produce generalized transducing particles during productive replication. As discussed previously, generalized transducing particles can be produced as a consequence of the initiation of headful packaging from pac-like sites in the host genome by TerS. Similarly, recognition of SaPI pac-like sequences within the host chromosome by SaPI TerS generates generalized transducing particles (Chen et al. 2015). Interestingly, SaPIs can promote the production of generalized transducing particles by non-helper phages in the absence of SaPI induction (Chen et al. 2015). As mentioned above, in contrast to many other temperate phage repressor proteins, the SaPI repressor protein is not inactivated upon activation of the SOS response. Surprisingly, expression of the SaPI terS gene nevertheless is induced by the SOS response, even in the absence of SaPI induction (Ubeda et al. 2007). Assuming another phage within the cell is also undergoing productive replication, the SaPI TerS protein could promote headful packaging initiating from the SaPI pac-like sites within the host chromosome as long as it could interact with the phage-encoded TerL protein (Chen et al. 2015; Novick and Ram 2016). As such, SaPIs promote not only their own dissemination but also the exchange of a wide variety of host genes within

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

175

a population by increasing the frequency of generalized transduction under conditions that induce the SOS stress response. It is worthwhile to note that certain antibiotics such as quinolones induce the SOS response and therefore may increase the risk of HGT by SaPIs (Ubeda et al. 2005; Maiques et al. 2006).

Methods to Detect and Characterize Transducing Particles Researchers have used a variety of methods to evaluate the role transduction plays in mediating HGT in bacteria. Some of the methods detect phenotypic changes in either the bacterial host or the transducing particle. Others detect genotypic changes of the bacterial host or, more often, the transducing particles. Some are used to characterize phage stocks generated by infection of a host bacterium by a single phage isolate, while others are used to analyze environmentally sourced phage communities. Some methods require the use of a culturable bacterial host while others do not. Each method has its strengths and weaknesses, and a precise understanding of what each method measures is vital in interpreting transduction data.

Phenotypic Changes of the Bacterial Host The classical method used to detect transduction (generalized or specialized) is to select for bacterial hosts whose phenotype has been altered as a consequence of infection by a transducing particle. Generally this involves initially replicating a purified phage in a donor strain of bacteria with a selectable genetic marker (e.g., antibiotic resistance or nutrient source) to generate transducing particles (and infectious particles), some of which presumably would contain the gene(s) encoding the selectable marker. The particles produced by the donor strain are then used to infect a susceptible recipient strain that lacks the genetic marker. The rare transductants are then detected by plating the potentially transduced recipient cells on selective growth media. This method can be used to detect generalized transduction or specialized transduction when the particles are produced by a lysogenic donor strain as a consequence of induction. The reader is referred to the following protocols for more in-depth discussion of specific methods for detecting generalized transduction (Thierauf et al. 2009; Waddell et al. 2009) or specialized transduction following the induction of a lysogen (Raya and H'bert 2009; Varga et al. 2012). These methods are useful to detect transduction by temperate phages but must be modified to detect generalized transduction by non-temperate phages. For non-temperate phages, transduction of the recipient cell must be done at low MOIs to prevent the lysis of recipient cells by co-infecting wild-type particles. Additional options to reduce recipient cell lysis by co-infection with wild-type phage include (1) UV irradiation of the phage lysate, (2) the inclusion of antiserum following the initial adsorption phase to limit infection of a single cell by more than one phage particle, and (3) the use of phage mutants which are conditionally unable to lethally infect bacteria (Waddell et al. 2009).

176

C. L. Schneider

A variety of controls should be included in these types of experiments to ensure that resulting phenotypic conversion is a consequence of transduction. To prevent transformation, the phage lysate must be free of non-encapsidated DNA. This is typically accomplished by treating the lysate with DNase before infecting the recipient strain. To ensure that the bacterial growth on selective plates is not a result of donor-cell contamination, the presumptively transducing lysate is first treated with chloroform to kill residual donor cells. The inclusion of a “no-phage” control (recipient cells only) and a “no-cell” control (phage lysate only) is also necessary to ensure the phenotypic conversion is dependent on both phage particles and infection. Additionally, the ability to block phenotypic conversion by blocking adsorption using phage antiserum provides stronger support for transduction. The biggest advantage of this classical approach is that it measures the transfer and phenotypic expression of the selectable marker, in other words bona fide transduction. The transductants are typically complete transductants, although microcolonies of abortive transductants (cells in which the foreign DNA is present and expressed but never integrated or replicated) may also be isolated (for review, see Masters 1996). One disadvantage is that it limits the detection of transduction to selectable markers, typically those genes that can be rapidly phenotypically expressed. Another drawback is that it is limited to phages with culturable bacterial host strains. Finally, this method is not able to detect transducing particles from phage populations isolated from environmental samples.

Phenotypic Changes of the Transducing Particle Density centrifugation is the classic method that has been used to isolate transducing particles from phage lysates for further characterization. Since the phage DNA and host DNA often differ in their GC content, the density of the transducing particles will often differ from that of the infectious particles (Ikeda and Tomizawa 1965a). In fact, even the relatively small change in specialized transducing fragments can result in distinguishable differences in density as seen for λ (Weigle et al. 1959). Additionally, sometimes the proteins packaged within the transducing particle may differ from those packaged in the infectious particle (e.g., phage P1 transducing particles), which may alter the density of the transducing particles (Ikeda and Tomizawa 1965b). This method is advantageous when the researcher would like to make comparisons between the transducing particles and infectious particles for a purified phage but clearly is limited to those transducing particles that have a density that is distinguishable from the infectious particles and is not applicable to phage mixtures.

Genotypic Changes of the Bacterial Host Detecting the transfer of DNA from a transducing particle into a single bacterial cell was impossible until quite recently. A method developed to detect individual genes within a living bacterial cell (Kenzaka et al. 2005) called cycling primed in situ

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

177

amplification-fluorescent in situ hybridization (CPRINS-FISH) has been used recently to detect transduction within a single cell (Kenzaka et al. 2007; Kenzaka et al. 2010). This sensitive method combines the features of FISH and in situ PCR to detect the transfer of specific pieces of DNA to individual bacterial cells. Several technical hurdles had to be overcome in the development of the CPRINSFISH method (Kenzaka et al. 2005). Conditions to permeabilize the bacterial cells had to be developed so that the PCR reagents could enter the cell without allowing the PCR products to exit. The PCR reaction was modified to use a single primer, thus generating a single-stranded product that is several kilobases in length to limit its propensity to diffuse out of the cell. The use of multiply-labeled fluorescent probe sets instead of a single probe improved the specific detection of the PCR product using epifluorescence microscopy. To permit the use of this technique on environmental samples that typically require filter-based concentration prior to analysis, the researchers developed a procedure to filter the sample onto polycarbonate filters and subsequently coat them in gelatin. The reader is referred to the original paper by Kenzaka et al. (2005) for a more detailed discussion of the CPRINS-FISH method. The application of this method to detect transduction requires only a few procedural additions. Similar to the classical method of detecting transduction, the transducing phage is replicated in a host with a particular genetic marker to generate transducing particles. To purify the particles and ensure they are free of contaminating non-encapsidated DNA, the virions are precipitated using polyethylene glycol (PEG), treated with DNase, and then purified by CsCl density centrifugation. The reader is referred to the following papers for a more detailed discussion of the application of the CPRINS-FISH method to the detection of transduction (Kenzaka et al. 2007; Kenzaka et al. 2010). The advantages of this method are that it permits the detection of a chosen gene at the level of a single cell in a highly sensitive manner, even in complex environmental samples. One disadvantage to this method is that the detection is again limited to a gene chosen by the researcher. Some caveats to consider when using this method to infer transduction is that although it permits the detection of a gene within a single cell, the gene may not be expressed or integrated into the host genome (e.g., abortive transductants). In fact, the detection of abortive transductants by this method would presumably contribute to the considerably higher rates of transduction seen using this method as well as the ability of a given transducing particle to transfer bacterial genes to a broad range of bacterial hosts (Kenzaka et al. 2010).

Genotypic Changes of the Transducing Particle The revolutionary advances in PCR and qPCR techniques have opened the door to sensitive methods to detect transducing particles by virtue of the bacterial DNA they contain either from a single phage lysate or from environmental samples containing a mixture of phages. Several groups have used this method to detect bacterial genes in phage samples from water (Colomer-Lluch et al. 2011), soil (Ross and Topp 2015),

178

C. L. Schneider

sludge (Calero-Caceres et al. 2014), and purified phage lysates (Chlebowicz et al. 2014). Although the methods differ slightly depending on the source of particles, most are similar in many respects to the approach originally published by Sander and Schmeiger (2001). In general, phage particles are purified in some manner to remove contaminating bacteria and non-encapsidated bacterial DNA so that PCR detection of bacterial genes can be used as an indication of the presence of transducing DNA. Initially the lysates or environmental samples are filtered (0.22 or 0.45 μm) to remove contaminating bacterial cells. Environmental samples are typically further concentrated either by precipitation with PEG or concentration using centrifugal filters. Chloroform treatment is occasionally included following filtration to fully disrupt infected cell remnants but is often omitted to prevent the disruption of enveloped phages from uncharacterized environmental samples. The lysates are then treated with DNase to remove contaminating bacterial DNA that is not packaged in a phage particle, and the DNase is then heat inactivated, or the particles are further purified by CsCl density centrifugation. Phage capsids are then disrupted by treatment with proteinase K and SDS, followed by the purification of nucleic acid using either traditional methods (phenol-chloroform-isoamyl extraction and ethanol precipitation) or commercially available kits. The samples are then analyzed by PCR or qPCR to detect bacteria-specific genes. The reader is referred to the following papers for a more in-depth description of these methods (Ross and Topp 2015; Colomer-Lluch et al. 2011; Calero-Caceres et al. 2014). If the results are intended to infer the potential for transduction by these phage particles, then it is essential to include a variety of controls to demonstrate that all contaminating bacterial DNA has been removed prior to PCR. One control commonly used by researchers to demonstrate the lack of contaminating bacterial DNA is to test samples that have been treated with DNase, but not yet de-encapsidated, for the presence of the r16S gene using PCR (Colomer-Lluch et al. 2011). Samples that are negative for a 16S PCR product prior to de-encapsidation are commonly inferred to be free of contaminating bacterial DNA. There are, however, other causes for a negative result, such as inhibition of the PCR reaction by environmental contaminants. PCR inhibition can easily be addressed by including two positive controls, one of which includes the phage sample being tested and the control template and one that contains the control template only (Colomer-Lluch et al. 2011). Some researchers also include a sample in which a known amount of plasmid or other bacterial DNA is added to the phage lysate before DNase treatment (Chlebowicz et al. 2014). This provides a control to verify that the DNase treatment was effective by demonstrating the presence of the DNA by PCR before DNase treatment but not after. This method has certain distinct advantages, the biggest of which is that it provides a sensitive method to detect transducing particles in a host-independent manner. This permits the analysis of mixtures of phages from environmental samples without introducing the sample bias that occurs when using enrichment protocols. Additionally, researchers are able to determine absolute concentrations (gene copies per ml) of bacterial genes packaged within phage particles using quantitative PCR.

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

179

One disadvantage is that although the method detects the presence of bacterial genes within a DNase protected structure, this structure may or may not be a bona fide phage capsid capable of transduction. Purified phage particles can also be subjected to metagenomic analysis to infer the presence of transducing particles through the detection of bacterial genes. These methods are similar to those described above, but typically the particles are purified through successive filtrations followed by CsCl density centrifugation (Thurber et al. 2009; Reyes et al. 2010) prior to de-encapsidation and purification of the nucleic acid. As with the PCR method described above, the inclusion of similar controls to ensure complete removal of contaminating bacterial DNA is vital. The purified viral nucleic acid is then subjected to shotgun sequencing, and bacterial genes are detected using bioinformatics. The advantage to this method over the previous method is that since the entire metagenome is determined, the findings are not limited to genes selected by the researcher. This method has similar limitations to the PCR approach (false positives due to contamination, inability to demonstrate bona fide transduction), but additionally the false-positive rate might be higher as a result of the stringency cutoffs used in the bioinformatics analysis. A study highlighting the need for caution when interpreting data like these evaluated the bioinformatics criteria used to predict antibiotic resistance genes (ARGs) sequenced from phage DNA and evaluated four predicted ARGs (Enault et al. 2016). The results of this study suggest the need for more stringent bioinformatics cutoffs (>70 bit-score thresholds for exploratory searches) to reduce the false-positive rate. Additionally, the fact that none of the four predicted ARGs they characterized conferred antibiotic resistance to E. coli underscores the risk of relying completely on bioinformatics predictions.

Potential Risks to Human Health Posed by Transduction Although Zinder and Lederberg first demonstrated that bacteriophages mediate horizontal gene transfer in 1951, the important role transduction plays in shaping the evolution of bacteria has only recently come to light. Thanks to the development of more sensitive techniques to quantify viruses in various environmental samples, we discovered the sheer abundance of phages in several environments in which their bacterial hosts are also prevalent (e.g., aquatic and human colon). In addition, the development of shotgun sequencing and its application to metagenomic analysis of the viral compartment (i.e., virome) of various environments has revealed that bacteriophages provide a reservoir of novel genes of unknown function; in fact, bacteriophages have been described as the “dark matter of the biological universe” (Pedulla et al. 2003). Although phages can affect the physiology of their bacterial host in many important ways, the focus for this final section will be confined to the influence of transduction on issues important to human health such as the spread of antibiotic resistance genes and the risk that transduction might pose to usefulness of phage therapy.

180

C. L. Schneider

Transduction and the Spread of Antibiotic Resistance Genes According to the Center for Disease Control and Prevention (CDC), roughly two million people become infected with antibiotic-resistant bacteria in the USA each year, and of those, roughly 23,000 die as a consequence of these bacterial infections. The alarmingly rapid acquisition of antibiotic resistance genes by human pathogens poses a global public health risk that threatens to undermine the significant medical advances that have occurred in the last 50 years. As of 2013, the CDC classified 15 antibiotic-resistant bacterial human pathogens as either “urgent” or “serious” threats to human health. To underscore the importance of this issue, in 2016 the US Congress appropriated $160 million to the CDC to combat antibiotic resistance. Clearly, an understanding of the source of antibiotic resistance genes and the means by which they can be transferred to human pathogens is vital in combatting this issue. Recent studies indicate that not only are antibiotic resistance genes (ARGs) ancient, but they are also prevalent in commensal bacteria from environments wherein they likely protect individual bacteria from bioactive molecules produced by other bacteria (for review, see Perry et al. 2014; von Wintersdorff et al. 2016; Perry et al. 2016; van Schaik 2015; Gaze et al. 2013). Several groups have identified ARGs in permafrost samples ranging from 5000–30,000 years old (Perron et al. 2015; D'Costa et al. 2011; Petrova et al. 2009), demonstrating their origins predate the use of antibiotics in human medicine. The mobilization and transfer of ARGs via HGT is also likely to be ancient, predating the clinical use of antibiotics (for review, see von Wintersdorf 2016). However, the overuse of antibiotics in humans and animals has certainly increased the prevalence of ARGs (Knapp et al. 2010), and it can increase the prevalence of ARGs on mobile genetic elements like plasmids (Datta and Hughes 1983). Aquatic ecosystems may provide an excellent environment to facilitate the spread of ARGs by transduction (Balcazar 2014). Not only are naturally occurring ARGs likely to be prevalent within the environmental bacterial species, but the pollutants and antibiotics that enter the waterways from wastewater treatment discharges and agricultural runoff provide a continual selective pressure that not only maintains ARGs but promotes the mobilization of ARGs and the evolution of new ARGs (for review, see Marti et al. 2014b). Bacteriophages are also prevalent in aquatic environments and oftentimes outnumber their bacterial hosts by a factor of 10 (for review, see Clokie et al. 2011). Thus, even if the transduction rate for an individual phage is low, the sheer number of phages present in these environments would make transduction a significant contributor to HGT. Several studies now have suggested that the role of transduction in the spread of ARGs has been underestimated (for review, see Balcazar 2014). Generalized transduction frequencies of 108–105 per PFU have been demonstrated in freshwater and marine systems using culture-based detection methods (for review, see Weinbauer and Rassoulzadegan 2004), and frequencies of 104–103 per PFU have been measured using non-culture-based methods (Kenzaka et al. 2010). Given that human bacterial pathogens can also enter these waterways either through

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

181

inadequate removal by traditional wastewater treatment procedures or by direct contamination, environmental waterways could play an important role in the spread of antibiotic resistance to clinically relevant bacteria (for review, see Fletcher 2015; Marti et al. 2014a).

Transduction and the Implications to Phage Therapy The risk that antibiotic-resistant bacteria pose to human health and the future of medical treatments is significant. The urgency in combatting antibiotic resistance is underscored by the congressional approval for $160 million to support the CDC’s National Action Plan for Combatting Antibiotic Resistant Bacteria (https://www. cdc.gov/drugresistance/solutions-initiative/). The National Action Plan includes five goals, one of which is to “Accelerate basic and applied research and development for new antibiotics, other therapeutics, and vaccines” (https://www.whitehouse.gov/ sites/default/files/docs/national_action_plan_for_combating_antibotic-resistant_bacteria. pdf). The use of phage to kill bacterial pathogens (phage therapy) is an alternative therapeutic approach receiving renewed interest (see chapters ▶ “Regulatory Aspects of the Therapeutic Use of Bacteriophages: Europe,” and ▶ “Clinical Trials of Bacteriophage Therapeutics”). Phage therapy has been used safely in humans for almost a century, predominantly in France, Georgia, and Poland. The reader is referred to several in-depth reviews on the usefulness of phage therapy (Abedon 2012; Brussow 2005; Wittebole et al. 2014; Kutter et al. 2010; Hraiech et al. 2015; Sulakvelidze et al. 2001; Loc-Carrillo and Abedon 2011). Phage therapy offers several advantages over traditional antibiotics. Given the narrow host range an individual phage typically possesses, phage therapy can provide a more targeted approach to specifically kill pathogenic bacteria without incurring global changes to the microbiota that typically follow antibiotic treatment (Bruttin and Brussow 2005; Chibani-Chennoufi et al. 2004; Sarker et al. 2012). Given the plethora of recent studies suggesting links between dysbiosis and a wide variety of human diseases, the targeted approach offered by phage therapy has considerable appeal. Additionally, since the replication of a phage requires the presence of a susceptible bacterial host, the “dose” is automatically controlled based on accessibility of the phage to their host (Abedon and Thomas-Abedon 2010). Significantly, although bacteria can develop resistance to an individual phage, phage therapy has two advantages that should reduce the likelihood of bacterial resistance. Since the mechanisms by which phages lyse their bacterial hosts vary, the use of a mixture of diverse phages that target a particular bacterial host (phage cocktail) should limit the ability of the bacteria to develop resistance to all phages within the cocktail. Secondly, unlike antibiotic chemicals, phages evolve in response to the evolution of their bacterial hosts. This means that even if the bacteria become resistant to the wild-type phage, there may be mutants within the phage population that can overcome this resistance mechanism. Also, a bacterium that acquires phage resistance may also incur a fitness cost making it unable to compete for resources with other bacterial species in the microbiome that are not

182

C. L. Schneider

susceptible to infection by the phage (Brussow 2005). These and other features of phage therapy make it an attractive alternative to combat bacterial infections, especially given the increasing prevalence of antibiotic-resistant bacterial pathogens. The ability of phage to transduce virulence genes including antibiotic resistance genes is a potential risk to consider when implementing phage therapy. The risk of transduction can be significantly minimized through careful screening and selection of phages for phage therapy. Several criteria for the isolation and characterization of phages for use in phage therapy are recommended (Łobocka et al. 2014; WeberDabrowska et al. 2016; Gill and Hyman 2010). The use of non-temperate phage, with minimal capacity to perform generalized transduction, should be sufficient to minimize the concerns of transduction. Non-temperate phages are less likely to perform transduction because they often degrade the host DNA for use as building blocks for their own replication (Waddell et al. 2009). This minimizes the potential for generalized transduction by eliminating or severely limiting the amount of host DNA present during viral DNA packaging (Masters 2004). Phages should be evaluated phenotypically (plaque morphology, etc.) as well as genotypically (absence of integrase genes) to verify they are in fact non-temperate (Łobocka et al. 2014). Phages should also be evaluated for their propensity to perform generalized transduction either by their ability to transfer antibiotic resistance genes or by detecting host DNA within phage particles (Łobocka et al. 2014). Finally, an understanding of a phage’s DNA packaging mechanism might be useful in evaluating the risk of generalized transduction. Phages that initiate packaging of multiple phage heads from a single sequence (headful packaging) are more likely to perform generalized transduction than those phages that require a unique packaging sequence to initiate packaging of each phage head (Klumpp et al. 2008; Casjens and Gilcrease 2009). Interest in phage therapy is increasing not only due to the global antibiotic resistance crisis but also in response to studies highlighting the importance of the human microbiota in human health and disease (Gorski et al. 2009; Young and Gill 2015; Wittebole et al. 2014; Viertel et al. 2014). While phage therapy clearly has some significant technical and regulatory hurdles to overcome before these approaches gain approval for use in humans in Western countries like the USA and the UK (Skurnik et al. 2007; Servick 2016; Gilmore 2012), the potential benefits seem to outweigh the potential risks at this point (Thiel 2004). When the amount of DNA transduction naturally occurring in the environment is considered, the potential risks of transduction to phage therapy seem manageable. As Ryland Young stated, “Unless you’re completely compulsive, it doesn’t make a whole lot of sense to me to worry about transduction” (Thiel 2004).

Conclusions The vast diversity observed in bacteria is generated by both mutation and a variety of mechanisms of HGT, including transduction. Understanding the role that transduction plays in driving bacterial evolution in various environments is important in safeguarding human health. Several recent studies suggest that the contribution of

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

183

transduction to the spread of antibiotic resistance genes has likely been underestimated in environments like waterways and the human gut where phages are abundant and human pathogenic bacteria at least transiently interact with endogenous bacterial species. Commensals and environmental bacteria are important reservoirs of antibiotic resistance genes (for review, see von Wintersdorff et al. 2016), and the accidental capture of these genes by generalized transducing phages offers certain advantages over either transformation or conjugation. The stability of transducing phage particles protects the ARGs from degradation and permits transfer of the ARGs between bacterial species that do not necessarily occupy the same location at the same time. Additionally, since certain types of antibiotics increase the production of generalized transducing particles by some bacteria through the induction of prophages, the clinical or agricultural use of these antibiotics and their persistence in environmental waterways may also increase the role of transduction in mediating ARG spread (Bearson and Brunelle 2015; Bearson et al. 2014; Beaber et al. 2004). Newer methodologies using metagenomic approaches or PCR have been useful in detecting ARGs in phage particles, and CRISPN-FISH has been used to demonstrate the transfer of ARGs to an individual bacterium by a transducing particle. These techniques open the door to better characterize the importance of transduction in mediating HGT in these environments, especially when coupled with more classical methods to detect bona fide transduction. Although transduction was discovered more than 50 years ago, the contribution of transduction to bacterial HGT is still an enigma worthy of continued scientific research.

Cross-References ▶ Clinical Trials of Bacteriophage Therapeutics ▶ Phage Infection and Lysis ▶ Regulatory Aspects of the Therapeutic Use of Bacteriophages: Europe ▶ Temperate Phages, Prophages, and Lysogeny

References Abedon ST (2012) Phage-therapy best practices. In: Hyman P, Abedon ST (eds) Bacteriophages in health and disease. CABI Press, Wallingford, pp 256–272 Abedon ST, Thomas-Abedon C (2010) Phage therapy pharmacology. Curr Pharm Biotechnol 11(1):28–47 Arber W (1974) DNA modification and restriction. Prog Nucleic Acid Res Mol Biol 14(0):1–37 Austin S, Ziese M, Sternberg N (1981) A novel role for site-specific recombination in maintenance of bacterial replicons. Cell 25(3):729–736. doi:0092-8674(81)90180-X Bachi B, Arber W (1977) Physical mapping of BglII, BamHI, EcoRI, HindIII and PstI restriction fragments of bacteriophage P1 DNA. Mol Gen Genet 153(3):311–324 Balcazar JL (2014) Bacteriophages as vehicles for antibiotic resistance genes in the environment. PLoS Pathog 10(7):e1004219. https://doi.org/10.1371/journal.ppat.1004219 Ball CA, Johnson RC (1991) Multiple effects of Fis on integration and the control of lysogeny in phage lambda. J Bacteriol 173(13):4032–4038

184

C. L. Schneider

Beaber JW, Hochhut B, Waldor MK (2004) SOS response promotes horizontal dissemination of antibiotic resistance genes. Nature 427(6969):72–74. https://doi.org/10.1038/nature02241 Bearson BL, Allen HK, Brunelle BW et al (2014) The agricultural antibiotic carbadox induces phage-mediated gene transfer in Salmonella. Front Microbiol 5:52. https://doi.org/10.3389/ fmicb.2014.00052 Bearson BL, Brunelle BW (2015) Fluoroquinolone induction of phage-mediated gene transfer in multidrug-resistant Salmonella. Int J Antimicrob Agents 46(2):201–204. https://doi.org/ 10.1016/j.ijantimicag.2015.04.008 Benson NR, Roth J (1997) A Salmonella phage-P22 mutant defective in abortive transduction. Genetics 145(1):17–27 Bertani G, Weigle JJ (1953) Host controlled variation in bacterial viruses. J Bacteriol 65(2):113–121 Billard-Pomares T, Fouteau S, Jacquet ME et al (2014) Characterization of a P1-like bacteriophage carrying an SHV-2 extended-spectrum beta-lactamase from an Escherichia coli strain. Antimicrob Agents Chemother 58(11):6550–6557. https://doi.org/10.1128/AAC.03183-14 Black LW (1989) DNA packaging in dsDNA bacteriophages. Annu Rev Microbiol 43:267–292. https://doi.org/10.1146/annurev.mi.43.100189.001411 Botstein D, Matz MJ (1970) A recombination function essential to the growth of bacteriophage P22. J Mol Biol 54(3):417–440. doi:0022-2836(70)90119-1 Boulter J, Lee N (1975) Isolation of specialized transducing bacteriophage lambda carrying genes of the L-arabinose operon of Escherichia coli B/r. J Bacteriol 123(3):1043–1054 Brimacombe CA, Ding H, Beatty JT (2014) Rhodobacter capsulatus DprA is essential for RecAmediated gene transfer agent (RcGTA) recipient capability regulated by quorum-sensing and the CtrA response regulator. Mol Microbiol 92(6):1260–1278. https://doi.org/10.1111/mmi.12628 Brimacombe CA, Ding H, Johnson JA et al (2015) Homologues of genetic transformation DNA import genes are required for rhodobacter capsulatus gene transfer agent recipient capability regulated by the response regulator CtrA. J Bacteriol 197(16):2653–2663. https://doi.org/ 10.1128/JB.00332-15 Brussow H (2005) Phage therapy: the Escherichia coli experience. Microbiology 151(Pt 7):2133–2140. https://doi.org/10.1099/mic.0.27849-0 Bruttin A, Brussow H (2005) Human volunteers receiving Escherichia coli phage T4 orally: a safety test of phage therapy. Antimicrob Agents Chemother 49(7):2874–2878. https://doi.org/10.1128/ AAC.49.7.2874-2878.2005 Bukhari AI (1976) Bacteriophage mu as a transposition element. Annu Rev Genet 10:389–412. https://doi.org/10.1146/annurev.ge.10.120176.002133 Bukhari AI, Taylor AL (1975) Influence of insertions on packaging of host sequences covalently linked to bacteriophage Mu DNA. Proc Natl Acad Sci U S A 72(11):4399–4403 Calero-Caceres W, Melgarejo A, Colomer-Lluch M et al (2014) Sludge as a potential important source of antibiotic resistance genes in both the bacterial and bacteriophage fractions. Environ Sci Technol 48(13):7602–7611. https://doi.org/10.1021/es501851s Campbell A (1957) Transduction and segregation in Escherichia coli K12. Virology 4(2):366–384. doi:0042-6822(57)90070-3 Campbell A (1994) Comparative molecular biology of lambdoid phages. Annu Rev Microbiol 48:193–222. https://doi.org/10.1146/annurev.mi.48.100194.001205 Caro L, Berg CM (1971) P1 Transduction. Methods Enzymol 21D:444–458 Casjens S, Hayden M (1988) Analysis in vivo of the bacteriophage P22 headful nuclease. J Mol Biol 199(3):467–474. doi:0022-2836(88)90618-3 Casjens S, Sampson L, Randall S et al (1992) Molecular genetic analysis of bacteriophage P22 gene 3 product, a protein involved in the initiation of headful DNA packaging. J Mol Biol 227(4):1086–1099. doi:0022-2836(92)90523-M Casjens SR, Gilcrease EB (2009) Determining DNA packaging strategy by analysis of the termini of the chromosomes in tailed-bacteriophage virions. Methods Mol Biol 502:91–111. https://doi. org/10.1007/978-1-60327-565-1_7 Casjens SR, Hendrix RW (2015) Bacteriophage lambda: Early pioneer and still relevant. Virology 479–480:310–330. https://doi.org/10.1016/j.virol.2015.02.010

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

185

Catalano CE, Cue D, Feiss M (1995) Virus DNA packaging: the strategy used by phage lambda. Mol Microbiol 16(6):1075–1086 Chelala CA, Margolin P (1974) Effects of deletions on cotransduction linkage in Salmonella typhimurium: evidence that bacterial chromosome deletions affect the formation of transducing DNA fragments. Mol Gen Genet 131(2):97–112 Chen J, Ram G, Penades JR et al (2015) Pathogenicity island-directed transfer of unlinked chromosomal virulence genes. Mol Cell 57(1):138–149. https://doi.org/10.1016/j. molcel.2014.11.011 Chibani-Chennoufi S, Sidoti J, Bruttin A et al (2004) In vitro and in vivo bacteriolytic activities of Escherichia coli phages: implications for phage therapy. Antimicrob Agents Chemother 48(7):2558–2569. https://doi.org/10.1128/AAC.48.7.2558-2569.2004 Chlebowicz MA, Maslanova I, Kuntova L et al (2014) The Staphylococcal Cassette Chromosome mec type V from Staphylococcus aureus ST398 is packaged into bacteriophage capsids. Int J Med Microbiol 304(5–6):764–774. https://doi.org/10.1016/j.ijmm.2014.05.010 Choi W, Jang S, Harshey RM (2014) Mu transpososome and RecBCD nuclease collaborate in the repair of simple Mu insertions. Proc Natl Acad Sci U S A 111(39):14112–14117. https://doi.org/ 10.1073/pnas.1407562111 Chung ST, Greenberg GR (1973) Loss of an essential function of Escherichia coli by deletions in the thyA region. J Bacteriol 116(3):1145–1149 Clokie MR, Millard AD, Letarov AV et al (2011) Phages in nature. Bacteriophage 1(1):31–45. https://doi.org/10.4161/bact.1.1.14942 Cohen G, Sternberg N (1989) Genetic analysis of the lytic replicon of bacteriophage P1. I Isolation and partial characterization. J Mol Biol 207(1):99–109. doi:0022-2836(89)90443-9 Colomer-Lluch M, Jofre J, Muniesa M (2011) Antibiotic resistance genes in the bacteriophage DNA fraction of environmental samples. PLoS One 6(3):e17549. https://doi.org/10.1371/journal.pone.0017549 Court DL, Oppenheim AB, Adhya SL (2007) A new look at bacteriophage lambda genetic networks. J Bacteriol 189(2):298–304. https://doi.org/10.1128/JB.01215-06 Datta N, Hughes VM (1983) Plasmids of the same Inc groups in Enterobacteria before and after the medical use of antibiotics. Nature 306(5943):616–617 Davis BD (1950) Nonfiltrability of the agents of genetic recombination in Escherichia coli. J Bacteriol 60(4):507–508 D'Costa VM, King CE, Kalan L et al (2011) Antibiotic resistance is ancient. Nature 477(7365):457–461. https://doi.org/10.1038/nature10388 Drexler H, Christensen JR (1979) Transduction of bacteriophage lambda by bacteriophage T1. J Virol 30(2):543–550 Ebel-Tsipis J, Botstein D, Fox MS (1972a) Generalized transduction by phage P22 in Salmonella typhimurium. I. Molecular origin of transducing DNA. J Mol Biol 71(2):433–448. doi:00222836(72)90361-0 Ebel-Tsipis J, Fox MS, Botstein D (1972b) Generalized transduction by bacteriophage P22 in Salmonella typhimurium. II. Mechanism of integration of transducing DNA. J Mol Biol 71(2):449–469. doi:0022-2836(72)90362-2 Echols H, Court DL (1971) The role of helper phage in gal transduction. In: Hershey AD (ed) The bacteriophage. Lambda The Cold Spring Harbor Laboratory, Cold Spring Harbor, pp 701–710 Echols H, Guarneros G (1983) Control of integration and excision. In: Hendrix RW, Roberts JW, Stahl FW et al (eds) Lambda II. Cold Spring Harbor Laboratory, Cold Spring Harbor, pp 75–92 Enault F, Briet A, Bouteille L et al (2016) Phages rarely encode antibiotic resistance genes: a cautionary tale for virome analyses. ISME J 11(1):237–247. https://doi.org/10.1038/ ismej.2016.90 Feiss M, Fisher RA, Crayton MA et al (1977) Packaging of the bacteriophage lambda chromosome: effect of chromosome length. Virology 77(1):281–293 Felsenstein J (1974) The evolutionary advantage of recombination. Genetics 78(2):737–756 Fletcher S (2015) Understanding the contribution of environmental factors in the spread of antimicrobial resistance. Environ Health Prev Med 20(4):243–252. https://doi.org/10.1007/ s12199-015-0468-0

186

C. L. Schneider

Gaze WH, Krone SM, Larsson DG et al (2013) Influence of humans on evolution and mobilization of environmental antibiotic resistome. Emerg Infect Dis 19(7). https://doi.org/10.3201/ eid1907.120871 Genthner FJ, Wall JD (1984) Isolation of a recombination-deficient mutant of Rhodopseudomonas capsulata. J Bacteriol 160(3):971–975 Gill JJ, Hyman P (2010) Phage choice, isolation, and preparation for phage therapy. Curr Pharm Biotechnol 11(1):2–14 Gilmore BF (2012) Bacteriophages as anti-infective agents: recent developments and regulatory challenges. Expert Rev Anti-Infect Ther 10(5):533–535. https://doi.org/10.1586/eri.12.30 Gorski A, Miedzybrodzki R, Borysowski J et al (2009) Bacteriophage therapy for the treatment of infections. Curr Opin Investig Drugs 10(8):766–774 Gottesman ME, Weisberg RA (2004) Little lambda, who made thee? Microbiol Mol Biol Rev 68(4):796–813. https://doi.org/10.1128/MMBR.68.4.796-813.2004 Griffith F (1928) The significance of pneumococcal types. J Hyg (Lond) 27(2):113–159 Groenen MA, van de Putte P (1985) Mapping of a site for packaging of bacteriophage Mu DNA. Virology 144(2):520–522 Hanks MC, Newman B, Oliver IR et al (1988) Packaging of transducing DNA by bacteriophage P1. Mol Gen Genet 214(3):523–532 Harel J, Duplessis L, Kahn JS et al (1990) The cis-acting DNA sequences required in vivo for bacteriophage Mu helper-mediated transposition and packaging. Arch Microbiol 154(1):67–72 Harriman PD (1972) A single-burst analysis of the production of P1 infectious and transducing particles. Virology 48(2):595–600. doi:0042-6822(72)90071-2 Harshey RM (1988) Phage Mu. In: Calendar R (ed) The bacteriophages, vol 1. Plenum Press, New York, pp 193–234 Harshey RM (2012) The Mu story: how a maverick phage moved the field forward. Mob DNA 3(1):21-8753-3-21. https://doi.org/10.1186/1759-8753-3-21 Harshey RM (2014) Transposable phage Mu. Microbiol Spectr 2(5). https://doi.org/10.1128/microbiolspec.MDNA3-0007-2014 Harshey RM, Bukhari AI (1983) Infecting bacteriophage mu DNA forms a circular DNA-protein complex. J Mol Biol 167(2):427–441 Hendrix RW, Roberts J, Stahl F et al (eds) (1983) Lambda II. Cold Spring Harbor Laboratory, Cold Spring Harbor Hendrix RW (2003) Bacteriophage genomics. Curr Opin Microbiol 6(5):506–511. doi: S1369527403001152 Hershey AD, Burgi E (1965) Complementary structure of interacting sites at the ends of lambda DNA molecules. Proc Natl Acad Sci U S A 53:325–328 Hershey AD, Burgi E, Ingraham L (1963) Cohesion of DNA molecules isolated from phage lambda. Proc Natl Acad Sci U S A 49(5):748–755 Howe MM (1973) Transduction by bacteriophage MU-1. Virology 55(1):103–117 Howe MM, Bade EG (1975) Molecular biology of bacteriophage mu. Science 190(4215):624–632 Hraiech S, Bregeon F, Rolain JM (2015) Bacteriophage-based therapy in cystic fibrosis-associated Pseudomonas aeruginosa infections: rationale and current status. Drug Des Devel Ther 9:3653–3663. https://doi.org/10.2147/DDDT.S53123 Huang H, Masters M (2014) Bacteriophage P1 pac sites inserted into the chromosome greatly increase packaging and transduction of Escherichia coli genomic DNA. Virology 468–470:274–282. https://doi.org/10.1016/j.virol.2014.07.029 Hynes AP, Mercer RG, Watton DE et al (2012) DNA packaging bias and differential expression of gene transfer agent genes within a population during production and release of the Rhodobacter capsulatus gene transfer agent, RcGTA. Mol Microbiol 85(2):314–325. https://doi.org/10.1111/ j.1365-2958.2012.08113.x Iida S, Arber W (1977) Plaque forming specialized transducing phage P1: isolation of P1CmSmSu, a precursor of P1Cm. Mol Gen Genet 153(3):259–269 Iida S, Hanni C, Echarti C et al (1981) Is the IS1-flanked r-determinant of the R plasmid NR1 a transposon? J Gen Microbiol 126(2):413–425. https://doi.org/10.1099/00221287126-2-413

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

187

Iida S, Meyer J, Arber W (1978) The insertion element IS1 is a natural constituent of coliphage P1 DNA. Plasmid 1(3):357–365. doi:0147-619X(78)90051-3 Ikeda H, Shimizu H, Ukita T et al (1995) A novel assay for illegitimate recombination in Escherichia coli: stimulation of lambda bio transducing phage formation by ultra-violet light and its independence from RecA function. Adv Biophys 31:197–208. doi:0065227X95993923 Ikeda H, Tomizawa J (1968) Prophage P1, and extrachromosomal replication unit. Cold Spring Harb Symp Quant Biol 33:791–798 Ikeda H, Tomizawa JI (1965a) Transducing fragments in generalized transduction by phage P1. I. Molecular origin of the fragments. J Mol Biol 14(1):85–109 Ikeda H, Tomizawa JI (1965b) Transducing fragments in generalized transduction by phage P1. II. Association of DNA and protein in the fragments. J Mol Biol 14(1):110–119 Jacob F, Monod J (1961) Genetic regulatory mechanisms in the synthesis of proteins. J Mol Biol 3:318–356 Jain A, Srivastava P (2013) Broad host range plasmids. FEMS Microbiol Lett 348(2):87–96. https:// doi.org/10.1111/1574-6968.12241 Kallmeyer J, Pockalny R, Adhikari RR et al (2012) Global distribution of microbial abundance and biomass in subseafloor sediment. Proc Natl Acad Sci U S A 109(40):16213–16216. https://doi. org/10.1073/pnas.1203849109 Katz L (1970) Selection of araB and araC mutants of Escherichia coli B-r by resistance to ribitol. J Bacteriol 102(2):593–595 Kenzaka T, Tamaki S, Yamaguchi N et al (2005) Recognition of individual genes in diverse microorganisms by cycling primed in situ amplification. Appl Environ Microbiol 71(11):7236–7244. https://doi.org/10.1128/AEM.71.11.7236-7244.2005 Kenzaka T, Tani K, Nasu M (2010) High-frequency phage-mediated gene transfer in freshwater environments determined at single-cell level. ISME J 4(5):648–659. https://doi.org/10.1038/ ismej.2009.145 Kenzaka T, Tani K, Sakotani A et al (2007) High-frequency phage-mediated gene transfer among Escherichia coli cells, determined at the single-cell level. Appl Environ Microbiol 73(10):3291–3299. https://doi.org/10.1128/AEM.02890-06 Klumpp J, Dorscht J, Lurz R et al (2008) The terminally redundant, nonpermuted genome of Listeria bacteriophage A511: a model for the SPO1-like myoviruses of gram-positive bacteria. J Bacteriol 190(17):5753–5765. https://doi.org/10.1128/JB.00461-08 Knapp CW, Dolfing J, Ehlert PA et al (2010) Evidence of increasing antibiotic resistance gene abundances in archived soils since 1940. Environ Sci Technol 44(2):580–587. https://doi.org/ 10.1021/es901221x Kondo E, Mitsuhashi S (1964) Drug resistance of enteric bacteria. Iv. Active transducing bacteriophage P1 Cm produced by the combination of R factor with bacteriophage P1. J Bacteriol 88:1266–1276 Kutter E, De Vos D, Gvasalia G et al (2010) Phage therapy in clinical practice: treatment of human infections. Curr Pharm Biotechnol 11(1):69–86 Kwoh DY, Kemper J (1978) Bacteriophage P22-mediated specialized transduction in Salmonella typhimurium: high frequency of aberrant prophage excision. J Virol 27(3):519–534 Lang AS, Zhaxybayeva O, Beatty JT (2012) Gene transfer agents: phage-like elements of genetic exchange. Nat Rev Microbiol 10(7):472–482. https://doi.org/10.1038/nrmicro2802 Lederberg EM (1951) Lysogeny in E. coli K-12. Genetics 36:560 Lederberg EM, Lederberg J (1953) Genetic studies of lysogenicity in Escherichia coli. Genetics 38(1):51–64 Lederberg J, Tatum EL (1946) Gene recombination in Escherichia coli. Nature 158(4016):558 Lindahl G, Sironi G, Bialy H et al (1970) Bacteriophage lambda; abortive infection of bacteria lysogenic for phage P2. Proc Natl Acad Sci U S A 66(3):587–594 Lindsay JA, Holden MT (2006) Understanding the rise of the superbug: investigation of the evolution and genomic variation of Staphylococcus aureus. Funct Integr Genomics 6(3):186–201. https://doi.org/10.1007/s10142-005-0019-7 Lindsay JA, Ruzin A, Ross HF et al (1998) The gene for toxic shock toxin is carried by a family of mobile pathogenicity islands in Staphylococcus aureus. Mol Microbiol 29(2):527–543

188

C. L. Schneider

Łobocka M, Hejnowicz MS, Gągała U et al (2014) The first step to bacteriophage therapy – how to choose the correct phage. In: Borysowski J, Międzybrodzki R, Górski A (eds) Phage therapy: current research and applications. Caister Academic Press, Norfolk, pp 23–67 Lobocka MB, Rose DJ, Plunkett G 3rd et al (2004) Genome of bacteriophage P1. J Bacteriol 186(21):7032–7068. https://doi.org/10.1128/JB.186.21.7032-7068.2004 Loc-Carrillo C, Abedon ST (2011) Pros and cons of phage therapy. Bacteriophage 1(2):111–114. https://doi.org/10.4161/bact.1.2.14590 Lu SD, Lu D, Gottesman M (1989) Stimulation of IS1 excision by bacteriophage P1 ref function. J Bacteriol 171(6):3427–3432 Maiques E, Ubeda C, Campoy S et al (2006) beta-lactam antibiotics induce the SOS response and horizontal transfer of virulence factors in Staphylococcus aureus. J Bacteriol 188(7):2726–2729. https://doi.org/10.1128/JB.188.7.2726-2729.2006 Marrs B (1974) Genetic recombination in Rhodopseudomonas capsulata. Proc Natl Acad Sci U S A 71(3):971–973 Marti E, Variatza E, Balcazar JL (2014a) Bacteriophages as a reservoir of extended-spectrum betalactamase and fluoroquinolone resistance genes in the environment. Clin Microbiol Infect 20(7): O456-9. https://doi.org/10.1111/1469-0691.12446 Marti E, Variatza E, Balcazar JL (2014b) The role of aquatic ecosystems as reservoirs of antibiotic resistance. Trends Microbiol 22(1):36–41. https://doi.org/10.1016/j.tim.2013.11.001 Masters M (1977) The frequency of P1 transduction of the genes of Escherichia coli as a function of chromosomal position: preferential transduction of the origin of replication. Mol Gen Genet 155(2):197–202 Masters M (1996) Generalized transduction. In: Neidhardt FC, Curtiss R III, Ingraham JJ et al (eds) Escherichia coli and salmonella typhimurium: cellular and molecular biology, 2nd edn. American Society for Microbiology, Washington DC, pp 2421–2448 Masters M (2004) Transduction: host DNA transfer by bacteriophages. In: Schaechter M (ed) The desk encyclopedia of microbiology. Elsevier/Academic Press, London, pp 1000–1012 McDaniel LD, Young EC, Ritchie KB et al (2012) Environmental factors influencing gene transfer agent (GTA) mediated transduction in the subtropical ocean. PLoS One 7(8):e43506. https://doi. org/10.1371/journal.pone.0043506 Mise K, Arber W (1976) Plaque-forming transducing bacteriophage P1 derivatives and their behaviour in lysogenic conditions. Virology 69(1):191–205. doi:0042-6822(76)90206-3 Morse ML, Lederberg EM, Lederberg J (1956a) Transduction in Escherichia coli K-12. Genetics 41(1):142–156 Morse ML, Lederberg EM, Lederberg J (1956b) Transductional heterogenotes in Escherichia coli. Genetics 41(5):758–779 Muller HJ (1964) The relation of recombination to mutational advance. Mutat Res 106:2–9 Nash HA (1981) Integration and excision of bacteriophage lambda: the mechanism of conservation site specific recombination. Annu Rev Genet 15:143–167. https://doi.org/10.1146/annurev. ge.15.120181.001043 Newman BJ, Masters M (1980) The variation in frequency with which markers are transduced by phage P1 is primarily a result of discrimination during recombination. Mol Gen Genet 180(3):585–589 Novick RP (2003) Mobile genetic elements and bacterial toxinoses: the superantigen-encoding pathogenicity islands of Staphylococcus aureus. Plasmid 49(2):93–105. doi: S0147619X02001579 Novick RP, Ram G (2016) The Floating (Pathogenicity) Island: A Genomic Dessert. Trends Genet 32(2):114–126. https://doi.org/10.1016/j.tig.2015.11.005 Paul JH (2008) Prophages in marine bacteria: dangerous molecular time bombs or the key to survival in the seas? ISME J 2(6):579–589. https://doi.org/10.1038/ismej.2008.35 Pedulla ML, Ford ME, Houtz JM et al (2003) Origins of highly mosaic mycobacteriophage genomes. Cell 113(2):171–182. doi:S0092867403002332

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

189

Penades JR, Chen J, Quiles-Puchalt N et al (2015) Bacteriophage-mediated spread of bacterial virulence genes. Curr Opin Microbiol 23:171–178. https://doi.org/10.1016/j.mib.2014.11.019 Penades JR, Christie GE (2015) The phage-inducible chromosomal Islands: a family of highly evolved molecular parasites. Annu Rev Virol 2(1):181–201. https://doi.org/10.1146/annurevvirology-031413-085446 Perron GG, Whyte L, Turnbaugh PJ et al (2015) Functional characterization of bacteria isolated from ancient arctic soil exposes diverse resistance mechanisms to modern antibiotics. PLoS One 10(3):e0069533. https://doi.org/10.1371/journal.pone.0069533 Perry J, Waglechner N, Wright G (2016) The prehistory of antibiotic resistance. Cold Spring Harb Perspect Med 6(6). https://doi.org/10.1101/cshperspect.a025197 Perry JA, Westman EL, Wright GD (2014) The antibiotic resistome: what's new? Curr Opin Microbiol 21:45–50. https://doi.org/10.1016/j.mib.2014.09.002 Petrova M, Gorlenko Z, Mindlin S (2009) Molecular structure and translocation of a multiple antibiotic resistance region of a Psychrobacter psychrophilus permafrost strain. FEMS Microbiol Lett 296(2):190–197. https://doi.org/10.1111/j.1574-6968.2009.01635.x Poteete AR (1988) Bacteriophage P22. In: Calendar R (ed) The bacteriophages, vol 2. Plenum Press, New York, pp 647–681 Puspurs AH, Trun NJ, Reeve JN (1983) Bacteriophage Mu DNA circularizes following infection of Escherichia coli. EMBO J 2(3):345–352 Quiles-Puchalt N, Carpena N, Alonso JC et al (2014a) Staphylococcal pathogenicity island DNA packaging system involving cos-site packaging and phage-encoded HNH endonucleases. Proc Natl Acad Sci U S A 111(16):6016–6021. https://doi.org/10.1073/pnas.1320538111 Quiles-Puchalt N, Martinez-Rubio R, Ram G et al (2014b) Unravelling bacteriophage varphi11 requirements for packaging and transfer of mobile genetic elements in Staphylococcus aureus. Mol Microbiol 91(3):423–437. https://doi.org/10.1111/mmi.12445 Ram G, Chen J, Kumar K et al (2012) Staphylococcal pathogenicity island interference with helper phage reproduction is a paradigm of molecular parasitism. Proc Natl Acad Sci U S A 109(40):16300–16305. https://doi.org/10.1073/pnas.1204615109 Ram G, Chen J, Ross HF et al (2014) Precisely modulated pathogenicity island interference with late phage gene transcription. Proc Natl Acad Sci U S A 111(40):14536–14541. https://doi.org/ 10.1073/pnas.1406749111 Raya RR, H'bert EM (2009) Isolation of phage via induction of lysogens. Methods Mol Biol 501:23–32. https://doi.org/10.1007/978-1-60327-164-6_3 Reyes A, Haynes M, Hanson N et al (2010) Viruses in the faecal microbiota of monozygotic twins and their mothers. Nature 466(7304):334–338. https://doi.org/10.1038/nature09199 Roberts JW, Devoret R (1983) Lysogenic induction. In: Hendrix RW, Roberts JW, Stahl FW et al (eds) Lambda II. Cold Spring Harbor Laboratory, Cold Spring Harbor, pp 123–144 Roberts MD, Drexler H (1981a) Isolation and genetic characterization of T1-transducing mutants with increased transduction frequency. Virology 112(2):662–669 Roberts MD, Drexler H (1981b) T1 mutants with increased transduction frequency are defective in host chromosome degradation. Virology 112(2):670–677 Ross J, Topp E (2015) Abundance of antibiotic resistance genes in bacteriophage following soil fertilization with dairy manure or municipal biosolids, and evidence for potential transduction. Appl Environ Microbiol 81(22):7905–7913. https://doi.org/10.1128/AEM.02363-15 Rutkai E, Gyorgy A, Dorgai L et al (2006) Role of secondary attachment sites in changing the specificity of site-specific recombination. J Bacteriol 188(9):3409–3411. https://doi.org/ 10.1128/JB.188.9.3409-3411.2006 Ruzin A, Lindsay J, Novick RP (2001) Molecular genetics of SaPI1–a mobile pathogenicity island in Staphylococcus aureus. Mol Microbiol 41(2):365–377. doi:mmi2488 Sander M, Schmieger H (2001) Method for Host-Independent Detection of Generalized Transducing Bacteriophages in Natural Habitats. Appl Environ Microbiol 67(4):1490–3 Sandri RM, Berger H (1980) Bacteriophage P1-mediated generalized transduction in Escherichia coli: fate of transduced DNA in rec+ and recA- recipients. Virology 106(1):14–29

190

C. L. Schneider

Sandulache R, Prehm P, Kamp D (1984) Cell wall receptor for bacteriophage Mu G(+). J Bacteriol 160(1):299–303 Sarker SA, McCallin S, Barretto C et al (2012) Oral T4-like phage cocktail application to healthy adult volunteers from Bangladesh. Virology 434(2):222–232. https://doi.org/10.1016/j. virol.2012.09.002 Sato K, Campbell A (1970) Specialized transduction of galactose by lambda phage from a deletion lysogen. Virology 41(3):474–487. doi:0042-6822(70)90169-8 Schmieger H (1970) The molecular structure of the transducing particles of Salmonella phage P22. II. Density gradient analysis of DNA. Mol Gen Genet 109(4):323–337 Schmieger H (1972) Phage P22-mutants with increased or decreased transduction abilities. Mol Gen Genet 119(1):75–88 Schroeder W, Bade EG, Delius H (1974) Participation of Escherichia coli DNA in the replication of temperate bacteriophage Mu1. Virology 60(2):534–542 Segev N, Laub A, Cohen G (1980) A circular form of bacteriophage P1 DNA made in lytically infected cells of Escherichia coli. Characterization and kinetics of formation. Virology 101(1):261–271 Servick K (2016) DRUG DEVELOPMENT. Beleaguered phage therapy trial presses on. Science 352(6293):1506. https://doi.org/10.1126/science.352.6293.1506 Shanado Y, Kato J, Ikeda H (1997) Fis is required for illegitimate recombination during formation of lambda bio transducing phage. J Bacteriol 179(13):4239–4245 Shimada K, Weisberg RA, Gottesman ME (1972) Prophage lambda at unusual chromosomal locations. I. Location of the secondary attachment sites and the properties of the lysogens. J Mol Biol 63(3):483–503. doi:0022-2836(72)90443-3 Shimada K, Weisberg RA, Gottesman ME (1973) Prophage lambda at unusual chromosomal locations. II. Mutations induced by bacteriophage lambda in Escherichia coli K12. J Mol Biol 80(2):297–314. doi:0022-2836(73)90174-5 Shimada K, Weisberg RA, Gottesman ME (1975) Prophage lambda at unusual chromosomal locations. III. The components of the secondary attachment sites. J Mol Biol 93(4):415–429. doi:0022-2836(75)90237-5 Skurnik M, Pajunen M, Kiljunen S (2007) Biotechnological challenges of phage therapy. Biotechnol Lett 29(7):995–1003. https://doi.org/10.1007/s10529-007-9346-1 Smith HW (1972) Ampicillin resistance in Escherichia coli by phage infection. Nat New Biol 238(85):205–206 Sternberg N (1986) The production of generalized transducing phage by bacteriophage lambda. Gene 50(1–3):69–85. doi:0378-1119(86)90311-2 Sternberg N (1990) Bacteriophage P1 cloning system for the isolation, amplification, and recovery of DNA fragments as large as 100 kilobase pairs. Proc Natl Acad Sci U S A 87(1):103–107 Sternberg N, Cohen G (1989) Genetic analysis of the lytic replicon of bacteriophage P1. II. Organization of replicon elements. J Mol Biol 207(1):111–133. doi:0022-2836(89)90444-0 Sternberg N, Coulby J (1987) Recognition and cleavage of the bacteriophage P1 packaging site (pac). II. Functional limits of pac and location of pac cleavage termini. J Mol Biol 194(3):469–479. doi:0022-2836(87)90675-9 Sternberg N, Hoess R (1983) The molecular genetics of bacteriophage P1. Annu Rev Genet 17:123–154. https://doi.org/10.1146/annurev.ge.17.120183.001011 Sternberg N, Weisberg R (1975) Packaging of prophage and host DNA by coliphage lambda. Nature 256(5513):97–103 Sternberg NL, Maurer R (1991) Bacteriophage-mediated generalized transduction in Escherichia coli and Salmonella typhimurium. Methods Enzymol 204:18–43. doi:00766879(91)04004-8 Streisinger G, Emrich J, Stahl MM (1967) Chromosome structure in phage t4, iii. Terminal redundancy and length determination. Proc Natl Acad Sci U S A 57(2):292–295 Stresinger G, Edgar RS, Denhardt GH (1964) Chromosome structure in phage T4. I. Circularity of the linkage map. Proc Natl Acad Sci U S A 51:775–779 Sulakvelidze A, Alavidze Z, Morris JG Jr (2001) Bacteriophage therapy. Antimicrob Agents Chemother 45(3):649–659. https://doi.org/10.1128/AAC.45.3.649-659.2001

Bacteriophage-Mediated Horizontal Gene Transfer: Transduction

191

Susskind MM, Botstein D (1978) Molecular genetics of bacteriophage P22. Microbiol Rev 42(2):385–413 Takahaski S (1975) Physiological transition of a coliphage lambda DNA replication. Biochim Biophys Acta 395(3):306–313 Tavares P, Zinn-Justin S, Orlova EV (2012) Genome gating in tailed bacteriophage capsids. Adv Exp Med Biol 726:585–600. https://doi.org/10.1007/978-1-4614-0980-9_25 Taylor AL (1963) Bacteriophage-Induced Mutation in Escherichia Coli. Proc Natl Acad Sci U S A 50:1043–1051 Thiel K (2004) Old dogma, new tricks–21st Century phage therapy. Nat Biotechnol 22(1):31–36. https://doi.org/10.1038/nbt0104-31 Thierauf A, Perez G, Maloy AS (2009) Generalized transduction. Methods Mol Biol 501:267–286. https://doi.org/10.1007/978-1-60327-164-6_23 Thomas CM, Nielsen KM (2005) Mechanisms of, and barriers to, horizontal gene transfer between bacteria. Nat Rev Microbiol 3(9):711–721. https://doi.org/10.1038/nrmicro1234 Thurber RV, Haynes M, Breitbart M et al (2009) Laboratory procedures to generate viral metagenomes. Nat Protoc 4(4):470–483. https://doi.org/10.1038/nprot.2009.10 Tormo MA, Ferrer MD, Maiques E et al (2008) Staphylococcus aureus pathogenicity island DNA is packaged in particles composed of phage proteins. J Bacteriol 190(7):2434–2440. https://doi. org/10.1128/JB.01349-07 Tye BK, Huberman JA, Botstein D (1974) Non-random circular permutation of phage P22 DNA. J Mol Biol 85(4):501–528. doi:0022-2836(74)90312-X Ubeda C, Maiques E, Barry P et al (2008) SaPI mutations affecting replication and transfer and enabling autonomous replication in the absence of helper phage. Mol Microbiol 67(3):493–503. https://doi.org/10.1111/j.1365-2958.2007.06027.x Ubeda C, Maiques E, Knecht E et al (2005) Antibiotic-induced SOS response promotes horizontal dissemination of pathogenicity island-encoded virulence factors in staphylococci. Mol Microbiol 56(3):836–844. https://doi.org/10.1111/j.1365-2958.2005.04584.x Ubeda C, Maiques E, Tormo MA et al (2007) SaPI operon I is required for SaPI packaging and is controlled by LexA. Mol Microbiol 65(1):41–50. https://doi.org/10.1111/j.1365-2958.2007.05758.x van Schaik W (2015) The human gut resistome. Philos Trans R Soc Lond Ser B Biol Sci 370(1670):20140087. https://doi.org/10.1098/rstb.2014.0087 Varga M, Kuntova L, Pantucek R et al (2012) Efficient transfer of antibiotic resistance plasmids by transduction within methicillin-resistant Staphylococcus aureus USA300 clone. FEMS Microbiol Lett 332(2):146–152. https://doi.org/10.1111/j.1574-6968.2012.02589.x Viertel TM, Ritter K, Horz HP (2014) Viruses versus bacteria-novel approaches to phage therapy as a tool against multidrug-resistant pathogens. J Antimicrob Chemother 69(9):2326–2336. https:// doi.org/10.1093/jac/dku173 von Wintersdorff CJ, Penders J, van Niekerk JM et al (2016) Dissemination of antimicrobial resistance in microbial ecosystems through horizontal gene transfer. Front Microbiol 7:173. https://doi.org/10.3389/fmicb.2016.00173 Waddell TE, Franklin K, Mazzocco A et al (2009) Generalized transduction by lytic bacteriophages. Methods Mol Biol 501:293–303. https://doi.org/10.1007/978-1-60327-164-6_25 Wang BM, Liu L, Groisman EA et al (1987) High frequency generalized transduction by miniMu plasmid phage. Genetics 116(2):201–206 Weber-Dabrowska B, Jonczyk-Matysiak E, Zaczek M et al (2016) Bacteriophage procurement for therapeutic purposes. Front Microbiol 7:1177. https://doi.org/10.3389/fmicb.2016.01177 Weigle J, Meselson M, Paigen K (1959) Density alterations associated with transducing ability in bacteriophage lambda. J Mol Biol 1:379 Weigle JJ, Bertani G (1953) Variations of bacteriophage conditioned by host bacteria. Ann Inst Pasteur (Paris) 84(1):175–179 Weinbauer MG, Rassoulzadegan F (2004) Are viruses driving microbial diversification and diversity? Environ Microbiol 6(1):1–11. doi:539 [pii] Weisberg RA (1987) Specialized transduction. In: Niedhardt FC (ed) Escherichia coli and Salmonella: cellular and molecular biology, 1st edn. American Society for Microbiology, Washington DC, pp 1169–1176

192

C. L. Schneider

Weisberg RA (1996) Specialized transduction. In: Neidhardt FC (ed) Escherichia coli and Salmonella: cellular and molecular biology, 2nd edn, vol 2. ASM Press, Washington DC, pp 2442–2448 Weisberg RA, Gottesman ME (1969) The integration and excision defect of bacteriophage lambdadg. J Mol Biol 46(3):565–580. doi:0022-2836(69)90196-X Wilson GG, Young KY, Edlin GJ et al (1979) High-frequency generalised transduction by bacteriophage T4. Nature 280(5717):80–82 Wittebole X, De Roock S, Opal SM (2014) A historical overview of bacteriophage therapy as an alternative to antibiotics for the treatment of bacterial pathogens. Virulence 5(1):226–235. https://doi.org/10.4161/viru.25991 Wolf RE Jr (1980) Integration of specialized transducing bacteriophage lambda cI857 St68 h80 dgnd his by an unusual pathway promotes formation of deletions and generates a new translocatable element. J Bacteriol 142(2):588–602 Wollman EL (1953) Sur le déterminisme génétique de la lysogénie. Ann. Inst. Pasteur (Paris) 84:281–293 Wollman EL, Jacob F, Hayes W (1956) Conjugation and genetic recombination in Escherichia coli K-12. Cold Spring Harb Symp Quant Biol 21:141–162 Wu H, Sampson L, Parr R et al (2002) The DNA site utilized by bacteriophage P22 for initiation of DNA packaging. Mol Microbiol 45(6):1631–1646. doi:3114 [pii] Yarmolinsky MB, Sternberg N (1988) Bacteriophage P1. In: Calendar R (ed) The bacteriophages, vol 1. Plenum Publishing Corporation, New York, pp 291–438 Yen HC, Hu NT, Marrs BL (1979) Characterization of the gene transfer agent made by an overproducer mutant of Rhodopseudomonas capsulata. J Mol Biol 131(2):157–168. doi:00222836(79)90071-8 Young KK, Edlin GJ, Wilson GG (1982) Genetic analysis of bacteriophage T4 transducing bacteriophages. J Virol 41(1):345–347 Young R, Gill JJ (2015) MICROBIOLOGY. Phage therapy redux–what is to be done? Science 350(6265):1163–1164. https://doi.org/10.1126/science.aad6791 Zinder ND, Lederberg J (1952) Genetic exchange in Salmonella. J Bacteriol 64(5):679–699

Genetics and Genomics of Bacteriophages The Evolution of Bacteriophage Genomes and Genomic Research Aidan Casey, Aidan Coffey, and Olivia McAuliffe

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Nature of Bacteriophage Genomes: Evolution and Mosaicism . . . . . . . . . . . . . . . . . . . . . . . . . . . Main Structural Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Structural Components of Tailed Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Structural Components of PFP Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bacteriophage T4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hallmark Genes and Variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hallmark Genes and Bacteriophage Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comparative Genomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bacterial Comparative Genomics: The Impact of Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . . . Comparative Genomics of Bacteriophages: Phylogeny and Evolution . . . . . . . . . . . . . . . . . . . . Comparative Genomics and Bacteriophage Taxonomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comparative Metagenomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Genomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Genomics and Genetic Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Site-Directed Mutagenesis (SDM) and Gene Knockouts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transcriptomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Host Response to Bacteriophage Infection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion and Future Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

194 197 200 200 201 201 202 203 204 204 204 205 207 208 209 209 211 211 212 213

A. Casey · O. McAuliffe (*) Department of Food Biosciences, Teagasc Food Research Centre, Cork, Ireland e-mail: [email protected]; [email protected] A. Coffey Department of Biological Sciences, Cork Institute of Technology, Cork, Ireland e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_5

193

194

A. Casey et al.

Abstract

In recent years, advancements in sequencing technology and genome analysis software have broadened the horizons for bacteriophage research. As the repository of data generated continues to grow, the fundamental principles of bacteriophages in terms of their population numbers, diversity, and composition have become increasingly apparent. Comparative genomic analyses have facilitated the definition of key concepts such as “mosaicism,” providing researchers with an insight into the highly complex nature of bacteriophage evolution, while the identification of “viral hallmark genes” has established an underpinning connection between bacteriophages throughout the virosphere. Furthermore, large-scale metagenomic research has confirmed bacteriophages as potential candidates for exploitation in a number of biological applications, including rapid detection of pathogens, as well as in the development of antimicrobials and repair enzymes. In order to exploit bacteriophages to their fullest capabilities, developments and improvements to the current collection of analysis software must be made in order to handle the expansive wealth of data that continues to be generated.

Introduction Bacteriophages are widely considered to be the most abundant microorganisms on the planet. Current estimates indicate that there are approximately 1031 virus particles in the biosphere (Kristensen et al. 2013), with 1023 viral infections occurring every second in the world’s oceans (Suttle 2007). Given their abundance in virtually every environment, bacteriophages naturally exhibit extensive diversity in their relative morphological and genomic composition. They can be genetically composed of either DNA, dsDNA as in the case of the Tectiviridae and ssDNA in the case of Microviridae, or RNA, as seen for the Leviviridae, and Cystoviridae, which have, ssRNA and dsRNA genomes, respectively. Morphologically, they can exist as polyhedral, filamentous, tailed, or indeed pleomorphic particles (Ackermann 2011). Phages also vary greatly in their relative genome sizes, ranging from the smallest Leuconostoc phage L5 genome at just 2,435 bp in length (Hatfull 2008) to the largest known Bacillus megaterium phage G, which has a genome size of approximately 498,000 bp (Hendrix 2009). Approximately 96% of all known bacteriophages are of the order Caudovirales and are tailed viruses containing dsDNA genomes (Ackermann 2007). Bacteriophages may be further subdivided as either lytic or lysogenic, depending on their relative modes of replication. Phages which utilize the lytic pathway are deemed “virulent” in their nature, and by definition, infection by a lytic phage will result in lysis of the host, followed by release of phage progeny. Bacteriophages of this type are common to environments where the host is present in a relatively high abundance (Maurice et al. 2013). Infection with a lysogenic bacteriophage on the other hand results in the integration of the phage into the host genome, where it exists as a prophage, and phages of this type are common to

Genetics and Genomics of Bacteriophages

1400

195

Fully sequenced bacteriophage genomes available on the NCBI database (1996 - 2013)

1200 1000 Number of sequenced bacteriophage genomes deposited to the NCBI database

800 600 400 200 2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

0

Fig. 1 Increase in number of sequenced phage genomes deposited to the NCBI database between 1996 and 2013

environments whereby the conditions are somewhat unfavorable for survival; the virus will remain integrated within the host until such a time that it becomes induced as a result of external factors or stresses. Bacteriophage research dates back almost 100 years to their first discovery in the early twentieth century (Twort 1915; d’Herelle 1917), yet interest in phage genomic research has only progressed significantly since the foundation of whole genome sequencing. The first completed genome sequencing project was that of bacteriophage øX174 (Sanger et al. 1977), which was subsequently followed by a number of others, including G4 (Godson et al. 1978), lambda (Sanger et al. 1982), and T7 (Dunn et al. 1983) to name but a few. Prior to the scientific milestone of whole genome sequencing, comparative and functional research of phages relied on phenotypic characterization, and at the forefront of this research were the T-even phages, a group of lytic viruses first isolated in the mid 1940s (Abedon 2000). A member of this group, bacteriophage T4, is a tailed dsDNA member of the Myoviridae family that infects E. coli and has a 168kbp genome containing approximately 300 genes. T4 has represented a model for bacteriophage genetics over the past 80 years, upon which important fundamental phage structural and phenotypic characteristics have been established, despite the fact that the complete genome sequence of T4 has only been available for the past decade or so (Miller et al. 2003). Since the first phage genomes were assembled in the late 1970s, the process of whole genome sequencing has become considerably more sophisticated, resulting in a reduction in running costs, as well as an improvement in sequencing quality and efficiency, all of which make it a viable option today even for small research groups with relatively limited budgets. This is reflected by the considerable increase in numbers of fully sequenced bacteriophage genomes that have been deposited to the NCBI database in recent years, with the total figure rising from 600 by the end of 2010 to 1,303 by the time of writing (June 2014) (Figs. 1 and 2).

196

A. Casey et al. Nucleic Acid Composion of Bacteriophage Genomes

dsDNA - 94.1% ssDNA - 4.6% dsRNA - 0.4% ssRNA - 0.9%

Fig. 2 Nucleic acid composition of bacteriophage genomes available on the NCBI database

From a genomic perspective, bacteriophages are comprised of a set of interchangeable “modules” (composed of single or several genes) and as such are deemed to be “mosaic” in nature; every phage genome is composed of a combination of these modules. The modular theory of bacteriophage evolution, published over three decades ago (Botstein 1980), proposes that phages do not evolve on a whole genome scale but more so at the level of their modules. While all bacteriophages are to some extent mosaic, it is clear that not all the modules within a given genome will participate in mosaicism to the same degree (Hatfull and Hendrix 2011). Certain groups of genes (including structural, lysis, and DNA replication genes) which are essential to the comprehensive functioning of the virus are termed “core genes,” considering their high degree of sequence conservation, in addition to their prominence in large sets of phage genomes. In comparison to other more flexible areas of the genome, these core genes exhibit a minimal degree of mosaicism as a bacteriophage evolves. “Non-core genes” on the other hand, encompass those which are not common amongst related phage groups and are often found in what are known as hyperplastic regions (HPR) of many bacteriophage genomes (Comeau et al. 2007). Despite the wealth of functional and homology-based knowledge now available, there remain a sizeable number of sequenced and annotated bacteriophage genes for which no function has been attributed. Anywhere between 60–99% of sequences obtained from viral metagenomic studies share no amino acid similarity with previously observed genes, therefore classifying them as “unknown” (Mokili et al. 2012). Indeed, despite the fact that bacteriophage T4 is one of the most extensively studied phage genomes, only approximately half of its genes have been assigned a function (Miller et al. 2003). This chapter will deliver a brief account of the nature of bacteriophage genomes, as well as examining some of the main structural features and hallmark genes that are most commonly associated with phage particles. In addition, it will offer a review of some recent advances in the areas of comparative genomics, metagenomics, and host transcriptomics, as well as assessing the current state of phage functional genomics. To conclude, it will present an outlook on the future of bacteriophage research from a genetics and genomics perspective.

Genetics and Genomics of Bacteriophages

197

The Nature of Bacteriophage Genomes: Evolution and Mosaicism The extent of their phenotypic diversity in terms of morphology, host specificity, and mode of action, together with the apparent lack of a common genetic region such as the 16S rRNA of their bacterial counterparts, has made it difficult in the past for researchers to define relationships that exist between bacteriophage species. The perceived “mosaic” nature of phages, as outlined in Botstein’s modular theory (1980), dictates that they evolve laterally through horizontal gene transfer, rather than from any common ancestor as observed in their bacterial counterparts. Generally speaking, bacteriophage mosaicism arises from one of two types of recombination: homologous or illegitimate. Homologous recombination (Fig. 3a), by definition, involves the transfer of genes or divergent sequences between phages via the recognition of flanking homologous sequences (Swenson et al. 2013). Illegitimate recombination on the other hand encompasses genetic exchange between bacteriophages which appears to happen randomly, resulting in a very low relative frequency of viable phage progeny (Fig. 3b). Despite this, selectively advantageous products of illegitimate recombination represent an extremely creative process in which novel combinations of gene and protein domains are generated (Hatfull 2008). Considering the observation that phages in general lack a conserved sequence that is long enough to be recognized by recombination machinery (Pedulla et al. 2003), illegitimate recombination is widely thought to be the main driving force

Non-viable phage recombinants 3a

3b ATTCGAGC

Homologous flanking sequences

TTTCGAGG

X Y

Random genetic exchange between phages

Viable phage recombinant

Z Recombinant with new gene insertion present

Bacteriophage Genome Interchangeable Module (Core Gene) Interchangeable Modules (Non-Core Genes)

Fig. 3 Visual representation of bacteriophage recombination. (a) On the left hand side, depicts homologous recombination between phages via recognition of specific flanking sequences. (b) On the right hand side, depicts illegitimate recombination, which occurs randomly resulting in a very low relative frequency of viable phage progeny

198

A. Casey et al.

behind bacteriophage mosaicism. However, more recent research (De Paepe et al. 2014) suggests that Rad52-like recombinases may play a pivotal role in bacteriophage evolution. Indeed, the type of recombination that contributes to phage diversity appears to be heavily dependent on the differing selective pressures that are placed on viruses. While it is thought to occur indiscriminately across the genome, illegitimate recombination of one or more core genes could disrupt the essential functioning of the phage and would almost exclusively result in a nonviable phage recombinant. The low level of mosaicism observed amongst the core genes therefore comes as a result of vast elimination of recombinants which confer a selective disadvantage to the phage (Hendrix 2002). The relatively high degree of conservation between genomes makes the core genes relatively identifiable. A number of software tools are available for use in the identification of core genes, including CoreGenes 3.5 (Turner et al. 2013), which uses BLAST-based algorithms to establish proteins which are common to a set of input genomes. Non-core genes are much more tolerant of illegitimate recombination events, given the fact that they are not essential to the survival of the phage; hence, these regions are commonly found to be widely variable, even between phenotypically similar phages. Although the majority of noncore genes have an unknown function and are not believed to be essential to the survival of the phage, it is thought that they may represent a repository for the generation of novel genes through mutation and recombination, with Hatfull et al. (2011) describing these highly mosaic regions as a possible “gene nursery” utilized by the phage to aid in adapting to a particular environmental niche. In Botstein’s model (Botstein 1980), physical access to other bacteriophage is the limiting factor for recombination. Homologous phages arise as a result of their relative access to a common genetic pool of modules, but access to this genetic pool is not uniform for all bacteriophages (Hendrix et al. 1999; Summer et al. 2006). Recombination between physically distant bacteriophages occurs at a much lower frequency than between those in a particular environmental niche (Fig. 4). The current ICTV system (International Committee for the Taxonomy of Viruses) classifies phages on the basis of phenotypic traits such as morphology and host range, yet the modular theory proposes that evolutionary relationships are far more complex, considering two phages which share similar phenotypic attributes may differ entirely in their genomic makeup as a result of the lack of a common pool. Lysogenic bacteriophages are particularly mosaic in nature compared to their lytic counterparts, due to the higher frequency at which they engage and recombine with other viruses, for example, as a lysogen in a bacterial cell that is subject to an external infection by a lytic phage (Martinsohn et al. 2008). The true extent of the mosaic nature of bacteriophages has become more apparent with the recent advancements made in whole genome sequencing. Bacteriophages have been shown to utilize horizontal gene transfer not only for their own evolution but also in the evolution of their bacterial counterparts. Of particular relevance to this are gene transfer agents (GTA), which are a distinct class of defective transducing bacteriophages or “phage-like entities” that are present in a range of bacterial and archaeal genomes (Fig. 5). Genes encoding the phage-like

Genetics and Genomics of Bacteriophages

199

a

Bacteriophage genome

b

Interchangeable module (core gene) Interchangeable module (non-core gene)

Fig. 4 Homologous bacteriophage genomes arise as a result of their relative physical access to a pool of interchangeable modules. Two distinct environmental niche areas are denoted (a, b)

Host DNA packaged into capsid

GTA Genes

Construction of GTA particle Extracellular polysaccharide

Release upon host cell lysis Injection of bacterial DNA

Transformation of DNA into new host genome

Fig. 5 Gene transfer agents involved in horizontal gene transfer through transduction of bacterial DNA

200

A. Casey et al.

structure of GTAs are found within the genome of the host bacterium, and when the particle is produced, a random piece of the host DNA is packaged into the capsid, rather than any GTA-specific genes themselves, thus promoting genetic exchange between bacterial species (Lang et al. 2012). GTA entities are all thought to have a tailed structure (similar to that of the Caudovirales) and are presumably released from the bacterial cell upon lysis of the producing host. Their contribution to genetic exchange within prokaryotes holds widespread consequences (Swanson et al. 2012), including the emergence of novel bacterial pathogens through conferred resistance (Muniesa et al. 2013).

Main Structural Features Given the sheer magnitude of population size and diversity observed among bacteriophages in the biosphere, a proportionate level of variation between these viruses in terms of their relative structure could be expected. However in general, bacteriophages conform to quite a modest combination of arrangements upon which they are composed. The vast majority of isolated phages that have been examined under a microscope are tailed and belong to the order Caudovirales. The result of this is that genetically distantly related bacteriophages become assigned to the same family under the ICTV’s classification system, a system which has come under much scrutiny in recent times (Gibbs 2013).

Structural Components of Tailed Bacteriophages Tailed bacteriophages of the order Caudovirales comprise approximately 96% of all known bacterial viruses. The order itself is sub-categorized into one of three families, namely, the Siphoviridae (61%), the Myoviridae (24.5%), and the Podoviridae (14%). Each shares a similar capsid conformation consisting of a polyhedral head (composed of many copies of one or more proteins) enclosing the viral genome, but the families differ in their relative tail configurations. Siphoviridae have characteristically long noncontractile tails composed of a tip complex which is involved in host recognition, and the tail tube which acts as a pipeline for DNA transfer (Fokine and Rossmann 2014). Myoviridae have predominantly contractile tails with a similar structure to Siphoviridae, but which also possess an outer sheath encompassing the inner central tube, while Podoviridae have characteristically short tails (Veesler and Cambillau 2011). The physical length of the tail is dictated by the tape measure protein, which spans the tail tube (Maxwell and Davidson 2013). This protein represents a crucial structural feature where tailed bacteriophages are concerned. Not only is the protein responsible for tail length determination, it has also been implicated to have a role in phage DNA injection into the host cell upon infection (Xu et al. 2013). Connecting the capsid to the tail in these types of phages is a multi-protein complex, consisting of the head completion proteins gp15 and gp16, as well as the portal protein gp6 (Orlova et al. 2003), which plays a number of roles

Genetics and Genomics of Bacteriophages

201

in the bacteriophage, including DNA packaging into the capsid and DNA release into the bacterium upon successful establishment of infection. The baseplate is located at the end of the tail, upon which long tail fibers are attached to the periphery that serve in primary host recognition.

Structural Components of PFP Bacteriophages The remaining 4% of bacteriophages consist of polyhedral, filamentous, and pleomorphic (PFP) viruses. Filamentous bacteriophages, such as the Inoviridae, are structurally composed of several thousand copies of identical coat proteins arranged in a helical sheath around a central core containing ssDNA and are divided into two separate classes (I & II) on the basis of their diffraction patterns (Marvin et al. 2014). They consist of a very small number of genes and are widely considered to be one of the simplest known biological systems. Interestingly, where filamentous phages are concerned, the virus enters, replicates, and exits without causing lysis of the host cell. Other members of the PFP phage group include an array of morphologically unique families. For example, pleomorphic viruses characteristically have dsDNA genomes surrounded by an array of different structures including a lipoprotein envelope (Plasmaviridae), a lemon-shaped capsid (Fusseloviridae), a droplet-like capsid (Guttaviridae), or indeed a bottle-shaped capsid morphology (Ampullaviridae) (Krupovic et al. 2011). Polyhedral phages all have capsids with an icosahedral shape but differ generally on the basis of their genome types, including the aforementioned dsDNA genomes of the Tectiviridae and Corticoviridae, the ssDNA genome of the Microviridae, the ssRNA genome of the Leviviridae, and the dsRNA genome of the Cystoviridae. In addition to these common features, there have been a number of recorded occurrences of novel features that depart from the known conventional phage structures. One such study (Kuznetsov et al. 2013) identified unique appendages on marine cyanophage, consisting of up to four complex fibers located on the neck or baseplate which have a number of bulbs at their distal ends, thought to be involved in host recognition in the marine environment. Likewise, a study on a Lactococcus lactis bacteriophage by Cavanagh et al. (2013) identified by electron microscopy an elongated tail fiber that has not been previously observed in phages of this type, and this feature is thought to confer an enhanced ability for the phage to achieve infection.

Bacteriophage T4 In order to further examine the specific structural components of bacteriophages, it is necessary to choose an example that has been comprehensively described in the literature. The extensive research that has been carried out on phage T4 complements the fact that it has one of the most complex known structures. The capsid alone consists of over 3000 polypeptide chains of approximately 12 different protein types

202

A. Casey et al.

which comprise an icosahedral structure (Leiman et al. 2003). As a member of the Myoviridae family, T4 has a contractile tail, which is found to be equally as complex as the capsid. Approximately 14% of the genome (25kbp of a total of 168kbp) is dedicated to assembly of the phage tail, which consists of over 20 genes involved specifically in this process. As described previously for the Myoviridae, the tail is made up of an inner tube surrounded by an outer protective sheath, which contracts upon host infection by the phage. On the end of the tail sits the baseplate, a multiprotein complex (gp11,10,7,8,6,53,25) composed of several “wedges” surrounding a central tail spike hub (Kostyuchenko et al. 2003), which has both long and short tail fibers attached to it. Binding of a long tail fiber to an E. coli cell surface receptor results in signal transduction to the baseplate and the extension of the short fibers that subsequently bind irreversibly to the cell. This is followed by a conformational change of the physical plate structure, anchoring it to the membrane (Fokine and Rossmann 2014). Along the axis of the baseplate dome is a spike composed of the gp5 and gp27 proteins. Upon conformational change, the central spike punctures the host cell membrane, and the peptidoglycan layer is then digested via a gp5 lysozyme domain (Browning et al. 2012). It is thought that once penetration is complete, the gp27 protein of the tail spike interacts with a specific receptor on the cytoplasmic membrane to initiate DNA release into the host, allowing infection to begin (Leiman et al. 2003). On a genomic level, bacteriophage diversity is enormous throughout the biosphere, and yet phage structural diversity is comparably minute. Structural determinants and mechanisms of virion particle assembly represent some of the previously described “core genes” in bacteriophages, with any mutation in these particular modules resulting in abolishment of viability. Their genomic content and function make them biologically unique to viruses, and as such, they belong to an exclusive set known as the “viral hallmark genes.”

Hallmark Genes and Variability Bacteriophage genomes are modular by nature. As such, they are composed of highly conserved regions of DNA complemented with regions of equally high variability. While it is well known that there is no single gene that is common to all viruses, there are a number of genes which are found to be shared by a wide array of bacteriophage genomes, spanning the ensemble of morphological phage types. These are known as “hallmark genes” and are responsible for some of the more essential roles within bacteriophages, not only where structural formation is concerned but also with regard to viral DNA synthesis, replication, and host integration (Koonin and Dolja 2013). Two of the most striking examples of “hallmark genes” within bacteriophages come in the form of the genes encoding the jelly-roll capsid protein (JRC) and the superfamily 3 helicase protein (S3H) (Koonin et al. 2006). These genes are widespread amongst bacteriophages and are even found to cross the genetic boundaries between DNA and RNA viruses, forming completely unexpected links between phages that are distantly or otherwise unrelated. The tape

Genetics and Genomics of Bacteriophages

203

measure protein (TMP) is a classic example of a highly variable product of a hallmark gene and is one of the characteristic proteins across virtually all tailed bacteriophages. Yet as mentioned previously, it is variable in that the length of the protein product proportionally dictates the resulting length of the phage tail (Xu et al. 2013). Viral hallmark genes are by definition unique in their genomic composition in that they are either completely exclusive to viruses or possess only very distant homologues in bacterial genomes. They form a connection between viral genomes that are seemingly unrelated in any other sense. The origin of these particular genes is a topic of much debate considering the lack of closely related cellular homologues.

Hallmark Genes and Bacteriophage Evolution A review by Koonin et al. (2006) outlined three distinct hypotheses for the evolution of the viral hallmark genes. The first, and arguably least likely scenario, suggests that bacterial orthologs of these genes do in fact exist but are undetectable due to extensive sequence divergence. Considering distant relatives of hallmark genes have been identified already in other cellular organisms, it makes the possibility of the existence of undiscovered orthologs unlikely. The second hypothesis proposes that the origin of viruses predates the last universal cellular ancestor (LUCA), a theory for which there is growing evidence (Holmes 2011), suggesting that viruses have a monophyletic ancestry, with subsequent evolutionary loss of genes in certain groups resulting in the divergences that are observed between phages today. A third hypothesis contradicts the second, in that it postulates that viruses in fact evolved from a polyphyletic ancestry, and that commonalities in hallmark gene content between distantly related phages arose as a result of horizontal gene transfer rather than via any vertical means. This theory represents an inverted view on the monophyletic opinion. Hallmark genes, while unique, are not necessarily found within all virus types. An elementary example of this is observed in bacteriophages with differing lifestyles in terms of their method of host infection, i.e., lytic v lysogenic phages. Lysogenic infection, as described previously, is characterized by insertion of phage DNA into the host’s genome and is a process which requires viral integrase proteins. Certain genes that encode these integrases are considered to be “hallmark” to bacteriophages and yet are not found present in viruses which are strictly lytic in their lifestyle. Indeed, lytic phages have their own modules, absent in some lysogenic phages, which contain hallmark gene sets functioning in host eradication upon infection. A software program known as Phage Classification Tool Set (PHACTS) was recently developed in order to predict whether or not a bacteriophage had a lytic or lysogenic lifestyle based on modular similarities with a training set of other phages with known lifestyles (McNair et al. 2012). With a 99% precision rate, this tool represents not only a valuable asset in phenotype prediction for the vast array of unculturable phages but also a prime example of the power of harnessing viral hallmark genes to extrapolate information from a phage genome.

204

A. Casey et al.

In order to obtain a more complete impression of these hallmark genes, there is an inherent need for vast expansion of viral biodiversity research given the diminutive sample upon which inferences are currently built, and this must become a priority. If the true evolutionary origins of bacteriophages and indeed viruses as a whole are to be deduced, the hallmark genes certainly will prove to be key informants.

Comparative Genomics The advancement of whole genome and metagenome sequencing in recent times has expanded the wealth of biological data that is now available for comparative analyses of bacteriophages. However, the reality is that in terms of understanding the true diversity of phages, researchers are merely scratching the surface, as it has been conservatively estimated that there may be at least ten million species of tailed bacteriophages alone (Casjens 2005), a figure that does not encompass enveloped phages, filamentous phages, or indeed prophages, of which there are predicted to be an average of two to three regions per bacterial genome, based on research undertaken on sequenced bacteria (Fouts 2006; Akhter et al. 2012).

Bacterial Comparative Genomics: The Impact of Bacteriophages Genome sequencing and subsequent comparative analysis can identify at the precise genomic level the way in which bacteriophages affect the composition, functioning, and population dynamic of their bacterial counterparts. For example, recent research on Shigella flexneri identified a particular lysogenic phage, sfIV, to play a major role in the acquired pathogenicity of the bacteria through serotype conversion (Jakhetia et al. 2013). Comparative analysis between this particular lysogenic phage and other phages of S. flexneri identified five novel genes, exclusive to sfIV, which are believed to be involved in this process. In a similar fashion, the role for prophages in conferred bacterial pathogenicity has also been reported using comparative genomics in recent studies (Busby et al. 2013; Matos et al. 2013; Vannucci et al. 2013). Furthermore, comparative analyses have revealed how bacteriophages confer antibiotic resistance to bacteria through the process of horizontal gene transfer, as reported in studies on the sputum microbiota of cystic fibrosis patients (Fancello et al. 2011) as well as in other research on murine and human fecal phage populations (Modi et al. 2013; Quirós et al. 2014).

Comparative Genomics of Bacteriophages: Phylogeny and Evolution Comparative genomic research is pivotal in defining relationships between bacteriophages, particularly where large families of viruses are concerned. One such example of the use of large-scale comparative genomics to study a range of related bacteriophages comes from the 2010 study by Petrov et al. (2010), involving

Genetics and Genomics of Bacteriophages

205

genomes of the T4-related phages. As the name suggests, these phages are genetically similar to that of bacteriophage T4 but differ from T4 in a number of crucial aspects, including their host range, proteome composition, and genome sizes. This particular study assessed approximately 40 of these T4-related phages for which the complete genome sequences were obtained and using the available data, were able to identify a shared “core genome” of 30–33 genes which represent a unifying genetic backbone upon which this family is connected. The results of the comparative study propose that the diversity observed between members of this large family arises due to the adaptation of this core genome to both evolution and to the dynamic environmental conditions. A second example of a large-scale comparative genomic study is exemplified in research by Jacobs-Sera et al. (2012), in which 220 different mycobacteriophage genomes were analyzed, with a goal to not only generate a defined structure of relationships between the phages but to also advance the current knowledge on their role in nature. The fully sequenced genomes of these phages were analyzed on the basis of gene content and were as a result grouped into 15 clusters and 8 singletons, with each cluster being further subdivided into a number of subclusters. The results of the study indicated a direct correlation between the type of mycobacteriophage genome and its host range, with subsequent comparative analysis inferring that these host preferences act as a barrier to horizontal gene transfer between each of the phage subclusters. While the advent of comparative genomics has enhanced the understanding of numerous aspects of phages, it has in a similar sense revealed several complications for long-standing fundamentals of phage biology. This is particularly relevant to the current method employed in the classification of bacteriophages.

Comparative Genomics and Bacteriophage Taxonomy The International Committee on the Taxonomy of Viruses (ICTV) was established in 1966, tasked with developing a universal system for classification of all viruses. Since then, the committee has met on a regular basis in order to update taxonomy guidelines, resulting in the issuing of sequential reports, outlining new additions or changes to the system. The ninth such report was issued in 2011, in which approximately 2,300 virus types were now recognized by the ICTV (King et al. 2011), and as of July 2012, the current taxonomy release comprises seven orders which are subdivided into a further 25 families (Table 1). The ICTV system of bacteriophage classification characterizes phages on the basis of a number of physical traits, including shape, structure, genome size, and type, as well as phenotypic traits such as host range and solvent resistance (Ackermann 2011; King et al. 2011). In order to determine many of these characteristics, visualization of bacteriophage particles using a scanning electron microscope therefore is a prerequisite. However, given the recent surge in whole genome sequencing, in addition to advances in viral metagenomics, less emphasis has been put on imaging of newly isolated bacteriophages, with researchers instead opting to use alternative comparative genomic approaches in order to infer genomic evolution and relatedness to other phages.

206

A. Casey et al.

Table 1 Classification of virus families according to the latest ICTV (9th) report Order Caudovirales

Herpesvirales

Ligamenvirales Mononegavirales

Nidovirales

Picornavirales

Tymovirales

Family Myoviridae Podoviridae Siphoviridae Alloherpesviridae Herpesviridae Malacoherpesviridae Lipothrixviridae Rudiviridae Bornaviridae Filoviridae Paramyxoviridae Rhabdoviridae Arteriviridae Coronaviridae Mesoniviridae Roniviridae Dicistroviridae Iflaviridae Marnaviridae Picornaviridae Secoviridae Alphaflexiviridae Betaflexiviridae Gammaflexiviridae Tymoviridae

Detailed visualization and morphological determination of a large percentage of bacteriophages from many environmental samples would be practically impossible, given that these viruses need to be isolated against a known host before they can be imaged, and given that a number of virus species for which a host bacterium has yet to be identified have already been sequenced and characterized from metagenomic studies (Labonté and Suttle 2013). In addition, phage particle visualization cannot be applied in the classification of prophages. Therefore in recent times, there has been somewhat of a shift from the original morphology-based classification of bacteriophages to a more genomics-based approach. While numerous genomic studies continue to find some merit in the validity of morphological taxonomy (Comeau et al. 2012), the increase in genome sequencing and metagenomic data proceeds to proportionally affect the percentage of “unclassified bacteriophages” being deposited to online databases, and there is an urgent need for reform to the universal standard in which bacteriophages and indeed all viruses alike are classified. However, as mentioned previously, bacteriophages as a whole lack a common genetic region upon which they can be classified, and so the

Genetics and Genomics of Bacteriophages

207

goal of an alternative genome-based taxonomy for phages has proven to be quite difficult to realize. With this in mind, a number of groups have presented their respective approaches to addressing the phage taxonomy issue. One such approach (Rohwer and Edwards 2002), proposed a taxonomical system that is based on the predicted phage proteome; the idea being that highly related phages would express a highly similar complement of proteins. With some notable exceptions, the majority of the resulting analysis predicted phage groupings that were reciprocal to those outlined by the ICTV, indicating that “The Phage Proteomic Tree” represents a relatively successful method of classifying phages on a genomic basis that is compatible with the current universal system. Other early comparative genomics studies proposed phage taxonomy based on a single common structural gene module such as that of the DNA-packaging head gene cluster (Proux et al. 2002) or the portal protein (Sullivan et al. 2008), while more recent research offers a method for classification of bacteriophages on the basis of their relative collective similarities in gene content, gene orders, and gene positions (Li et al. 2008). Ultimately, there currently appears to be no single satisfactory method by which bacteriophages and viruses alike are classified. The most likely solution to this problem is a combinative approach which takes into account not only the structural and physical nature of the bacteriophage itself but also the similarities between phages in terms of their relative genomic content. In reality, bacteriophage taxonomy is a much more dynamic concept than the inexorable system upon which it is based; a concept which must adapt as the repository of information grows, especially considering the ever expanding abundance of data that continues to be generated in the area of viral metagenomic research.

Comparative Metagenomics Viral metagenomics (Breitbart et al. 2002) has expanded the boundaries of possibility for comparative genomics of bacteriophages, due to the wealth of genetic information that can now be acquired from these types of studies. The vast array of data generated is evidenced by the fact that the average human or environmental viral metagenome has been shown to contain hundreds to thousands of sequences from unique viral types (Breitbart et al. 2002; Ray et al. 2012; Labonté and Suttle 2013; Mizuno et al. 2013). Comparative metagenomic studies have revealed the extensive nature of global viral diversity and their effect on bacterial dynamics. With quantitative experiments estimating an average of 107 VLP/ml of surface seawater (Breitbart 2012), viral metagenomics has provided researchers with a real insight into their abundance and role in virtually all of the world’s environmental ecosystems. This is particularly true for marine environments, which have recently come to be the focus of numerous large-scale metagenomic studies (Angly et al. 2006; Suttle 2007; Holmfeldt et al. 2013; Kim et al. 2013; Labonté and Suttle 2013; Mizuno et al. 2013; Xia et al. 2013), as phages are believed to be responsible for killing 20% of the total biomass of microorganisms in the sea every day (Suttle 2007).

208

A. Casey et al.

While the effect they exert on their bacterial counterparts is evident, a number of factors must be taken into account when considering whether or not the findings of a particular metagenomic study give a true representation of diversity amongst the viral population of an environment. For example, isolating bacteriophages from an environmental sample using tangential-flow filtration (Thurber et al. 2009), will confer a particle-size bias toward smaller bacteriophages if the type of filter chosen (e.g., 0.2 μm) excludes some uncharacteristically large viruses (Koonin 2005). Furthermore, the technique to be employed in the construction of a phage metagenome also requires careful consideration, since viral DNA requires amplification prior to sequencing. Common preparation methods such as the linker-amplified shotgun library (LASL) method, and the multiple displacement amplification (MDA) method (Kim and Bae 2011), are associated with known virus-specific biases, having been shown to exclusively amplify dsDNA and ssDNA, respectively. Given that viral metagenomics has become an essential tool for elucidating phage diversity and dynamics, it is imperative that the manner in which research is undertaken must be calculated down to the most intricate detail, considering these extensive biases observed in particular metagenome preparations.

Functional Genomics The advent of gene annotation software, together with improvements in sequence comparison tools has changed the face of the approach to functional genomics when studying bacteriophages. Advances in the field of viral metagenomics have led to the development of new techniques and new applications for phages and their components in various areas of research, including their use as antimicrobials, anticancer agents, components of drug delivery mechanisms, and in development of repair enzymes to name but a few (Schoenfeld et al. 2010). Elucidating bacteriophage gene functions has been beneficial to these developments as well as resolving certain aspects of viral evolution. For example, a deeper understanding of functional genomics has allowed researchers to justify observed mosaicism within bacteriophages, particularly when it comes to the stark contrast between coevolving core genes and highly plastic non-core genes. Hatfull and Hendrix (2011) suggest that the rigidity observed in terms of genomic content for core genes exists not as a result of the genes themselves but more so as a result of the intricate and highly specific interactions that occur between their translated proteins. These interactions may be essential to the viability of the phage (e.g., in assembly of certain structural features such as the capsid); therefore, any recombination involving core genes would result in a nonfunctional progeny, even in cases where the recombination event resulted in the replacement of one gene with that of a homologue from a different phage. On the other hand, mutations or recombination events involving non-core genes are tolerated to a much higher degree from an evolutionary perspective, given the fact that these particular genes are not essential to the viability of the recombinant.

Genetics and Genomics of Bacteriophages

209

Functional Genomics and Genetic Characterization At present, large numbers of bacteriophage genes of unknown function continue to be identified. Generally speaking, there are three inherent reasons as to why this is the case. Firstly, as the number of viral metagenomic studies increases, so too does the enormous archive of data associated with them, making it virtually impossible to keep nucleotide and protein databases completely up to date. Secondly, due to the low level of evolutionary conservation of certain “non-core genes” in bacteriophages, an average of less than 30% of sequences produced from metagenomic studies show any kind of homology to a previously annotated gene, and only a small proportion of these genes have an identifiable function (Schoenfeld et al. 2010). Thirdly, the increased emphasis being put on homology-based prediction that is generally associated with metagenomic studies has shifted focus away from the more traditional techniques for elucidating gene function. Researchers are left with no alternative but to utilize homology-based approaches to assigning gene functions, given the fact that many of the sequenced bacteriophages cannot be individually isolated or cultured (Labonté and Suttle 2013). In an attempt to address this ever expanding repository of phage genes without a known function, researchers must look to alternative methods which focus more on genomic context and content. One such approach involves characterizing unknown genes on the basis of their position on the genome of the phage. Given the modular nature of viruses and the high degree of mosaicism, closely related phages will exhibit a similarity in the ordering and location of their genes (Klumpp et al. 2013), and so gene function can be inferred through comparative analysis. Another more sequence-based approach involves the use of automated protein structure prediction software, such as 3Drefine (Bhattacharya and Cheng 2013), Genome3D (Lewis et al. 2013), or HHpred (Hildebrand et al. 2009), all of which predict the function of a gene on the basis of the structure of its translated product. In addition, the accuracy of this type of software has an important role in the development of site-directed mutagenesis and gene knockout studies.

Site-Directed Mutagenesis (SDM) and Gene Knockouts The somewhat classic method for identifying individual functions within an organism is through gene inactivation and subsequent phenotypic observation. While this approach is highly informative, an SDM study is limited by the fact that it can only be undertaken both when a bacteriophage is amplifiable in the lab environment and when the complete genome sequence of the phage is available. Thus, SDM is not a viable technique for elucidating gene functions from viruses obtained through the likes of metagenomic studies. When such a study is possible, SDM represents a powerful tool in demonstrating specific pathways and interactions between phages and their hosts. Utilizing SDM as well as gene insertions and knockouts could resolve some of the most intimate of relationships between phages and their hosts,

210

A. Casey et al.

and the following are some examples of such research observed in published literature which reaffirm this hypothesis. Clustered regularly interspaced short palindromic repeats (or CRISPRs) together with CRISPR-associated (Cas) proteins form an adaptive immune system in many bacteria, consisting of short repeats, separated by variable spacer sequences (Garneau et al. 2010) These variable spacer sequences, first described over a decade ago (Jansen et al. 2002), correspond to foreign viral DNA which has been captured by the bacteria and incorporated into the host genome, so that subsequent foreign nucleic acids invading the cell can be recognized and cleaved. Recently, functional genomic studies have identified a number of bacteriophage genes with the ability to inactivate this CRISPR/Cas system in Pseudomonas aeruginosa (Bondy-Denomy et al. 2013). These functions were confirmed by inserting the “anti-CRISPR” genes into the genomes of susceptible phages and showing that they were now able to evade host defenses and infect. Likewise, another study (Seed et al. 2013) demonstrated through silent mutation knockouts how a phage-encoded CRISPR/Cas system is crucial to allowing certain Vibrio cholera phages to evade host innate immunity upon infection. In a similar fashion, Le et al. (2013) used functional genomics in order to establish how specific tail fiber genes in P. aeruginosa phages are responsible for their host specificity. The study focused on two such phages which were shown to have different host specificities, namely, JG004 and PaP1. Characterization of spontaneous phage mutants and subsequent functional genomic analysis of bacteriophage JG004 identified that a single point mutation in the tail fiber gene resulted in a broader host range for the phage. Subsequently, replacement of the tail fiber gene of phage PaP1 with that of the highly similar phage JG004 resulted in altered host specificity of PaP1 that was now identical to that of JG004, thus confirming the function of this gene in host recognition. The concept of so-called “reporter phages” which are used in bacterial detection would not be possible without a deep understanding of the functional genomics and phage-host interaction mechanisms of the phage itself, in order to select not only viable but also appropriate insertion sites for reporter genes. One research group (Piuri et al. 2013) developed affinity-tagged phages by insertion of a 10-amino-acid tag to a phage capsid subunit, knowing that modification of the tail may have consequences in terms of adsorption and infectivity, while another group (Schofield et al. 2013) utilized a particular bacteriophage with the ability to transduce a bioluminescent phenotype into target cell in order to develop a technique used for rapid detection of Bacillus anthracis. An amalgamation of the classic and modern techniques of genetic modification may hold the key for the advancement of gene function research. To this end, the recent development of a technique known as “bacteriophage recombineering of electroporated DNA (BRED)” has provided a novel approach to bacteriophage genome modification by using phage-encoded recombination systems in order to engineer deletions, point mutations, and small insertions into a desired location (Marinelli et al. 2012) and has been used previously in reporter phage development (da Silva et al. 2013). It is thought that this technique may become integral to the

Genetics and Genomics of Bacteriophages

211

elucidation of gene functions through physical insertion of an unknown gene into a constructed mutant followed by phenotypic experimentation.

Transcriptomics As the wealth of knowledge about bacteriophages continues to expand, so too does the number of unknown variables associated with the interaction between these viruses and their bacterial counterparts. The study of transcriptomics allows scientists to observe with precision the changes in gene expression of a microorganism under dynamic conditions, including that of exposure to any number of external stresses such as disinfectants, detergents, bacteriocins, and indeed phages themselves. To this end, understanding the response of a bacterium to phage infection may prove pivotal if these viruses and their proteins are to be exploited for pathogen detection and control.

Host Response to Bacteriophage Infection Lytic bacteriophages are commonly observed to gain control of their host through the translation of particular proteins that inactivate the host transcriptional regulators (Seco et al. 2013; Liu et al. 2014a; Tagami et al. 2014). Despite the detailed research to date on phage-host interactions and viral replication within the bacterial cell, there is still much to be learned about the genetic response of the cell itself to the presence of the phage. While this remains a relatively undiscovered area of exploration, the limited numbers of host transcriptome analyses have shed some light on the internal activity of the infected host. Fallico et al. (2011) investigated the transcriptional response of Lactococcus lactis during bacteriophage infection. They found that the bacterium treats the presence of the phage as an attack on the integrity of the cell and as a consequence upregulates transcription of a complex network of proteins which are involved in reinforcement of the cell wall and conservation of energy. They also observed an increase in cellular pathways involved in modification of the cell envelope lipoteichoic acids (LTAs) which may be associated with a non-specific attempt to prevent phage adsorption to the cell by shielding of the receptor binding site. Another study (Ainsworth et al. 2013) also involving L. lactis investigated the transcriptional response of the bacterium when exposed to two different bacteriophage species. Somewhat surprisingly, the results obtained from this research indicated that the response of the L. lactis strain to infection is very much phage specific; only a small portion of differentially expressed genes were found to be shared between the datasets. However, the shared observations in terms of increased energy production, cell wall reinforcement, and nucleotide biosynthesis do coincide with the previous findings by Fallico et al. Lavigne et al. (2013) carried out similar research on a strain of Pseudomonas aeruginosa following exposure to a strictly lytic bacteriophage. In this case, both the

212

A. Casey et al.

bacterium and the phage itself were assessed for their relative expression levels during infection. The foremost observation from this research indicated an association between accumulation of viral transcripts and a rapid depletion of bacterial mRNA, which is likely to hasten the liberation of bacterial transcriptional machinery for use by the infecting phage. Despite the rapid depletion of host mRNA, the analysis still identified over 200 genes which are upregulated in expression upon infection, the majority of which are involved in energy metabolism. Interestingly, there was a notable repression of three type IV pilus genes, whose products are thought to serve as cell surface receptors for that particular bacteriophage. The findings from each of these particular studies seem to agree that the transcriptional response of the host to phage infection focuses mainly on the upregulation of pathways involved in energy metabolism, cell wall reinforcement, and surface LTA modification. Such responses may not however be bacteriophage specific but more so arising as a result of the cell’s reaction to a more general external stress. For example, in the presence of bacteriocins (Liu et al. 2014b) or disinfectants (Casey et al. 2014), the host bacterium is observed to elicit responses concerning cell wall reinforcement and energy metabolism, responses largely similar to those seen upon phage infection.

Discussion and Future Perspectives The digital revolution and the dawn of the information age have modernized bacteriophage research over the last number of decades, as the repository of generated data proceeds to grow at a seemingly exponential rate. With this modernization, particularly in the area of viral metagenomics, comes the realization of the profound impact that bacteriophages have on virtually every environment within the biosphere. Phages represent a legitimate driving force behind the extensive level of bacterial diversification observed on a global scale. This diversification is achieved in a number of different ways, whether through reduction of a dominant bacterial species in an environment through a lytic pathway, or by conferring of resistance to bacteria through lysogeny, horizontal gene transfer, or transduction. There is much still to be learned about the evolution of phages and bacteria alike by investigating horizontal gene transfer and the central role that phages have in this process. Such research will prove pivotal, not only looking forward where understanding and controlling bacterial pathogenesis and acquired antimicrobial resistance is concerned but also looking backward to the evolution of bacteria and phages alike. The presence of so-called “hallmark genes” that appear to connect the virosphere represent the enigma that is bacteriophage evolution. There are a number of working theories as to the precise origins of these particles, theories which seem to contradict one another to a certain extent, yet all have their own merits based on genotypic evidence. The lack of closely related cellular homologues to the “hallmark genes” fuels the ever increasing opinion about an ancient virus world: that viruses predate the last universal cellular ancestor (LUCA). However, scientific discoveries continue

Genetics and Genomics of Bacteriophages

213

to drastically impact the most fundamental of evolutionary theories at every turn. For example, the discovery and sequencing of the mimivirus (La Scola et al. 2003; Raoult et al. 2004), a giant virus of amoebae containing a 1.2 Mb genome with less than 30% homology with any gene in either cellular or indeed viral genomes alike (Holmes 2011), has unquestionably “blurred the boundaries” between viruses and their bacterial counterparts. Although much is now known about the inner workings of bacteriophage genomes themselves, there remains an expanding wealth of genetic data for which an overall function has yet to be assigned. As advances in techniques such as X-ray crystallography and electron microscopy continue to be made, so too will the understanding of the finer structural intricacies of phages. These revelations may hold crucial novel information regarding the role for each gene component in the life cycle of the virus, a belief built on the classical assumption that “structure infers function.” Likewise, the future development of therapeutics against bacterial pathogens will rely heavily on progressing the understanding of phage-host interactions, not only from the perspective of the infecting phage but equally with regard to the response of the bacterium itself to the invading virus. In this sense, visualization of such microbial interactions may provide insights which otherwise could not be inferable from the genomic data alone. Transcriptomics has also become a powerful technique in recent times for comprehending some of the most complex interactions between bacteriophages and their hosts. The increasing efficiency and decreasing costs of microarrays, together with the inception of RNA-Seq, has enabled researchers to elucidate the definitive cellular responses of bacteria to viral infection, a handful of which have been described in this chapter. Assessing and understanding phage-host interactions need to be at the forefront of research given the vast insights it offers, both in antimicrobial development and conversely, in prevention of phage contamination in food processing. Genomic research has already proven to be a valuable asset in discovering the true extent of phage population numbers, diversity, and geographical coverage, as well as opening the doors of opportunity for exploiting phages for a vast array of biological applications. And yet, despite the modest progression observed in virtually every area of bacteriophage genetic and genomic research, the bottom line is that at present, there is simply insufficient data from which to draw any absolute conclusions. The only way to truly solve the mysteries of bacteriophages is to continue developing new tools, techniques, and ideas but most of all, to continue gathering evidence.

References Abedon ST (2000) The murky origin of snow white and her T-even dwarfs. Genetics 155 (2):481–486 Ackermann HW (2007) 5500 phages examined in the electron microscope. Arch Virol 152 (2):227–243 Ackermann HW (2011) Bacteriophage taxonomy. Microbiol Aust 32(2):90–94

214

A. Casey et al.

Ainsworth S, Zomer A, Mahony J, van Sinderen D (2013) Lytic infection of Lactococcus lactis by bacteriophages Tuc2009 and c2 triggers alternative transcriptional host responses. Appl Environ Microbiol 79(16):4786–4798 Akhter S, Aziz RK, Edwards RA (2012) PhiSpy: a novel algorithm for finding prophages in bacterial genomes that combines similarity-and composition-based strategies. Nucleic Acids Res 40(16):e126–e126 Angly FE, Felts B, Breitbart M, Salamon P, Edwards RA, Carlson C, Chan AM, Haynes M, Kelley S, Liu H (2006) The marine viromes of four oceanic regions. PLoS Biol 4(11):e368 Bhattacharya D, Cheng J (2013) 3Drefine: consistent protein structure refinement by optimizing hydrogen bonding network and atomic-level energy minimization. Proteins Struct Funct Bioinf 81(1):119–131 Bondy-Denomy J, Pawluk A, Maxwell KL, Davidson AR (2013) Bacteriophage genes that inactivate the CRISPR/Cas bacterial immune system. Nature 493(7432):429–432 Botstein D (1980) A theory of modular evolution for bacteriophages. Ann NY Acad Sci 354 (1):484–491 Breitbart M (2012) Marine viruses: truth or dare. Mar Sci 4:425–448 Breitbart M, Salamon P, Andresen B, Mahaffy JM, Segall AM, Mead D, Azam F, Rohwer F (2002) Genomic analysis of uncultured marine viral communities. Proc Natl Acad Sci 99 (22):14250–14255 Browning C, Shneider MM, Bowman VD, Schwarzer D, Leiman PG (2012) Phage pierces the host cell membrane with the iron-loaded spike. Structure 20(2):326–339 Busby B, Kristensen DM, Koonin EV (2013) Contribution of phage-derived genomic islands to the virulence of facultative bacterial pathogens. Environ Microbiol 15(2):307–312 Casey A, Fox EM, Schmitz-Esser S, Coffey A, McAuliffe O, Jordan K (2014) Transcriptome analysis of Listeria monocytogenes exposed to biocide stress reveals a multi-system response involving cell wall synthesis, sugar uptake, and motility. Front Microbiol 5(68):1–10 Casjens SR (2005) Comparative genomics and evolution of the tailed-bacteriophages. Curr Opin Microbiol 8(4):451–458 Cavanagh D, Guinane CM, Neve H, Coffey A, Ross RP, Fitzgerald GF, McAuliffe O (2013) Phages of non-dairy lactococci: isolation and characterization of ΦL47, a phage infecting the grass isolate Lactococcus lactis ssp. cremoris DPC6860. Front Microbiol 4(417):1–15 Comeau AM, Bertrand C, Letarov A, Tétart F, Krisch H (2007) Modular architecture of the T4 phage superfamily: a conserved core genome and a plastic periphery. Virology 362(2):384–396 Comeau AM, Tremblay D, Moineau S, Rattei T, Kushkina AI, Tovkach FI, Krisch HM, Ackermann H-W (2012) Phage morphology recapitulates phylogeny: the comparative genomics of a new group of myoviruses. PLoS One 7(7):e40102 d’Herelle F (1917) Sur un microbe invisible antagoniste des bacilles dysentériques. CR Acad Sci Paris 165:373–375 da Silva JL, Piuri M, Broussard G, Marinelli LJ, Bastos GM, Hirata RDC, Hatfull GF, Hirata MH (2013) Application of BRED technology to construct recombinant D29 reporter phage expressing EGFP. FEMS Microbiol Lett 344(2):166–172 De Paepe M, Hutinet G, Son O, Amarir-Bouhram J, Schbath S, Petit M-A (2014) Temperate phages acquire DNA from defective prophages by relaxed homologous recombination: the role of Rad52-like recombinases. PLoS Genet 10(3):e1004181 Dunn JJ, Studier FW, Gottesman M (1983) Complete nucleotide sequence of bacteriophage T7 DNA and the locations of T7 genetic elements. J Mol Biol 166(4):477–535 Fallico V, Ross RP, Fitzgerald GF, McAuliffe O (2011) Genetic response to bacteriophage infection in Lactococcus lactis reveals a four-strand approach involving induction of membrane stress proteins, D-alanylation of the cell wall, maintenance of proton motive force, and energy conservation. J Virol 85(22):12032–12042 Fancello L, Desnues C, Raoult D, Rolain JM (2011) Bacteriophages and diffusion of genes encoding antimicrobial resistance in cystic fibrosis sputum microbiota. J Antimicrob Chemother 66(11):2448–2454

Genetics and Genomics of Bacteriophages

215

Fokine A, Rossmann MG (2014) Molecular architecture of tailed double-stranded DNA phages. Bacteriophage 4(e28281):1–22 Fouts DE (2006) Phage_Finder: automated identification and classification of prophage regions in complete bacterial genome sequences. Nucleic Acids Res 34(20):5839–5851 Garneau JE, Dupuis M-È, Villion M, Romero DA, Barrangou R, Boyaval P, Fremaux C, Horvath P, Magadán AH, Moineau S (2010) The CRISPR/Cas bacterial immune system cleaves bacteriophage and plasmid DNA. Nature 468(7320):67–71 Gibbs AJ (2013) Viral taxonomy needs a spring clean; its exploration era is over. Virol J 10(1):254 Godson GN, Barrell BG, Staden R, Fiddes JC (1978) Nucleotide sequence of bacteriophage G4 DNA. Nature 276(5685):236–247 Hatfull GF (2008) Bacteriophage genomics. Curr Opin Microbiol 11(5):447–453 Hatfull GF, Hendrix RW (2011) Bacteriophages and their genomes. Curr Opin Virol 1(4):298–303 Hendrix RW (2002) Bacteriophages: evolution of the majority. Theor Popul Biol 61(4):471–480 Hendrix RW (2009) Jumbo bacteriophages. Curr Top Microbiol Immunol 328:229–240. Hendrix RW, Smith MCM, Burns RN, Ford ME, Hatfull GF (1999) Evolutionary relationships among diverse bacteriophages and prophages: all the world’s a phage. Proc Natl Acad Sci 96 (5):2192–2197 Hildebrand A, Remmert M, Biegert A, Söding J (2009) Fast and accurate automatic structure prediction with HHpred. Proteins Struct Funct Bioinf 77(S9):128–132 Holmes EC (2011) What does virus evolution tell us about virus origins? J Virol 85(11):5247–5251 Holmfeldt K, Solonenko N, Shah M, Corrier K, Riemann L, VerBerkmoes NC, Sullivan MB (2013) Twelve previously unknown phage genera are ubiquitous in global oceans. Proc Natl Acad Sci 110(31):12798–12803 Jacobs-Sera D, Marinelli LJ, Bowman C, Broussard GW, Guerrero Bustamante C, Boyle MM, Petrova ZO, Dedrick RM, Pope WH, Modlin RL (2012) On the nature of mycobacteriophage diversity and host preference. Virology 434(2):187–201 Jakhetia R, Talukder KA, Verma NK (2013) Isolation, characterization and comparative genomics of bacteriophage SfIV: a novel serotype converting phage from Shigella flexneri. BMC Genomics 14(1):677 Jansen R, Embden J, Gaastra W, Schouls L (2002) Identification of genes that are associated with DNA repeats in prokaryotes. Mol Microbiol 43(6):1565–1575 Kim K-H, Bae J-W (2011) Amplification methods bias metagenomic libraries of uncultured singlestranded and double-stranded DNA viruses. Appl Environ Microbiol 77(21):7663–7668 Kim M-S, Whon TW, Bae J-W (2013) Comparative viral metagenomics of environmental samples from Korea. Genome Inform 11(3):121–128 King AM, Adams MJ, Lefkowitz EJ, Carstens EB (2011) Virus taxonomy: IXth report of the international committee on taxonomy of viruses. Elsevier Academic Press, London Klumpp J, Fouts DE, Sozhamannan S (2013) Bacteriophage functional genomics and its role in bacterial pathogen detection. Brief Funct Genomics 12:354–365 Koonin EV (2005) Virology: Gulliver among the Lilliputians. Curr Biol 15(5):R167–R169 Koonin EV, Dolja VV (2013) A virocentric perspective on the evolution of life. Curr Opin Virol 3 (5):546–557 Koonin EV, Senkevich TG, Dolja VV (2006) The ancient virus world and evolution of cells. Biol Direct 1(1):29 Kostyuchenko VA, Leiman PG, Chipman PR, Kanamaru S, van Raaij MJ, Arisaka F, Mesyanzhinov VV, Rossmann MG (2003) Three-dimensional structure of bacteriophage T4 baseplate. Nat Struct Mol Biol 10(9):688–693 Kristensen DM, Waller AS, Yamada T, Bork P, Mushegian AR, Koonin EV (2013) Orthologous gene clusters and taxon signature genes for viruses of prokaryotes. J Bacteriol 195(5):941–950 Krupovic M, Prangishvili D, Hendrix RW, Bamford DH (2011) Genomics of bacterial and archaeal viruses: dynamics within the prokaryotic virosphere. Microbiol Mol Biol Rev 75(4):610–635 Kuznetsov YG, Chang S-C, Credaroli A, McPherson A (2013) Unique tail appendages of marine bacteriophages. Adv Microbiol 3:55

216

A. Casey et al.

La Scola B, Audic S, Robert C, Jungang L, de Lamballerie X, Drancourt M, Birtles R, Claverie J-M, Raoult D (2003) A giant virus in amoebae. Science 299(5615):2033–2033 Labonté JM, Suttle CA (2013) Metagenomic and whole-genome analysis reveals new lineages of gokushoviruses and biogeographic separation in the sea. Front Microbiol 4 (404):1–11 Lang AS, Zhaxybayeva O, Beatty JT (2012) Gene transfer agents: phage-like elements of genetic exchange. Nat Rev Microbiol 10(7):472–482 Lavigne R, Lecoutere E, Wagemans J, Cenens W, Aertsen A, Schoofs L, Landuyt B, Paeshuyse J, Scheer M, Schobert M (2013) A multifaceted study of Pseudomonas aeruginosa shutdown by virulent podovirus LUZ19. MBio 4(2):e00061-00013 Le S, He X, Tan Y, Huang G, Zhang L, Lux R, Shi W, Hu F (2013) Mapping the tail fiber as the receptor binding protein responsible for differential host specificity of Pseudomonas aeruginosa bacteriophages PaP1 and JG004. PLoS One 8(7):e68562 Leiman P, Kanamaru S, Mesyanzhinov V, Arisaka F, Rossmann M (2003) Structure and morphogenesis of bacteriophage T4. Cell Mol Life Sci 60(11):2356–2370 Lewis TE, Sillitoe I, Andreeva A, Blundell TL, Buchan DW, Chothia C, Cuff A, Dana JM, Filippis I, Gough J (2013) Genome3D: a UK collaborative project to annotate genomic sequences with predicted 3D structures based on SCOP and CATH domains. Nucleic Acids Res 41(D1):D499–D507 Li J, Halgamuge SK, Tang S-L (2008) Genome classification by gene distribution: an overlapping subspace clustering approach. BMC Evol Biol 8(1):116 Liu B, Shadrin A, Sheppard C, Mekler V, Xu Y, Severinov K, Matthews S, Wigneshweraraj S (2014a) The sabotage of the bacterial transcription machinery by a small bacteriophage protein. Bacteriophage 4(e28520):1–4 Liu X, Basu U, Miller P, McMullen LM (2014b) Stress and adaptation of Listeria monocytogenes 08-5923 exposed to a sublethal dose of carnocyclin A. Appl Environ Microbiol 80:3835–3841. 00350-00314 Marinelli LJ, Hatfull GF, Piuri M (2012) Recombineering: a powerful tool for modification of bacteriophage genomes. Bacteriophage 2(1):5–14 Martinsohn JT, Radman M, Petit M-A (2008) The λ red proteins promote efficient recombination between diverged sequences: implications for bacteriophage genome mosaicism. PLoS Genet 4 (5):e1000065 Marvin DA, Symmons MF, Straus SK (2014) Structure and assembly of filamentous bacteriophages. Prog Biophys Mol Biol 114(2):80–122 Matos RC, Lapaque N, Rigottier-Gois L, Debarbieux L, Meylheuc T, Gonzalez-Zorn B, Repoila F, Lopes M d F, Serror P (2013) Enterococcus faecalis prophage dynamics and contributions to pathogenic traits. PLoS Genet 9(6):e1003539 Maurice C, Bouvier C, Wit R, Bouvier T (2013) Linking the lytic and lysogenic bacteriophage cycles to environmental conditions, host physiology and their variability in coastal lagoons. Environ Microbiol 15(9):2463–2475 Maxwell KL, Davidson AR (2013) A shifty chaperone for phage tail assembly. J Mol Biol 426 (5):1001–1003 McNair K, Bailey BA, Edwards RA (2012) PHACTS, a computational approach to classifying the lifestyle of phages. Bioinformatics 28(5):614–618 Miller ES, Kutter E, Mosig G, Arisaka F, Kunisawa T, Rüger W (2003) Bacteriophage T4 genome. Microbiol Mol Biol Rev 67(1):86–156 Mizuno CM, Rodriguez-Valera F, Kimes NE, Ghai R (2013) Expanding the marine virosphere using metagenomics. PLoS Genet 9(12):e1003987 Modi SR, Lee HH, Spina CS, Collins JJ (2013) Antibiotic treatment expands the resistance reservoir and ecological network of the phage metagenome. Nature 499(7457):219–222 Mokili JL, Rohwer F, Dutilh BE (2012) Metagenomics and future perspectives in virus discovery. Curr Opin Virol 2(1):63–77 Muniesa M, Allué-Guardia A, Martínez-Castillo A (2013) Bacteriophage-driven emergence of novel pathogens. Futur Virol 8(4):323–325

Genetics and Genomics of Bacteriophages

217

Orlova EV, Gowen B, Dröge A, Stiege A, Weise F, Lurz R, van Heel M, Tavares P (2003) Structure of a viral DNA gatekeeper at 10 Å resolution by cryo-electron microscopy. EMBO J 22 (6):1255–1262 Pedulla ML, Ford ME, Houtz JM, Karthikeyan T, Wadsworth C, Lewis JA, Jacobs-Sera D, Falbo J, Gross J, Pannunzio NR (2003) Origins of highly mosaic mycobacteriophage genomes. Cell 113 (2):171–182 Petrov VM, Ratnayaka S, Nolan JM, Miller ES, Karam JD (2010) Genomes of the T 4-related bacteriophages as windows on microbial genome evolution. Virol J 7:292 Piuri M, Rondón L, Urdániz E, Hatfull GF (2013) Generation of affinity-tagged fluoromycobacteriophages by mixed assembly of phage capsids. Appl Environ Microbiol 79(18):5608–5615 Proux C, van Sinderen D, Suarez J, Garcia P, Ladero V, Fitzgerald GF, Desiere F, Brüssow H (2002) The dilemma of phage taxonomy illustrated by comparative genomics of Sfi21-like Siphoviridae in lactic acid bacteria. J Bacteriol 184(21):6026–6036 Quirós P, Colomer-Lluch M, Martínez-Castillo A, Miró E, Argente M, Jofre J, Navarro F, Muniesa M (2014) Antibiotic resistance genes in the bacteriophage DNA fraction of human fecal samples. Antimicrob Agents Chemother 58(1):606–609 Raoult D, Audic S, Robert C, Abergel C, Renesto P, Ogata H, La Scola B, Suzan M, Claverie J-M (2004) The 1.2-megabase genome sequence of mimivirus. Science 306(5700):1344–1350 Ray J, Dondrup M, Modha S, Steen IH, Sandaa R-A, Clokie M (2012) Finding a needle in the virus metagenome haystack-micro-metagenome analysis captures a snapshot of the diversity of a bacteriophage armoire. PLoS One 7(4):e34238 Rohwer F, Edwards R (2002) The phage proteomic tree: a genome-based taxonomy for phage. J Bacteriol 184(16):4529–4535 Sanger F, Air G, Barrell B, Brown N, Coulson A, Fiddes C, Hutchison C, Slocombe P, Smith M (1977) Nucleotide sequence of bacteriophage phi X174 DNA. Nature 265(5596):687–695 Sanger F, Coulson AR, Hong G, Hill D, Petersen G (1982) Nucleotide sequence of bacteriophage λ DNA. J Mol Biol 162(4):729–773 Schoenfeld T, Liles M, Wommack KE, Polson SW, Godiska R, Mead D (2010) Functional viral metagenomics and the next generation of molecular tools. Trends Microbiol 18(1):20–29 Schofield DA, Sharp NJ, Vandamm J, Molineux IJ, Spreng KA, Rajanna C, Westwater C, Stewart GC (2013) Bacillus anthracis diagnostic detection and rapid antibiotic susceptibility determination using ‘bioluminescent’ reporter phage. J Microbiol Methods 95(2):156–161 Seco EM, Zinder JC, Manhart CM, Piano AL, McHenry CS, Ayora S (2013) Bacteriophage SPP1 DNA replication strategies promote viral and disable host replication in vitro. Nucleic Acids Res 41(3):1711–1721 Seed KD, Lazinski DW, Calderwood SB, Camilli A (2013) A bacteriophage encodes its own CRISPR/Cas adaptive response to evade host innate immunity. Nature 494(7438):489–491 Sullivan MB, Coleman ML, Quinlivan V, Rosenkrantz JE, DeFrancesco AS, Tan G, Fu R, Lee JA, Waterbury JB, Bielawski JP (2008) Portal protein diversity and phage ecology. Environ Microbiol 10(10):2810–2823 Summer EJ, Gonzalez CF, Bomer M, Carlile T, Embry A, Kucherka AM, Lee J, Mebane L, Morrison WC, Mark L, King MD, LiPuma JJ, Vidaver AK, Young R (2006) Divergence and mosaicism among virulent soil phages of the Burkholderia cepacia complex. J Bacteriol 188 (1):255–268 Suttle CA (2007) Marine viruses – major players in the global ecosystem. Nat Rev Microbiol 5 (10):801–812 Swanson MM, Reavy B, Makarova KS, Cock PJ, Hopkins DW, Torrance L, Koonin EV, Taliansky M (2012) Novel bacteriophages containing a genome of another bacteriophage within their genomes. PLoS One 7(7):e40683 Swenson KM, Guertin P, Deschênes H, Bergeron A (2013) Reconstructing the modular recombination history of Staphylococcus aureus phages. BMC Bioinf 14(Suppl 15):S17 Tagami S, Sekine S-i, Minakhin L, Esyunina D, Akasaka R, Shirouzu M, Kulbachinskiy A, Severinov K, Yokoyama S (2014) Structural basis for promoter specificity switching of RNA polymerase by a phage factor. Genes Dev 28(5):521–531

218

A. Casey et al.

Thurber RV, Haynes M, Breitbart M, Wegley L, Rohwer F (2009) Laboratory procedures to generate viral metagenomes. Nat Protoc 4(4):470–483 Turner D, Reynolds D, Seto D, Mahadevan P (2013) CoreGenes3. 5: a webserver for the determination of core genes from sets of viral and small bacterial genomes. BMC Res Notes 6(1):140 Twort FW (1915) An investigation on the nature of ultra-microscopic viruses. Lancet 186 (4814):1241–1243 Vannucci FA, Kelley MR, Gebhart CJ (2013) Comparative genome sequencing identifies a prophage-associated genomic island linked to host adaptation of Lawsonia intracellularis infections. Vet Res 44(1):49 Veesler D, Cambillau C (2011) A common evolutionary origin for tailed-bacteriophage functional modules and bacterial machineries. Microbiol Mol Biol Rev 75(3):423–433 Xia H, Li T, Deng F, Hu Z (2013) Freshwater cyanophages. Virol Sin 28(5):253–259 Xu J, Hendrix RW, Duda RL (2013) Chaperone–protein interactions that mediate assembly of the bacteriophage lambda tail to the correct length. J Mol Biol 426(5):1004–1018

Bacteriophage Discovery and Genomics Graham F. Hatfull

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Discovery and Genomics as a Platform for Science Education . . . . . . . . . . . . . . . . . . . . . . . . . The Phage Hunters Integrating Research and Education (PHIRE) Program . . . . . . . . . . . . . . . The Mycobacterial Genetics Course . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Science Education Alliance Phage Hunters Advancing Genomics and Evolutionary Science (SEA-PHAGES) Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A View of Bacteriophage Genetic Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

220 221 223 224 225 226 228 228 228

Abstract

Bacteriophages have offered opportunities for promoting student education for many years. The advancement of high throughput DNA sequencing technologies, however, has now paved the way for fuller student engagement in phage discovery and genomics. Because of the extraordinarily high diversity of phage genomes, there is a near-limitless resource of bacteriophages that can be isolated from the environment using a bacterial host of one’s choice, with strong prospects of isolating a phage that is replete with genetic novelty. The phage discovery and genomics platform is imbued with flexibility and has been implemented as openended research experiences for high school and undergraduate students in the Phage Hunters Integrating Research and Education (PHIRE) program, as an intense 2-week workshop in the Mycobacterial Genetic Course in Durban South Africa, and as a community of institutions providing course-based research experiences for freshman undergraduates in the Science Education Alliance Phage Hunters Advancing Genomics and Evolutionary Science (SEA-PHAGES) G. F. Hatfull (*) Department of Biological Sciences, University of Pittsburgh, Pittsburgh, PA, USA e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_6

219

220

G. F. Hatfull

program. Here, I discuss the essential components of the phage discovery and genomic platform, three different configurations in which it has been implemented, and review the insights into phage diversity and genomics that have been gleaned.

Introduction In 2012, the President’s Council of Advisors on Science and Technology delivered a report on Science Education in the United States and presented five recommendations aimed at enhancing the number of undergraduate students graduating with degrees in Science, Technology, Engineering, and Mathematics (STEM) disciplines (PCAST 2012). The second of these recommendations was to “advocate and provide support for replacing standard laboratory courses with discovery-based research courses,” with a focus on students in their first 2 years of college. Although at first this might seem quite a simple directive, there are many challenges presented, such as identifying effective research questions, maintaining a focus on authentic scientific discovery, training instructors who may lack experience in the research topic, and implementing it all at scale to broadly impact students, not just a self-selected few. Standard introductory laboratory courses that are not research-based have developed in part because of these challenges and because teaching hundreds or thousands of students is simpler with cookbook instruction of techniques and approaches, using part-time or nontenure track faculty who do not have extensive research expertise. Replacing these standard laboratory courses with research-based courses is, however, highly attractive, as there is excellent evidence supporting gains in student learning when they engage in research activities (Seymour et al. 2004; Hunter et al. 2007; Lopatto 2007; Eagan et al. 2011, 2013). But “research” experiences can be highly varied and may focus on research mechanics and/or structured inquiry, rather than authentic discovery-based experiences generating insights that are publishable in the peer-reviewed literature. Identifying such authentic research activities that are suitable for early career students who may have little expert content knowledge, have no prior research experience, and are likely entering a laboratory for the first time represents a substantial challenge (Auchincloss et al. 2014). Even if suitable projects can be defined, implementing them at community and some 4-year colleges which – unlike major research universities – do not have robust research infrastructures is a tough proposition. Replacing one standard laboratory course with a research-based course at one institution does not seem so daunting, but will not succeed in a broad transformation of undergraduate education at a nationwide level. Bacteriophage research would seem to be a good place to seek research projects that early career students can engage in. It is 100 years since phages were discovered and the methods developed early on for isolating and enumerating phage particles have proven to be powerful and robust (Rohwer et al. 2014) (see chapter ▶ “Bacteriophage Ecology”). Phage studies have long been components of undergraduate curricula, although their use in undergraduate science education has

Bacteriophage Discovery and Genomics

221

arguably declined since their halcyon days of a few decades ago. General interest in phage biology, however, has perked up somewhat over the past few years, in part because of insights by phage ecologists regarding phage numbers, dynamics, and roles in biogeochemical cycles (Rohwer and Thurber 2009; Danovaro et al. 2011), in part because of anxiety about how to respond to the growing prevalence of antimicrobial resistance of pathogens (Chan et al. 2013; Nobrega et al. 2015), and in part because of the effective use of genomics to gain insights into viral origins and evolution (Hatfull 2015a). Given the increased interest in phage biology along with the relative ease with which phages can be both isolated and manipulated experimentally, here we consider how to return phages to the undergraduate curriculum as a research-based component of science training. We focus particularly on platforms of phage discovery and genomics as exemplified by programs such as the Phage Hunters Integrating Research and Education (PHIRE), the Mycobacterial Genetics Course (MGC), and the Science Education Alliance Phage Hunters Advancing Genomics and Evolutionary Science (SEA-PHAGES).

Phage Discovery and Genomics as a Platform for Science Education The underlying rationale to the phage discovery and genomics platform is a simple one. The phage population is huge with 1031 phage particles in the biosphere, the population is highly dynamic with an estimated 1023 productive infections per second on a global scale, and the population is old, perhaps as old as microbial life itself (Bergh et al. 1989; Hendrix 1999, 2002). It thus is no surprise that the phage population is highly diverse, i.e., that there are large numbers of different types of phages in the biosphere (Hambly and Suttle 2005; Pope et al. 2015). Because phages are a majority of biological entities in the natural world, it behooves us to seek to define and understand bacteriophage diversity, including in terms of the rich reservoir of undiscovered genes that they encode. There are two core approaches that can be applied to try and define this diversity, and both have important roles to play. The first is a metagenomic approach where total viral particles are harvested from environment samples and sequenced, providing huge amounts of sequence information (Angly et al. 2006; Mokili et al. 2012; Brum and Sullivan 2015). This approach has been especially powerful for illustrating the vast extent of viral diversity and for monitoring dynamic changes in viral populations, but it is still rare to obtain complete or near-complete phage genome assemblies from this information. The second approach is the genome-by-genome strategy that forms the basis of the phage discovery and genomics platform. In this approach, individual phages are isolated from the environment using a known bacterial host, the phages are purified, amplified, the DNA extracted, the genome sequenced and annotated, followed by comparative genomic analysis (Hanauer et al. 2006, 2009). The genome-by-genome approach has an advantage over the metagenomic approach in that it not only generates data that is stored in the

222

G. F. Hatfull

computer, but individual phages are stored in the freezer and thereby available for genetic and biochemical characterization. In general, virtually any bacterial strain could be used for the phage discovery and genomics platform. In practice, however, some are easier than others to implement in the classroom and the behaviors of different platforms are not easy to predict a priori. For use in an educational setting, there are several key criteria. First, the strain should be nonpathogenic and, ideally, designated as a risk group 1 organism. Second, the growth rate should be tolerable, such that it does not take more than a day or so to grow a confluent bacterial lawn. Third, it is advisable to choose bacteria that are easy and cheap to culture, thereby avoiding bacteria that have special growth requirements, including most anaerobes. Fourth, bacteria and their phages are typically found in similar environments, so it is helpful if the bacteria or closely related strains have been previously detected in similar samples from which the phages will be isolated. It is also useful if the genome sequence of the bacterium has been determined, and there are obvious benefits to working with a host that has relevance to human health, veterinary concerns, or agricultural or environmental issues. For example, phage discovery courses taught at Texas A&M University for more than 15 years have included hosts such as Bacillus megaterium, Caulobacter crescentus, Klebsiella pneumoniae, Citrobacter freundii, Escherichia coli, and Acinetobacter baumannii (Farmer et al. 2013; Bernal et al. 2015; Brizendine et al. 2015; Doan et al. 2015; Edwards et al. 2015; Lerma et al. 2015; Mijalis et al. 2015). The bacterial host that we have used most extensively is Mycobacterium smegmatis mc2155 (Snapper et al. 1990). This is a high efficiency transformation strain derived from the original M. smegmatis ATCC607 stock by Dr. Jacobs and colleagues and has become a workhorse for mycobacterial genetics. M. smegmatis is a nonpathogen and is classified as a risk group 1 agent by the Centers for Disease Control (although a small number of M. smegmatis infections have been reported). It is reported to be present in some soil samples, and other mycobacteria and Actinobacteria in general are common soil inhabitants. Phages are typically present in environments in which their hosts live, and M. smegmatis has been used since the 1950s to isolate phages from environmental samples. It is also known to be both CRISPR- and restriction-free (Jacobs et al. 1987). There are two common approaches to the phage isolation procedure: direct plating and enrichment. In direct plating, samples of soil or compost are briefly extracted with buffer, filtered, mixed with the bacterial host, and plated to form lawns. Using M. smegmatis, we typically see only a small number of plaques in any given sample, and only about 10% of samples yield plaques. In the enrichment procedure, the environmental sample is incubated in the presence of the host for several hours or days and then plated onto a lawn of bacterial host cells. Using M. smegmatis, the yield is substantially higher with more than 50% of samples producing plaques. The titer per enrichment culture also varies from a few plaques to 109 plaque forming units (pfu) per milliliter. Once plaques are recovered, they are purified through several rounds of re-plating and then amplified to generate high titer lysates. Sufficient DNA can be readily

Bacteriophage Discovery and Genomics

223

extracted from approximately 1010 pfu for DNA sequencing. Many different platforms are available for sequencing, although we currently favor multiplexing libraries on a single Illumina MiSeq run, which produces high quality assemblies, and from which the genome termini can often be predicted. Sequence ambiguities and genome end conformation can be clarified by direct sequencing from genome DNA templates using specific primers and phage genomic DNA. Annotation of phage genomes presents interesting challenges but also rich opportunities for science education. Phage genomes are typically replete with small open reading frames and large numbers of previously unidentified genes (Pope et al. 2015). Automated gene calling programs such as GeneMark and Glimmer do a reasonable job of providing a first approximation of gene positions, but careful manual inspection is required to enhance the veracity of putative gene start sites, to identify genes missed by automation, to avoid errantly overlapping genes, and to deal with programmed translational frameshifts, splicing, and other noncanonical gene features. We have made effective use of the DNA Master analysis program (http://cobamide2. bio.pitt.edu/), which facilitates rapid automated gene predictions and careful manual revision. The program Phamerator (Cresawn et al. 2011) enables effective genome comparisons and alignments of genomes according to both their nucleic acid and predicted protein products. The processes described above are all generally accessible without highly specialized content or technical knowledge and are therefore well suited for use by high school and undergraduate students, although effectively anyone with a sense of curiosity and a willingness and interest to explore the viral world can participate. The initial steps in phage isolation are simple with subsequent steps following a structured progress towards the more complex and abstract processes of genome annotation and comparative analysis. We have implemented this phage discovery and genomics platform in three distinct configurations aimed at different groups of students, different contexts, and different institutions. Each of these is described below.

The Phage Hunters Integrating Research and Education (PHIRE) Program The PHIRE program implements this basic phage discovery and genomics platform within our research laboratory at the University of Pittsburgh, as has been supported by a Howard Hughes Professorship Award starting in 2002 (Hanauer et al. 2006; Hatfull et al. 2006; Hatfull 2010, 2015b). It was designed as a means of introducing high school and undergraduate students to scientific research by immersing them in the culture and practices of the research laboratory. It also provided a means of exploring and developing the separate elements of the phage discovery and genomics platform. The scientific motivation was based on the desire to extend our understanding of mycobacteriophage genome diversity, building on a small collection of about a dozen genomes we had sequenced between 1993 and 2002 (Ford et al. 1998a, b; Mediavilla et al. 2000; Pedulla et al. 2003). The educational

224

G. F. Hatfull

challenge was to see how well this research experience was suited to novice scientists and what learning benefits accrued (Hanauer et al. 2009). Throughout it was the intent to more fully integrate our research and educational missions, which seem to have all too commonly become fractured and disparate. The phage isolation component of this discovery and genomics platform is clearly not new, and with M. smegmatis specifically Dr. Bill Jacobs was isolating phages (including Bxb1 and Bxz1) in the mid-late 1980s (Mediavilla et al. 2000; Pedulla et al. 2003). The ability to sequence the genomes and learn about diversity and evolutionary mechanisms, however, has only recently been just catching up. At the onset of the PHIRE program, automated sequencing methods were available using shotgun clones, but phage genomes in the 50–100 kbp size range still presented considerable challenges, especially for novice student scientists. Nonetheless, a substantial number of students moved through the PHIRE program, many of whom continued onto further education and careers in the sciences (Hanauer et al. 2009). The program also illustrated facets of the platform that we predicted to be associated with its overall impact, such as project ownership, a structured transition from concrete to abstract data, flexibility in timing and with milestones, and a parallel project structure that promotes peer-mentoring opportunities and facilitates inclusion of greater numbers of students (Hatfull et al. 2006; Hanauer et al. 2009). Most importantly, students can succeed in this research without extensive prior content or technical understanding. This latter point is a particularly important one. Only a program that does not select students according to prior “academic success” or some type of “gifted” status can engage the full diversity – demographic and intellectual – of students displaying any degree of interest in STEM subjects. Thus, students have the opportunity to discover whether they have an aptitude for scientific research before being selected against by some artificial criteria. The PHIRE program also produced a substantial number of peer-reviewed publications, many with student co-authors, illustrating the authenticity of the research experience (Hatfull 2015b).

The Mycobacterial Genetics Course A second configuration of the phage discovery and genomics platform is its implementation within an intense 2-week Mycobacterial Genetics Course at the University of KwaZulu-Natal (UKZN) in Durban, South Africa. This was offered initially in 2008 and was the brainchild of Drs. Michelle Larsen and Bill Jacobs, and further advanced by Deborah Jacobs-Sera. It was developed in parallel with the HHMI initiative to advance the KwaZulu-Natal Research Institue for TB-HIV (K-RITH) at the UKZN Medical School. It has been offered each year since then and typically enrolls 24–30 students, ranging from undergraduates and honors students to masters students, but typically represents a first experience in doing research. Because of the tight timeframe, it is not plausible (yet) to have the students annotate the sequences of the genomes they have isolated. But typically each of the students is able to isolate a phage from the Durban area, purify it, and amplify it. They also have the

Bacteriophage Discovery and Genomics

225

opportunity to examine the viral morphologies by electron microscopy, and other phage characteristics such as patterns of host resistance or superinfection immunity. The genome annotation effort focuses on phage genomes isolated by students in the prior year’s course, and typically 4–6 genomes are fully annotated, reviewed for quality control, and submitted to GenBank prior to the conclusion of the course. The implementation of the phage discovery and genomics platform in the compressed 2-week course illustrates much of its inherent flexibility. Even though some compromises are needed – including annotation of previously isolated genomes – assessment of student gains suggests that much of the education advantages observed in the PHIRE and SEA-PHAGES programs (see below) are preserved. This therefore would seem to represent a useful model for implementation in other more time-constrained contexts such as summer bridge programs or courses to introduce the general public as well as administrators, politicians, and policy makers to science and scientific research.

The Science Education Alliance Phage Hunters Advancing Genomics and Evolutionary Science (SEA-PHAGES) Program The most expansive implementation of the phage discovery and genomics platform is in the SEA-PHAGES program, in which trained faculty at multiple institutions offer a course in phage discovery and genomics to early career undergraduate students (Jordan et al. 2014). The SEA-PHAGES initiative began in 2007 in discussions with HHMI staff, and the first cohort of 12 institutions offered the course in 2008. SEA-PHAGES builds upon the concepts emerging from the PHIRE program, especially the opportunities afforded by authentic research, an open inclusion for all students, and a parallel project structure supporting implementation at-scale. The dissemination and expansion in numbers of students requires a number of structural components that are critical to program success, including systems for training faculty and for organizing data. The program has been jointly administered since 2008 by our group at the University of Pittsburgh, Dr. Cresawn at James Madison University, and HHMI staff. Through addition of member institutions each year, since 2008, the program has grown to include about 95 participating schools and over 3,200 students in the 2015–2016 school year. Implementation of the SEA-PHAGES course differs somewhat from institution to institution, while maintaining the core components. For the most part, the course is offered for first year students in substitution for a more standard introductory biology laboratory course. It is taught over two terms (usually but not exclusively Fall – Spring), typically for 4 h each week in two 2-h sessions, with the first term encompassing the microbiological components of phage isolation, purification, amplification, and DNA isolation, and the second term focusing primarily on genome annotation and comparative genomic analyses. Phage genomes are sequenced centrally during the period from mid-November to early January and returned for annotation in the Spring term. Many schools find it convenient to teach a single section of 18–24 students when first offering the course, with the prospects of

226

G. F. Hatfull

expanding to multiple sections in future years and enrolling substantial proportions of all first year biology students; a few schools target either honors students or, alternatively, at-risk students. The size and disseminated nature of the SEA-PHAGES program requires a considerable supporting structure. Essential structural components are two databases with web interfaces: phagesdb.org (http://phagesdb.org) and seaphagesdb.org (http:// seaphagesdb.org). Phagesdb.org is a database containing all of the information about the individual phages, such as where they were found, who found them, genome length, sequence information, electron micrographs. Individual students enter and update their information on the phages they have isolated, and the information is readily available to the larger SEA-PHAGES and scientific communities. Seaphages. org is a separate database that contains programmatic information, such as the institution and faculty information, numbers of students enrolled, and includes forums for faculty discussions, news items, and other programmatic information. Both databases are publicly available. An attractive feature of the SEA-PHAGES program is that the course can be implemented at many different types of institutions, including those that do not have robust research infrastructures. Furthermore, although many instructors may have some knowledge of microbiology, few have experience in phage biology, and two 1-week training workshops prepare faculty to teach the course (one focusing on microbiology, the other on bioinformatics). Newly joining schools are paired with a “buddy” school that – along with program administrators – can provide ready assistance for troubleshooting. An annual scientific symposium is held at the HHMI Janelia Research Campus with faculty and students from each of the participating schools attending. The main scientific focus dominating the first few years of the SEA-PHAGES program was to describe the genetic diversity of phages infecting M. smegmatis mc2155 (Pope et al. 2015). Because of constraints on genome sequencing both with regards to time and costs, in the early years only one of the phages isolated in the first term at each school was sequenced and then jointly annotated by students in the second term. As the sequencing technologies have advanced, one phage for each student section could be sequenced, and the capacity is gradually expanding to encompass two or more phages per section of students. As a consequence, the size of the collection of sequenced mycobacteriophages has exploded, rising from about a dozen at the start of the program in 2008, to over 900 at the time of writing. This has also prompted a broadening of the scientific scope to include phages of all bacterial hosts within the phylum Actinobacteria, with the aim of sampling deeply (>500 genomes) as many of the thousands of strains within this phylogenetic space as possible.

A View of Bacteriophage Genetic Diversity These 900 sequenced mycobacteriophages – by far the largest collection of sequenced phages known to infect a single common bacterial host strain – display remarkable diversity and insights into viral evolution (Pope et al. 2015). In general,

Bacteriophage Discovery and Genomics

227

they are all double-stranded DNA tailed phages (members of the Caudovirales) and all have either flexible noncontractile tails (siphoviridae) or contractile tails (myoviridae); the lack of phages with short, stubby tails (podoviridae) may reflect the challenges imposed to phage infection by the thick, complex, and lipid-rich mycobacterial cell walls. There are only two instances in which two identical phages were isolated at different locations and different times (there are other instances of identity where cross contamination is a plausible explanation), and all others are “different.” However, the types of differences are highly varied and range from genomes that share no DNA sequence similarity, to those that have a few single nucleotide changes, and everything in-between including extensive nucleotide sequence similarity over parts of the genomes, but dissimilarity over the rest, or having extensive nucleotide sequence similarity but differing in gene content due to insertions and deletions. To recognize that there are groups of phages with similar genomes and genome organizations we have grouped them into “clusters.” The primary parameter for assigning membership into the same cluster is identifiable nucleotide sequence similarity spanning greater than 50% of the genome lengths, but also taken into consideration are average nucleotide identities and gene content (Hatfull et al. 2010). Currently there are 23 clusters (Clusters A – W; each of which can also be quite diverse within the cluster) and seven singletons (individual phages with no close relatives), or about 30 types in total that do not share extensive sequence similarity between types. The relationships are complex, however, in large part because of the mosaic nature of the genomes, with segments, often single genes, shared among other dissimilar genomes. As such the grouping into clusters is a taxonomy of convenience and there are no well-defined lines separating many of the clusters. As the collection has grown, so too has the evidence that the diversity is best viewed as a continuum, within which some parts are more highly sampled than others (Pope et al. 2015). We predict that sequencing large numbers of phages of closely related hosts will further populate and embellish this continuum. It is notable that there are several examples of phages that have distinctly different gene coding profiles from their host M. smegmatis. For some phages (e.g., Cluster L), this is readily apparent by comparing coding predictions using GeneMark and comparing the profiles reported using an M. smegmatis model and one using a heuristic model. Other examples are seen in the Cluster U phage, Patience, where the GC% content (50.3%) is substantially different to M. smegmatis (67.3%) and the codon usage tables are, not surprisingly, also very different (Pope et al. 2014). These examples support a model and a general view that phages migrate rather rapidly across the landscape of potential bacterial hosts and – provided that hosts are available – do so faster than their genomes ameliorate to have profiles similar to their hosts (Jacobs-Sera et al. 2012). These phage genomic profiles thus likely reflect the current host ranges of the phages, but also the hosts that they have infected along their evolutionary paths. Thus, if we had broad and deep sampling of phages across the entire spectrum of hosts in the phylum, we may be able to at least partially reconstruct these pathways. We hope that future efforts in phage genomics fueled by active integration of science education and viral discovery will test these hypotheses.

228

G. F. Hatfull

Concluding Remarks The three programs described here represent just a few of the possible ways in which phage discovery and genomics can be used in educational settings. Ry Young and colleagues have a successful program at Texas A&M University in phage genomics that has been offered since the 1999–2000 academic year and engages about 20 students in a three-credit course targeted at juniors and seniors. The hosts used are environmentally important ones and are phylogenetically diverse, and numerous genomes have been sequenced and published. Continued expansion and extension of the SEA-PHAGES course is anticipated to contribute substantially to our understanding of phages infecting hosts across the phylum Actinobacteria and that this will provide phylogenetic coherence and further insights into specific pathways in phage evolution. The pace of phage discovery and genome sequencing technologies suggests that these efforts will generate a vast amount of sequence information and the identification of thousands if not millions of genes of unknown function. The challenge will be to find methods that can be effectively implemented for gene function determination, using genetic, biochemical, and structural approaches. The freezers are filling with thousands of individual phage isolates providing a biological context for rich discoveries in the years to come.

Cross-References ▶ Genetics and Genomics of Bacteriophages Acknowledgments I thank Deborah Jacobs-Sera and Welkin Pope for comments on the manuscript. This work was supported by a grant from the Howard Hughes Medical Institute.

References Angly FE, Felts B, Breitbart M, Salamon P, Edwards RA, Carlson C, Chan AM, Haynes M, Kelley S, Liu H, Mahaffy JM, Mueller JE, Nulton J, Olson R, Parsons R, Rayhawk S, Suttle CA, Rohwer F (2006) The marine viromes of four oceanic regions. PLoS Biol 4(11):e368 Auchincloss LC, Laursen SL, Branchaw JL, Eagan K, Graham M, Hanauer DI, Lawrie G, McLinn CM, Pelaez N, Rowland S, Towns M, Trautmann NM, Varma-Nelson P, Weston TJ, Dolan EL (2014) Assessment of course-based undergraduate research experiences: a meeting report. CBE Life Sci Educ 13(1):29–40 Bergh O, Borsheim KY, Bratbak G, Heldal M (1989) High abundance of viruses found in aquatic environments. Nature 340(6233):467–468 Bernal CL, Berkowitz VE, Cahill JL, Rasche ES, Kuty Everett GF (2015) Complete genome sequence of Citrobacter freundii myophage Michonne. Genome Announc 3(5). https://doi. org/10.1128/genomeA.01134–15. PubMed PMID: 26430047; PubMed Central PMCID: PMCPMC4591319 Brizendine AM, Rousseau S, Hernandez AC, Kuty Everett GF (2015) Complete genome sequence of Bacillus megaterium siphophage Stahl. Genome Announc 3(4). https://doi.org/10.1128/ genomeA.00857–15. PubMed PMID: 26251491; PubMed Central PMCID: PMCPMC4541283

Bacteriophage Discovery and Genomics

229

Brum JR, Sullivan MB (2015) Rising to the challenge: accelerated pace of discovery transforms marine virology. Nat Rev Microbiol 13(3):147–159 Chan BK, Abedon ST, Loc-Carrillo C (2013) Phage cocktails and the future of phage therapy. Future Microbiol 8(6):769–783 Cresawn SG, Bogel M, Day N, Jacobs-Sera D, Hendrix RW, Hatfull GF (2011) Phamerator: a bioinformatic tool for comparative bacteriophage genomics. BMC Bioinformatics 12:395 Danovaro R, Corinaldesi C, Dell’anno A, Fuhrman JA, Middelburg JJ, Noble RT, Suttle CA (2011) Marine viruses and global climate change. FEMS Microbiol Rev 35(6):993–1034 Doan DP, Lessor LE, Hernandez AC, Kuty Everett GF (2015) Complete genome sequence of enterotoxigenic Escherichia coli siphophage Seurat. Genome Announc 3(1). https://doi.org/10. 1128/genomeA.00044–15. PubMed PMID: 25720682; PubMed Central PMCID: PMCPMC4342423 Eagan MK Jr, Sharkness J, Hurtado S, Mosqueda CM, Chang MJ (2011) Engaging undergraduates in science research: not just about faculty willingness. Res High Educ 52(2):151–177 Eagan MK, Hurtado S, Chang MJ, Garcia GA, Herrera FA, Garibay JC (2013) Making a difference in science education: the impact of undergraduate research programs. Am Educ Res J 50(4):683–713 Edwards GB, Luna AJ, Hernandez AC, Kuty Everett GF (2015) Complete genome sequence of Citrobacter freundii myophage Moon. Genome Announc 3(1). https://doi.org/10.1128/genomeA. 01427–14. PubMed PMID: 25635027; PubMed Central PMCID: PMCPMC4319498 Farmer NG, Wood TL, Chamakura KR, Kuty Everett GF (2013) Complete genome of Acinetobacter baumannii N4-Like Podophage Presley. Genome Announc 1(6). https://doi.org/ 10.1128/genomeA.00852–13. PubMed PMID: 24309722; PubMed Central PMCID: PMCPMC3853045 Ford ME, Sarkis GJ, Belanger AE, Hendrix RW, Hatfull GF (1998a) Genome structure of mycobacteriophage D29: implications for phage evolution. J Mol Biol 279(1):143–164 Ford ME, Stenstrom C, Hendrix RW, Hatfull GF (1998b) Mycobacteriophage TM4: genome structure and gene expression. Tuber Lung Dis 79(2):63–73 Hambly E, Suttle CA (2005) The viriosphere, diversity, and genetic exchange within phage communities. Curr Opin Microbiol 8(4):444–450 Hanauer DI, Jacobs-Sera D, Pedulla ML, Cresawn SG, Hendrix RW, Hatfull GF (2006) Inquiry learning. Teaching scientific inquiry. Science 314(5807):1880–1881 Hanauer D, Hatfull GF, Jacobs-Sera D (2009) Active assessment: assessing scientific inquiry. Springer, New York Hatfull GF (2010) Bacteriophage research: gateway to learning science. Microbe 5:243–250 Hatfull GF (2015a) Dark matter of the biosphere: the amazing world of bacteriophage diversity. J Virol 89(16):8107–8110 Hatfull GF (2015b) Innovations in undergraduate science education: going viral. J Virol 89(16):8111–8113 Hatfull GF, Pedulla ML, Jacobs-Sera D, Cichon PM, Foley A, Ford ME, Gonda RM, Houtz JM, Hryckowian AJ, Kelchner VA, Namburi S, Pajcini KV, Popovich MG, Schleicher DT, Simanek BZ, Smith AL, Zdanowicz GM, Kumar V, Peebles CL, Jacobs WR Jr, Lawrence JG, Hendrix RW (2006) Exploring the mycobacteriophage metaproteome: phage genomics as an educational platform. PLoS Genet 2(6):e92 Hatfull GF, Jacobs-Sera D, Lawrence JG, Pope WH, Russell DA, Ko CC, Weber RJ, Patel MC, Germane KL, Edgar RH, Hoyte NN, Bowman CA, Tantoco AT, Paladin EC, Myers MS, Smith AL, Grace MS, Pham TT, O’Brien MB, Vogelsberger AM, Hryckowian AJ, Wynalek JL, Donis-Keller H, Bogel MW, Peebles CL, Cresawn SG, Hendrix RW (2010) Comparative genomic analysis of 60 Mycobacteriophage genomes: genome clustering, gene acquisition, and gene size. J Mol Biol 397(1):119–143 Hendrix RW (1999) The long evolutionary reach of viruses. Curr Biol 9(24):914–917 Hendrix RW (2002) Bacteriophages: evolution of the majority. Theor Popul Biol 61(4):471–480 Hunter A-B, Laursen SL, Seymour E (2007) Becoming a scientist: the role of undergraduate research in students’ cognitive, personal, and professional development. Sci Educ 91(1):36–74 Jacobs WR Jr, Tuckman M, Bloom BR (1987) Introduction of foreign DNA into mycobacteria using a shuttle phasmid. Nature 327(6122):532–535

230

G. F. Hatfull

Jacobs-Sera D, Marinelli LJ, Bowman C, Broussard GW, Guerrero Bustamante C, Boyle MM, Petrova ZO, Dedrick RM, Pope WH, Science Education Alliance Phage Hunters Advancing Genomics and Evolutionary Science Sea-Phages Program, Modlin RL, Hendrix RW, Hatfull GF (2012) On the nature of mycobacteriophage diversity and host preference. Virology 434(2):187–201 Jordan TC, Burnett SH, Carson S, Caruso SM, Clase K, DeJong RJ, Dennehy JJ, Denver DR, Dunbar D, Elgin SC, Findley AM, Gissendanner CR, Golebiewska UP, Guild N, Hartzog GA, Grillo WH, Hollowell GP, Hughes LE, Johnson A, King RA, Lewis LO, Li W, Rosenzweig F, Rubin MR, Saha MS, Sandoz J, Shaffer CD, Taylor B, Temple L, Vazquez E, Ware VC, Barker LP, Bradley KW, Jacobs-Sera D, Pope WH, Russell DA, Cresawn SG, Lopatto D, Bailey CP, Hatfull GF (2014) A broadly implementable research course in phage discovery and genomics for first-year undergraduate students. MBio 5(1):e01051–e01013 Lerma RA, Tidwell TJ, Cahill JL, Rasche ES, Kuty Everett GF (2015) Complete genome sequence of Caulobacter crescentus podophage Percy. Genome Announc 3(6). https://doi.org/10.1128/ genomeA.01373–15. PubMed PMID: 26607888; PubMed Central PMCID: PMCPMC4661307 Lopatto D (2007) Undergraduate research experiences support science career decisions and active learning. CBE Life Sci Educ 6(4):297–306 Mediavilla J, Jain S, Kriakov J, Ford ME, Duda RL, Jacobs WR Jr, Hendrix RW, Hatfull GF (2000) Genome organization and characterization of mycobacteriophage Bxb1. Mol Microbiol 38(5):955–970 Mijalis EM, Lessor LE, Cahill JL, Rasche ES, Kuty Everett GF (2015) Complete genome sequence of Klebsiella pneumoniae carbapenemase-producing K. pneumoniae myophage Miro. Genome Announc 3(5). https://doi.org/10.1128/genomeA.01137–15. PubMed PMID: 26430050; PubMed Central PMCID: PMCPMC4591322 Mokili JL, Rohwer F, Dutilh BE (2012) Metagenomics and future perspectives in virus discovery. Curr Opin Virol 2(1):63–77 Nobrega FL, Costa AR, Kluskens LD, Azeredo J (2015) Revisiting phage therapy: new applications for old resources. Trends Microbiol 23(4):185–191 Pedulla ML, Ford ME, Houtz JM, Karthikeyan T, Wadsworth C, Lewis JA, Jacobs-Sera D, Falbo J, Gross J, Pannunzio NR, Brucker W, Kumar V, Kandasamy J, Keenan L, Bardarov S, Kriakov J, Lawrence JG, Jacobs WR, Hendrix RW, Hatfull GF (2003) Origins of highly mosaic mycobacteriophage genomes. Cell 113(2):171–182 Pope WH, Jacobs-Sera D, Russell DA, Rubin DH, Kajee A, Msibi ZN, Larsen MH, Jacobs WR Jr, Lawrence JG, Hendrix RW, Hatfull GF (2014) Genomics and proteomics of mycobacteriophage patience, an accidental tourist in the Mycobacterium neighborhood. MBio 5(6):e02145 Pope WH, Bowman CA, Russell DA, Jacobs-Sera D, Asai DJ, Cresawn SG, Jacobs WR, Hendrix RW, Lawrence JG, Hatfull GF, Science Education Alliance Phage Hunters Advancing Genomics and Evolutionary Science, Phage Hunters Integrating Research and Education, Mycobacterial Genetics Course (2015) Whole genome comparison of a large collection of mycobacteriophages reveals a continuum of phage genetic diversity. Elife 4:e06416 President’s Council of Advisors on Science and Technology (PCAST) (2012) Engage to excel: producing one million additional college graduates with degrees in science, technology, engineering, and mathematics. Executive Office of the President, Washington, DC Rohwer F, Thurber RV (2009) Viruses manipulate the marine environment. Nature 459(7244):207–212 Rohwer F, Youle M, Maughan H, Hisakawa N (2014) Life in our phage world: a centennial field guide to the earth’s most diverse inhabitants. Wholon, San Diego Seymour E, Hunter A-B, Laursen SL, DeAntoni T (2004) Establishing the benefits of undergraduate research for undergraduates in the sciences: first findings from a three-year study. Sci Educ 88:493–594 Snapper SB, Melton RE, Mustafa S, Kieser T, Jacobs WR Jr (1990) Isolation and characterization of efficient plasmid transformation mutants of Mycobacterium smegmatis. Mol Microbiol 4(11):1911–1919

Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy Michael A. Brockhurst, Britt Koskella, and Quan-Guo Zhang

Contents The Coevolutionary Process and the Red Queen’s Race . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Experimental Studies of Bacteria-Bacteriophage Antagonistic Coevolution . . . . . . . . . . . . . . . . . . Mechanisms of Bacterial Resistance to Lytic Bacteriophage and Phage Counter-Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Effects of Resistance and Counter-Resistance on Fitness and Phenotype . . . . . . . . . . . . . . . . . . . . . (Co)Evolutionary Considerations of Phage Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

232 234 235 240 242 245 245

Abstract

The ubiquity of bacteria-phage interactions across biomes on earth has resulted in a diverse suite of adaptations conferring either bacterial resistance or phage infectivity. Understanding the mechanisms underlying these adaptations has important implications for the use of phages as therapeutic agents, but also offers key insights into how bacterial populations and communities are structured across time and space. In this chapter, we provide, first, an overview of coevolutionary theory relevant to bacteria-phage interactions. Next, we summarize the findings of experimental coevolution studies, focusing on the insights provided into the M. A. Brockhurst (*) Department of Animal and Plant Sciences, University of Sheffield, Sheffield, UK University of York, York, UK e-mail: m.brockhurst@sheffield.ac.uk B. Koskella Department of Integrative Biology, University of California, Berkeley, Berkeley, CA, USA e-mail: [email protected] Q.-G. Zhang College of Life Sciences, Beijing Normal University, Beijing, China e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_7

231

232

M. A. Brockhurst et al.

bacteria-phage coevolutionary processes. Although most experimental studies of bacteria-phage coevolution focus on mutational resistance and counter-resistance, we next survey the variety of resistance and counter-resistance strategies described in nature and consider their implications for bacteria-phage coevolution. We conclude by considering the implications of coevolution for developing phage therapies.

The Coevolutionary Process and the Red Queen’s Race Coevolution, the reciprocal evolution of adaptation and counter-adaptation between ecologically interacting species, is an important evolutionary process, which is believed to play a role, among many other effects, in the emergence and maintenance of diversity, the evolution of parasite virulence, and species extinction (Thompson 1994, 2005). In particular, antagonistic coevolution, whereby adaptation by each species reduces the fitness of the other (as between bacteria and bacteriophages), provides a potent and pervasive source of natural selection thought to be responsible for some of the fastest rates of evolutionary change yet observed (Brockhurst et al. 2014). This idea, that interspecific biotic conflict is a prime driving force of evolution, is best encapsulated in the Red Queen hypothesis, proposed by Van Valen (1973). The eponymous Red Queen refers to the character in Lewis Carroll’s Through the Looking Glass, with whom Alice has a race but soon realizes that despite running as fast as they can, neither of them is moving anywhere. This metaphor neatly captures the hypothesized situation in biological communities: Due to antagonistic ecological interactions, species must continually evolve simply to keep up with their evolving biological enemies, with the result that despite continual evolutionary adaptation, each species’ fitness never improves because any temporary gains are rapidly nullified by counteracting adaptation(s) (Lively 2010; Brockhurst et al. 2014). Correspondingly, much research has attempted to understand the dynamics and outcomes of the coevolutionary races between bacteria and lytic bacteriophages (other phage life-histories, such as temperate and filamentous phages, remain far less studied from a coevolutionary perspective) (Koskella and Brockhurst 2014). Before coming to the empirical data, we first provide a brief overview of the theory underpinning our current view of host-parasite coevolutionary processes. Most models of host-parasite coevolution assume a genetic basis of infection, whereby infection is a product of the interaction between parasite and host alleles at loci encoding infectivity and resistance, respectively. (Although, see Nuismer et al. (2005) and Best et al. (2010) for examples of models where infection is determined by quantitative traits.) The most commonly used models of infection genetics are matching allele (MA) and gene-for-gene (GFG) (Agrawal and Lively 2002), although variants of these (inverse matching allele, inverse gene-forgene) as well as alternatives (e.g., lock and key) exist. Which model of infection genetics best applies to bacteria-phage interactions is a matter of debate, with authors variously arguing the merits of MA, GFG, multilocus-GFG, modified-GFG, IGFG, and relaxed lock and key (Hurst et al. 2008; Fenton et al. 2009; Williams 2013).

Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy

233

However, it is likely that none accurately describes any specific mechanism of bacteria-phage interaction and, moreover, that within many bacteria-phage infection processes, multiple different resistance mechanisms corresponding to different infection genetics operate sequentially or simultaneously (Fenton et al. 2012). In the absence of detailed mechanistic understanding of the appropriate infection genetics, most researchers have instead focused on determining the mode(s) of coevolution operating in bacteria-phage interactions. Thus, we next describe the modes of coevolutionary dynamic that emerge from models of the best-studied infection genetics, MA and GFG. Under MA, parasites cause infection when their infectivity allele matches the host resistance allele, such that each parasite genotype is a specialist that can only infect a single corresponding host genotype (Agrawal and Lively 2002); this mode of infection genetics was derived from interactions between pathogens and invertebrate innate immunity (Luijckx et al. 2013). Under GFG, hosts carry either a Resistant or a Susceptible allele, while parasites carry either a Virulent or an Avirulent allele, Aallele parasites can infect only S-allele hosts whereas V-allele parasites can infect both R- and S-allele hosts; as such GFG allows for the evolution of generalist parasites (i.e., carrying the V-allele) and hosts (i.e., carrying the R-allele), and was derived from plant-pathogen interactions (Thompson and Burdon 1992). The contrasting properties of MA and GFG infection genetics give rise to different selection dynamics in coevolutionary models: MA models typically display timelagged allele frequency oscillations driven by negative frequency-dependent selection (nFDS) (Agrawal and Lively 2002). Here, parasite allele frequencies track host allele frequencies such that at any given time, selection favors rare host alleles, which are less prone to infection by the prevailing, common parasite genotype(s); under these circumstances, high levels of allelic diversity in both hosts and parasites is maintained over time. GFG models typically display directional selection favoring fixation of the generalist host and parasite alleles (R and V, respectively), where coevolution corresponds to an arms race for increasing resistance and infectivity over time (Agrawal and Lively 2002). (However, readers should note that more complex GFG models, incorporating pleiotropic costs of resistance and infectivity, can instead give rise to balanced polymorphism or sustained coevolutionary cycling, similar to that seen in MA models (Sasaki 2000).) Coevolutionary theory therefore suggests two predominant modes of dynamical antagonistic coevolution; these have been named: (i) Fluctuating Selection Dynamics, where host and parasite allele frequencies undergo time-lagged oscillations driven by nFDS; (ii) Arms Race Dynamics, where host and parasite alleles undergo time-lagged selective sweeps driven by directional selection for increased resistance and infectivity, respectively, over time (Gandon et al. 2008). Distinguishing these modes of coevolution in host-parasite populations has been achieved using timeshift experiments (Gandon et al. 2008; Gaba and Ebert 2009). These are crossinfection experiments where hosts (or parasites) from a given point in time are exposed to parasites (or hosts) from past, contemporary, and future time-points. Plots of infectivity or resistance against time-shift, i.e., the difference in time between the host and the parasite sample, can then be used to infer the mode of

234

M. A. Brockhurst et al.

coevolution (Brockhurst and Koskella 2013). Whereas ARD gives rise to monotonic relationships of resistance or infectivity against time-shift, more complex patterns are predicted for FSD depending on the phase of the coevolutionary cycle, including monotonic, V-shaped, and unimodal relationships; as such, it is essential that studies of coevolutionary mode perform time-shift assays for multiple sampling points over time (Gandon et al. 2008).

Experimental Studies of Bacteria-Bacteriophage Antagonistic Coevolution Coculture studies across a range of bacteria-bacteriophage associations reveal the potential for rapid evolutionary responses to phage-imposed selection in bacterial populations, and the potential for evolutionary counter-adaptation in bacteriophage infectivity (Koskella and Brockhurst 2014). Early studies using the classic laboratory model bacterium, Escherichia coli B, and various T-even, T-odd, and λ-vir bacteriophages suggested that while bacteria-bacteriophage coevolution occurred, it was limited to one or two cycles of adaptation and counter-adaptation (Dennehy 2012). To detail one example, coculture of E. coli and T7 in chemostats led to a predictable coevolutionary sequence: The evolution of a resistant bacterial mutants, followed by the evolution of a bacteriophage host-range mutant infecting both the ancestral bacterium and the resistant mutant, followed by the further evolution of bacterial mutants to resist both the ancestral phage and the host-range phage mutant, whereupon coevolution ceased (i.e., 1.5 cycles of coevolution) (Chao et al. 1977). Despite the limited nature of coevolution in this experiment, it nevertheless revealed that the evolution of generalist resistance and generalist infectivity can occur, suggesting therefore that, at least under laboratory conditions, E. coli – T7 coevolution followed the ARD scenario. Similar patterns of limited ARD coevolution have been reported in the other lab studies of E. coli – lytic phage interactions, where resistance and infectivity evolved through spontaneous de novo mutation (Bohannan and Lenski 2000, Dennehy 2012). E. coli B has a long history of adaptation to the lab environment (Daegelen et al. 2009) and is known to harbor defects in cell wall moieties and lipopolysaccharides (Yoon et al. 2012) commonly used by bacteriophages as targets for adsorption (Lenski and Levin 1985). Therefore, its coevolutionary relationship with bacteriophages may not be representative of bacteria-bacteriophage associations less far-removed from the natural environment. Indeed, lab coculture studies of a range of other bacteria-bacteriophage interactions have revealed the potential for dynamical coevolutionary change to be sustained over many 100 s of bacterial generations (reviewed in Koskella and Brockhurst (2014)). The best-studied interaction is that between the plant-associated soil bacterium Pseudomonas fluorescens SBW25 and the T7-like bacteriophage φ2 (Buckling and Rainey 2002; Brockhurst et al. 2007). Over the first approx. 300 bacterial generations, P. fluorescens – φ2 coevolution conforms to ARD: For multiple sampled time-points, bacterial resistance and phage infectivity displayed positive monotonic relationships with time-

Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy

235

shift, consistent with recurrent time-lagged selective sweeps of de novo adaptive mutations driven by directional selection (Brockhurst et al. 2003; Paterson et al. 2010; Hall et al. 2011). By contrast, after approx. 300 bacterial generations, the coevolutionary mode shifts toward FSD: No further increases in bacterial resistance or phage infectivity ranges were observed, indicating a weakening of the response to directional selection, yet coevolution proceeded as sustained oscillations of bacterial resistance and phage infectivity types of approximately equal breadth but different specificity (Hall et al. 2011). The transition, from ARD to FSD, appears to be driven by costs associated with increased resistance and infectivity in this system, which, past some threshold, act to impede further responses to directional selection (Hall et al. 2011). The balance of costs and benefits of resistance/infectivity traits seems to play a key role in determining the dynamics of bacteria-bacteriophage coevolution. Manipulation of environmental parameters can alter this balance, with predictable effects on coevolutionary dynamics. For example, population mixing elevates bacteriaphage encounter rates, which increases the benefit of resistance and thereby accelerates the tempo of coevolution (Brockhurst et al. 2003). Similarly, increasing the supply of resources to P. fluorescens has demographic (increased bacteria-phage population sizes, and therefore higher encounter rates), population genetic (increased mutational supply), and physiological (reduced costs of resistance) effects, which reduce the costs and increase the benefits of resistance, and thereby accelerates the tempo of coevolution (Lopez-Pascua and Buckling 2008). The abiotic environment can also affect the mode of coevolution: P. fluorescens – φ2 coculture experiments in a soil environment demonstrate that coevolution conformed to FSD, with bacteria showing highest resistance against their contemporary phages. This was the case even during the early stages of coevolution, which show ARD in rich liquid media, and was caused by high costs of bacterial resistance in soil environments (Gomez and Buckling 2011).

Mechanisms of Bacterial Resistance to Lytic Bacteriophage and Phage Counter-Resistance The adaptive race between bacteria and their phages has resulted in a striking diversity of mechanisms for infection and resistance. Current understanding of these mechanisms has been reviewed elsewhere (Nechaev and Severinov 2008; Hyman and Abedon 2010; Labrie et al. 2010; Westra et al. 2012; Molineux and Panja 2013; Samson et al. 2013; Young 2013), so here we aim to update and summarize the breadth of this work by highlighting examples of each mechanism and discussing the potential impact each has on the interaction. Bacterial cells can resist phages by blocking phage infection, reproduction, and/or transmission using a variety of resistance mechanisms (Hyman and Abedon 2010). These mechanisms differ from one another in many ways, from the ease at which phages can overcome them (Samson et al. 2013), to the potential fitness costs a host pays for employing them (Bohannan and Lenski 2000), to the breadth of phage resistance

236

M. A. Brockhurst et al.

attained (Hyman and Abedon 2010), and these differences have the potential to influence the way in which bacteria coevolve with different phages in the environment (Jessup and Bohannan 2008; Gomez and Buckling 2011; Hall et al. 2011; Betts et al. 2014). As a first line of defense, bacteria can become resistant to particular phages by preventing phage adsorption through either: the loss or alteration of target receptors; the production of extracellular polysaccharide matrices; or the production of competitive inhibitors that bind to the phage attachment site (reviewed in Hyman and Abedon (2010) and Labrie et al. (2010)). Common phage attachment sites include bacterial pili, flagella, and lipopolysaccharides (LPS) from the outer membrane (Lindberg 1973). Intriguingly, there is also recent evidence that phage attachment might require the presence of multiple receptors for successful attachment (ReyesCortés et al. 2012), and some phages are known to use motility appendages such as flagella and pili to move toward the bacterial cell surface (Bender et al. 1989; Guerrero-Ferreira et al. 2011). Loss or alteration of these key structures in the cell wall are predicted to lead to decreased bacterial fitness, and evidence for such costs of resistance is now substantial. Phage-resistant mutants of Yersinia pestis were found to have decreased production of LPS core biosynthesis enzymes, and this lead to slower growth rates and increased death rate (Filippov et al. 2011). In this case, loss of fitness resulted in attenuated virulence when tested on mice suggesting that phage resistance critically alters the interaction between the pathogenic bacterium and its host. Similarly, modification of a capsular polysaccharide (CPS) resulting in a nonfunctional gene product was found to confer phage resistance in mutant Campylobacter jejuni (Sørensen et al. 2011). Phage resistance of this food-borne pathogen found in the avian gut has also been associated with decreased fitness in the host (Scott et al. 2007). Such attenuated virulence of bacterial pathogens as a result of phage-mediated selection is proving to be a common feature of bacteria-phage systems (Le et al. 2014; Capparelli et al. 2010; Meaden and Koskella 2013). Thus, although receptor loss or modification offers a rapid acquisition of resistance to circulating phages, it is also likely to be among the most costly mechanisms of resistance and therefore may well be lost over time in favor of less costly mechanisms. Alternatively, bacteria are also capable of varying expression of receptors either stochastically or in response to particular stimuli, therefore only paying the cost when risk of infection is high. This “phase variation” can confer resistance of a subset of genetically identical bacterial cells to phage. For example, Vibrio cholerae exhibit intrastrain heterogeneity for expression of the lipopolysaccharide O1 antigen, a known phage receptor (Seed et al. 2012), and variable expression of O antigens in Salmonella enterica has also been associated with phage resistance (Kim and Ryu 2012). Finally, some bacteria are able to block the entry of phage DNA into host cells through the acquisition of prophages (Labrie et al. 2010). These superinfection exclusion mechanisms have been demonstrated for prophages of both Gram-negative (Maillou and Dreiseikelmann 1990) and Gram-positive bacteria (Mahony et al. 2008), and are generally thought to involve phage-encoded proteins that are found in the inner membrane of cells and block the injection of DNA from a subset of other lytic phages.

Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy

237

Like loss or modification of receptors, production of extracellular polysaccharides has long been thought to interfere with phage adsorption, although the effect is not universal (Gu et al. 2011; Roach et al. 2013). For example, phage-resistant mutants of Bacillus anthracis were found to have a mucoid colony phenotype, with increased production of extracellular matrix, as a result of single mutational steps in the csaB gene encoding a cell surface anchoring protein (Bishop-Lilly et al. 2012). However, while the production of exopolysaccharides by the plant pathogen, Erwinia amylovora, has been found to be negatively associated with infection by Myoviridae phages, it seems to be essential for infection and reproduction of Podoviridae phages (Roach et al. 2013), highlighting that this resistance mechanism is not yet clear cut. Finally, there is some evidence for competitive inhibition of phage binding due to the production of particular molecules by the host cell. For example, there is evidence that the siderophore Ferrichrome interferes with binding of phage T5 to its receptor on the surface of Escherichia coli (Luckey et al. 1975). Similarly, coincubation of Staphylococcus aureus with immunoglobulin G from either human and rabbit serum was found to block absorption of phage (Nordström et al. 1974), and host cell-produced protein A was also found to inhibit phage absorption (Nordström and Forsgren 1974). In addition, support for the potential of competitive inhibition as a mechanism for increasing resistance to phages comes from studies in which synthetic peptides designed to block receptors are used to confirm phage binding sites (Killmann et al. 1995). However, this latter mechanism is currently less well supported as a widespread natural mechanism of resistance. Phages are able to counter-adapt to adsorption-blocking resistance mechanisms of their hosts by altering their tail fibers to either recognize newly altered receptors or by switching to entirely new receptors (Hyman and Abedon 2010; Meyer et al. 2012; Munsch-Alatossava and Alatossava 2013), by degrading extracellular polysaccharide matrices (Yan et al. 2014), or by stochastically altering receptor recognition (Samson et al. 2013). For example, analysis of phage genomes after experimental evolution with Pseudomonas fluorescens identified a number of independent mutations in the tail fiber genes required for adsorption (Scanlan et al. 2011). As a more extreme response, experimental evolution of a lytic derivative of phage λ that attaches to the LamB outer membrane protein of E. coli was found to evolve the ability to also use OmpF as a receptor in a glucose-limited environment (Meyer et al. 2012). Furthermore, phages infective to Bordetella spp. have been shown to use diversity-generating retroelements that direct mutagenesis to specific sites in order to vary their recognition of bacterial surface receptors (Doulatov et al. 2004). In response to bacterial production of extracellular polysaccharide matrices, phages are known to produce hydrolyzing enzymes that breakdown the matrix and allow the phage to come into contact with the receptor. For example, phage infection of P. aeruginosa strains isolated from patients with cystic fibrosis was recently shown to involve hydrolysis of the bacterial exopolysaccharide secretion, which resulted in production of clear halos around phage plaques as a result of overproduction of polysaccharide-degrading enzymes leading to bacteria without surrounding capsules (Glonti et al. 2010). Similarly, a number of phages have recently been characterized

238

M. A. Brockhurst et al.

that are capable of degrading biofilms produced by Escherichia coli strains associated with urinary tract infections (Chibeu et al. 2012). If the phage is able to successfully adsorb to its bacterial host and inject its own genetic material into the host cell, there are still a number of possible routes to resistance for the cell; phage DNA entering the host cell can be degraded by hostencoded enzymes and/or replication of DNA can be blocked by the host directly. First, a number of restriction-modification systems have been found to be widespread across bacterial taxa. These systems act to recognize and degrade unmethylated DNA in the cell, including that of phages, through production of restriction enzymes (Wilson and Murray 1991). Restriction endonucleases are generally effective against DNA phages as well as plasmids, but phages that are successfully methylated within the cell, thus evading degradation, are not only able to reproduce but also produce progenies that can evade recognition by future host cells with the same restriction-modification system. Although restrictionmodification systems are often considered quite specific, recent work has demonstrated that bacteria harboring highly promiscuous restriction-modification systems are better able to deal with phage anti-restriction mechanisms than those carrying high-fidelity systems (Vasu et al. 2012). In addition, there is evidence that restrictionmodification-mediated resistance may be temperature dependent; phage resistance of Listeria monocytogenes across varying temperatures was found to be due to differential expression of the restriction endonuclease (Kim et al. 2012). Phages can overcome restriction-modification systems by accumulating mutations at endonuclease recognition sites (Krüger et al. 1987), by acquiring their own methylase genes (Hill et al. 1991, McGrath et al. 1999), by switching to more unusual nucleic acids (Krüger and Bickle 1983), or by modifying the host’s restriction-modification system directly (King and Murray 1995). In addition to restriction-modification systems, the majority of bacterial and archaeal genomes examined to date contain arrays of clustered, regularly interspaced short palindromic repeats (CRISPRs) which have been linked to, among other functions, resistance against phages (Barrangou et al. 2007; Bondy-Denomy and Davidson 2014). In contrast to the restriction-modification systems, the CRISPR-Cas system is highly specific to infecting phages, as the bacterial host must match a short segment of the phage genome in order to block replication. In order to achieve such specific recognition, bacteria incorporate small segments of foreign DNA into their own genomes (referred to as “spacers”), and these sequences are used to generate RNA that is carried by the ribonucleoprotein Cas complex to cleave the matching phage DNA (Bhaya et al. 2011; Westra et al. 2012). The acquisition of spacers is known to be highly biased away from sequences that match the host DNA and also toward specific phage genome locations (Paez-Espino et al. 2013). Interestingly, there is building evidence that the presence of “old” spacers (i.e., those which no longer directly match the foreign genetic elements in the local environment) might “prime” the CRISPR-cas system for more rapid acquisition of new spacers in response to mutations in the DNA targets (the protospacers), which could help explain the rapid acquired immunity observed (Fineran et al. 2014). As with each of the previous resistance mechanisms discussed, there is building evidence that coevolving phages

Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy

239

can rapidly escape recognition by the CRISPR-Cas system (Samson et al. 2013). A number of “anti-CRISPR” genes have already been found within the genomes of Pseudomonas aeruginosa phages (Bondy-Denomy et al. 2012), and some phages have even been found to carry their own CRISPR-cas system to target a chromosomal island of the bacterial host (Seed et al. 2013). Furthermore, experimentally evolved phages infecting Streptococcus thermophiles hosts were found to rapidly escape CRISPR-Cas based immunity via mutations in the proto-spacer adjacent motif (PAM) (Sun et al. 2013). Finally, there is recent evidence that the CRISPRCas and restriction-modification systems can work synergistically to increase bacterial resistance to phages, perhaps going some way to explain the ubiquity of the two systems across the bacterial tree of life (Dupuis et al. 2013). A final known mechanism of bacterial resistance is the abortive infection (Abi) system. This mechanism of cell suicide, encoded by a toxin-antitoxin system, has been shown to protect bacterial hosts against infection by multiple phages and seems to be relatively difficult for phages to circumvent (Fineran et al. 2009). Most Abi systems appear to be plasmid-encoded (Chopin et al. 2005), suggesting that they can move readily among bacterial species. Furthermore, despite the great diversity of Abi systems uncovered to date, there do seem to be some common features. In most cases, phage infection leads to the activation of dormant enzymes, and this activation results in cleavage of highly conserved and essential components of the cellular translational apparatus (which itself is known to be hijacked by RNA viruses for replication; Bushell and Sarnow (2002)). This response has recently been demonstrated in Escherichia coli where it was found to be a low cost and effective strategy at the population level (Refardt et al. 2013) that is particularly favored in spatially structured environments where protection is conferred primarily to highly related neighbors (Berngruber et al. 2013). Not surprisingly, there is evidence that phages are able to adapt in response to Abi systems. Isolation of mutant lytic phages that showed resistance against one or two Abi systems, AbiK and/or AbiT, in Lactococcus lactis were found to have acquired this resistance via extensive homologous and nonhomologous recombination with prophages within the bacterial genome, exchanging as much as 79% of their genomes (Labrie and Moineau 2007). More recently, evidence for phage production of a “pseudotoxin” that mimics ToxI in order to suppress ToxN has been uncovered (Blower et al. 2012). Abi escape mutants of the lytic phage ΦTE, infecting Pectobacterium atrosepticum, were found to have acquired a noncoding RNA sequence with expanded repeats that allowed the phage to replicate in the presence of ToxIN. Finally, comparative genomic approaches to explore bacterial and archaeal defense systems has uncovered an intriguing hypothesis: that the immunity systems whereby bacteria target and either degrade or block replication of phage DNA are functionally coupled to systems that lead to cell death or dormancy upon phage infection (Makarova et al. 2013). Under this model, the latter defense pathway would only be triggered upon failure of the first, thereby both minimizing the probability of cell death and maximizing protection at the bacterial population level. The variety of resistance and infectivity mechanisms observed so far represents the result of an ancient, obligate association and ongoing coevolution. Although

240

M. A. Brockhurst et al.

the exploration of how these mechanisms might differentially affect the outcome of coevolution is in its infancy (Hall et al. 2011; Betts et al. 2014), the data thus far provide intriguing support for the translation of mechanistic understanding into predictive power. One aspect of bacterial resistance and phage counter-adaptation that warrants further research is in regard to how such mechanisms might vary across environments and time, regardless of bacterial genetics. Many bacterial resistance mechanisms are remarkably labile and can often be regulated by reversible switching of phenotypic expression in response to environmental cues (reviewed in Hoskisson and Smith (2007)), and this variation will have key implications for the coevolutionary trajectory of the two players. The known resistance mechanisms are of course not mutually exclusive, and there is now good reason to think that multiple mechanisms will be acting simultaneously in a population, even within the same pairwise interaction. Moving forward, it will also become important to determine when and how these defenses interact to protect bacterial populations and communities against circulating phages (Tables 1 and 2).

Effects of Resistance and Counter-Resistance on Fitness and Phenotype Since phage receptors are often components of the bacterial cell membrane like LPS or membrane-associated proteins like transporters, bacterial evolution of phage resistance often has deleterious side-effects. As such, bacterial resistance mutations often reduce bacterial fitness in the absence of phages through antagonistic pleiotropy, wherein the molecular changes that cause resistance by, for instance, altering a phage-binding site impair the normal biological function of the molecule. Furthermore, the costs of multiple resistance mutations in an individual bacterial host may interact through epistasis. Epistasis may constrain resistance evolution if the combined costs are greater than the sum of the costs of each individual mutation (negative epistasis), or conversely could promote resistance evolution if the cost of each subsequent mutation is less than additive (positive epistasis). Examples of both forms of epistasis have been observed in bacteria-phage experiments (Bohannan et al. 1999; Buckling et al. 2006; Koskella et al. 2012). In addition to the direct effects of resistance mutations, the evolution of resistance can affect the bacterial phenotype due to linkage between the primary resistance mutation and coincident mutations at other loci in the genome. Here, selective sweeps of resistance mutations could also lead to fixation of other mutations by hitchhiking. This has been observed in populations of P. fluorescens wherein mutations causing colony morphology variants can hitchhike with resistance, leading to phenotypic divergence between populations due to the chance associations between linked mutations (Buckling and Rainey 2003, Brockhurst et al. 2004). On the other side of the coin, the consequence of coevolution for phage fitness and phenotypic traits has been less well studied. However, this is particularly relevant under the ARD coevolution scenario: fitness cost of infectivity is a most parsimonious explanation for the absence of phage phenotypes with “universal

Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy

241

Table 1 Mechanisms of bacterial resistance to phage, with specific examples from wellcharacterized systems Mechanism Modification or loss of phage receptor

Details Reduces phage recognition of host cell and adsorption

Phase variation in expression of phage receptors Production of polysaccharide matrix

Varied expression of receptor confers resistance to subset of cells Reduces phage adsorption by blocking access to cell surface Production of peptide which binds to phage receptor

Competitive inhibition of phage receptor

Restrictionmodification system CRISPR-Cas system Abortive infection (Abi) system

Recognition and restriction of foreign DNA by REase Blocks phage transcription once inside the cell Cell apoptosis upon infection by phage

Species Prochlorococcus cyanobacteria, Staphylococcus aureus, Yersinia pestis, Escherichia coli, Pseudomonas aeruginosa, Campylobacter jejuni Salmonella enterica, Bordetella spp., Vibrio cholera, Campylobacter jejuni Bacillus anthracis, Erwinia amylovora

Escherichia coli, Staphylococcus aureus

Klebsiella pneumonia, Haemophilus influenzae

Widespread

Lactococcus lactis, Erwinia carotovora, Escherichia coli

Reference (Le et al. 2014; Killmann and Braun 1992; Capparelli et al. 2010; Avrani et al. 2011; Filippov et al. 2011; Sørensen et al. 2011)

(Liu et al. 2002; Kim and Ryu 2012; Seed et al. 2012; Sørensen et al. 2012) (Bishop-Lilly et al. 2012; Roach et al. 2013)

(Nordström and Forsgren 1974; Nordström et al. 1974; Luckey et al. 1975; Destoumieux-Garzón et al. 2005) (Zaleski et al. 2005; Vasu et al. 2012)

(Bhaya et al. 2011; BondyDenomy and Davidson 2014) (Parma et al. 1992; Chopin et al. 2005; Fineran et al. 2009)

infectivity” (one that can infect all host genotypes) (Agrawal and Lively 2002). Empirical evidence has mainly come from two model systems. In the E. coli – T7 system, the more infective phage phenotype (T71) often reaches a lower population size than the ancestral phenotype (T70) (Forde et al. 2007). In the P. fluorescens – φ2 system, phage phenotypes with broader infectivity ranges have lower growth rate than those with narrow infectivity ranges when infecting the ancestral, susceptible, bacterial genotype (Poullain et al. 2008); and fixation of phage phenotypes that form tiny plaques on bacterial lawn were occasionally observed in phage populations coevolving with bacteria but not in those always infecting the ancestral, susceptible, bacterial phenotype (Q.-G. Zhang, personal observations). The evolution of infectivity may also show pleiotropic effect on phage traits. In recent work with the P. fluorescens – φ2 system, phages with broader infectivity showed higher sensitivity to temperature elevation, with a negative consequence for the population-level

242

M. A. Brockhurst et al.

Table 2 Mechanisms of phage infectivity against previously resistant bacteria, with examples from well-characterized systems. For a more detailed review, see Samson et al. (2013) Bacterial resistance mechanism Modification of phage receptor Modification of phage receptor Loss of phage receptor Production of extracellular polysaccharide matrix Restrictionmodification system CRISPR-Cas system CRISPR-Cas system CRISPR-Cas system Abortive infection (Abi) system Abortive infection (Abi) system

Phage infectivity mechanism Mutations in tail fiber genes Diversity-generating retroelements (DGRs) Switch to novel receptor

Bacterial host species Pseudomonas fluorescens Bordetella spp. Escherichia coli

Reference (Paterson et al. 2010; Scanlan et al. 2011) (Doulatov et al. 2004) (Meyer et al. 2012)

Production of hydrolyzing enzymes to degrade/depolymerize matrix Genetic exchange of host methylase gene

Streptococcus pyogenes, Escherichia coli, Pseudomonas aeruginosa Lactococcus lactis

(Baker et al. 2002; Glonti et al. 2010; Chibeu et al. 2012)

Mutations in protospacer adjacent motif (PAM) Inactivation of CRISPRcas system Phage-encoded CRISPR-Cas system to evade host response Recombination with prophage within bacterial genome Mimicry of the RNA antitoxin, ToxI

Streptococcus thermophiles

(Paez-Espino et al. 2013)

Pseudomonas aeruginosa Vibrio cholerae

(Bondy-Denomy et al. 2012) (Seed et al. 2013)

Lactococcus lactis

(Labrie and Moineau 2007)

Pectobacterium atrosepticum

(Blower et al. 2012)

(Hill et al. 1991).

adaptation: phages coevolved with bacteria were less able to evolutionarily adapt to higher temperature compared with those which always infected the ancestral, susceptible, bacterial phenotype (Zhang and Buckling 2011). This supports a notion that reciprocal coevolutionary adaptation may come at the expense of adaptation to the physical environments (Thompson 2013), and highlight the importance to consider coevolutionary dynamics when studying evolutionary rescue effects.

(Co)Evolutionary Considerations of Phage Therapy Therapeutic uses of phages for treating bacterial infections have received a resurgent attention in the past two decades, due to the challenge from the unprecedented level of bacterial antibiotic resistance. As bacteria may readily evolve resistance to phages too, a naturally arising question is whether or not we will also be faced with a

Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy

243

“tragedy of the commons” in phage applications: overuse of phages results in the spread of phage-resistant bacteria, leaving us with no weapon to use to combat bacterial pathogens (Meaden and Koskella 2013). Parallel with the debate about single-drug versus multiple-drug uses, there are different opinions in the design of phage treatments: the conventional wisdom is to use cocktails of different types of phages infecting the same species or strains to reduce the chance of emergence of resistant bacteria (Gu et al. 2012); and the opposite opinion is to use only a single type of phage to treat a specific infection in order to prevent the emergence of bacteria that are resistant to broad-host range cocktails of phages (Krylov et al. 2012). One positive aspect with phage therapy is that phages can evolve novel infectivity phenotypes at a speed much faster than humans can develop new antibiotic drugs. Considering what we have learned about bacteria-phage coevolution, it is reasonable to be more optimistic about phage therapy than antibiotic uses. For either antibiotic or phage therapy, the emergence of resistant bacteria is unlikely to be preventable (Levin and Bull 2004). But the spread and persistence of bacterial resistance in natural environments matter a lot. The cost of antibiotic resistance in bacteria is often relatively small and may often be compensated for by further compensatory evolution, leading to long-term persistence of such resistance in nature (Andersson and Hughes 2011). The situation with bacterial resistance to phages is likely to be different; bacteria have been continuously attacked by phages during their evolutionary history (in most habitats the virus-to-bacterium ratio is over 10; Srinivasiah et al. (2008)), but such strong selection did not result in fixation of broadly phageresistant bacteria: biogeography studies of phage infection patterns show that phages often be infective to bacterial host cells sampled from distant locations (Clokie et al. 2011). Experimental evolution work with conventional lab environment (nutrientrich liquid medium) suggests that broadly phage-resistant bacteria can emerge, and such broad resistance may not necessarily confer a large fitness cost (Buckling and Rainey 2002; Forde et al. 2008b; Koskella et al. 2012). However, when such evolution experiments were done in more natural environments such as soil or tree leaves (Koskella 2014), bacteria-phage coevolution is more likely to follow an FSD pattern, and the bacteria do not maintain broad resistance to phages (Gomez and Buckling 2011). In these more naturalistic environments, it may be very costly for the bacteria to maintain resistance to phage types no longer present in their local environments. This suggests that even the broadly phage-resistant bacteria do emerge in clinic settings, their spread and persistence in natural environments may be limited. Meanwhile, the evolution of phages against bacteria does not guarantee desirable therapeutic effectiveness. What we hope phages do in therapeutic applications (driving bacteria to very low density or even extinct) is often not what phages manage to do in nature (Wommack and Colwell 2000; Danovaro et al. 2011). Appropriate human interventions are often necessary for better therapeutic practice. In fact, we may make use of experimental evolution protocols for producing more effective phage preparations. First, many phages isolated from natural environments do not kill bacteria very efficiently, e.g., due to low growth rate but not a lack of

244

M. A. Brockhurst et al.

infectivity, because selection for other traits such as lower decay rates may have limited the evolution of high virulence of phages in nature. Here, simple evolutionary training, repeated serial passages on susceptible bacterial types in the lab, may significantly increase the phage virulence (Betts et al. 2013). Second, under the FSD scenario, every phage type and the bacterial type susceptible to it undergo timelagged oscillations in frequency and population size, leading to limited effect of phages on the total population size of bacteria due to temporal mismatches. In this case, repeated introduction of cocktails of phages infective against different bacterial types may achieve much better control of bacterial density. Third, we may adopt an experimental approach for “breeding” very broadly infective phages (and unavoidably broadly resistant bacteria) that may have not emerged or not widely persisted in nature, and use them for therapeutic use. Of course cautions are needed here to deal with the broadly resistant bacteria yielded in such breeding work, although as mentioned above, such broadly phage-resistant bacteria are unlikely to persist in nature. Previous experimental work in lab environments have provided several clues for ideal culture conditions for such breeding work: (i) large culture volumes, as larger population sizes increase the rate of supply of novel mutations; (ii) nutrientrich medium that promotes ARD-like coevolution by increasing phage population sizes as well as decreasing the fitness cost of broad resistance in bacteria (Forde et al. 2008a; Lopez-Pascua and Buckling 2008; Lopez Pascua et al. 2014); (iii) population mixing that increases the rate of encounter between phages and susceptible bacteria and thus leads to stronger selection for bacterial defense which in turn increases the strength of selection for phage infectivity (Brockhurst et al. 2003); and (iv) repeated immigration of susceptible bacteria at an appropriate migration rate; here the bacterial immigration needs to be strong enough to increase the population sizes of the phages but not so strong to select against the evolution of broad infectivity ranges in the phages (Benmayor et al. 2009). An important difference between phage treatment of bacteria and biocontrol of higher organisms (such as weeds or insects) is that phage uses can be combined with chemical treatments. Phages are usually insensitive to antibiotics, a recent study with the P. fluorescens – φ2 system showed that simultaneous use of the phage and an antibiotic (kanamycin) greatly reduce the chance of bacterial survival that requires resistance evolution; and furthermore, in the very rare cases where the bacteria survived the combined treatments, the resistant bacteria had very low fitness (Zhang and Buckling 2012). In another study with the same bacterium-phage system, under simultaneous treatments of the phage and an antibiotic (rifampicin), bacteria evolved to be resistant and survived, but later some biofilm-forming bacterial genotypes in the populations showed some levels of reversion to antibiotic susceptibility; but this did not happen in the populations treated with the antibiotic only. Presumably this biofilm-forming genotype can, through forming a biofilm on the surface of liquid medium, protect a portion of its cells from antibiotic stress. Moreover, because antibiotic resistance had been very costly when the populations were also exposed to phages, this fitness cost drove the resistance reversion at the population level (Escobar-Páramo et al. 2012). However, the fitness cost conferred by resistance to phages may not generally augment that of resistance to antibiotics.

Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy

245

In a more recent study also with P. fluorescens – φ2, bacteria were exposed to phage treatment first, and then evolved on concentration gradients of three single antibiotics (cefotaxime, chloramphenicol, and kanamycin), with migration along the gradients (mimicking the situations where bacteria evolve in more realistic heterogeneous drug environment), additive or antagonistic, but not synergistic, interaction between fitness costs of resistance to the phage and those of resistance to antibiotics were observed (Zhang 2014). It is possible that in more realistic environments, migration of bacteria from low-drug environments can promote compensatory evolution that reduces the fitness cost of antibiotic resistance in the high-drug environments. Therefore, phage therapy does provide a very promising alternative to antibiotic uses, particularly because the evolution of bacterial resistance to phages might be a less serious problem compared with bacterial antibiotic resistance. Studies of bacteria-phages coevolution can provide much valuable knowledge for improving the practice of phage therapy.

Cross-References ▶ Bacteriophage as Biocontrol Agents ▶ Bacteriophage Ecology ▶ Genetics and Genomics of Bacteriophages ▶ Phage Infection and Lysis

References Agrawal A, Lively CM (2002) Infection genetics: gene-for-gene versus matching-alleles models and all points in between. Evol Ecol Res 4:79–90 Andersson DI, Hughes D (2011) Persistence of antibiotic resistance in bacterial populations. FEMS Microbiol Rev 35:901–911 Avrani S, Wurtzel O, Sharon I, Sorek R, Lindell D (2011) Genomic island variability facilitates Prochlorococcus-virus coexistence. Nature 474:604–608 Baker J, Dong S, Pritchard D (2002) The hyaluronan lyase of Streptococcus pyogenes bacteriophage H4489A. Biochem J 365:317–322 Barrangou R, Fremaux C, Deveau H, Richards M, Boyaval P, Moineau S, Romero DA, Horvath P (2007) CRISPR provides acquired resistance against viruses in prokaryotes. Science 315:1709–1712 Bender RA, Refson CM, O'Neill EA (1989) Role of the flagellum in cell-cycle-dependent expression of bacteriophage receptor activity in Caulobacter crescentus. J Bacteriol 171:1035–1040 Benmayor R, Hodgson DJ, Perron GG, Buckling A (2009) Host mixing and disease emergence. Curr Biol 19:764–767 Berngruber TW, Lion S, Gandon S (2013) Evolution of suicide as a defence strategy against pathogens in a spatially structured environment. Ecol Lett 16:446–453 Best A, White A, Kisdi E, Antonovics J, Brockhurst MA, Boots M (2010) The evolution of hostparasite range. Am Nat 176:63–71

246

M. A. Brockhurst et al.

Betts A, Vasse M, Kaltz O, Hochberg ME (2013) Back to the future: evolving bacteriophages to increase their effectiveness against the pathogen Pseudomonas aeruginosa PAO1. Evol Appl 6:1054–1063 Betts A, Kaltz O, Hochberg ME (2014) Contrasted coevolutionary dynamics between a bacterial pathogen and its bacteriophages. Proc Natl Acad Sci USA 111:11109–11114 Bhaya D, Davison M, Barrangou R (2011) CRISPR-Cas Systems in Bacteria and Archaea: versatile small RNAs for adaptive defense and regulation. Ann Rev Genet 45:273–297 Bishop-Lilly K, Plaut R, Chen P, Akmal A, Willner K, Butani A, Dorsey S, Mokashi V, Mateczun A, Chapman C, George M, Luu T, Read T, Calendar R, Stibitz S, Sozhamannan S (2012) Whole genome sequencing of phage resistant Bacillus anthracis mutants reveals an essential role for cell surface anchoring protein CsaB in phage AP50c adsorption. Virol J 9:246 Blower TR, Evans TJ, Przybilski R, Fineran PC, Salmond GP (2012) Viral evasion of a bacterial suicide system by RNA-based molecular mimicry enables infectious altruism. PLoS Genet 8:e1003023 Bohannan BJM, Lenski RE (2000) Linking genetic change to community evolution: insights from studies of bacteria and bacteriophage. Ecol Lett 3:362–377 Bohannan BJM, Travisano M, Lenski RE (1999) Epistatic interactions can lower the cost of resistance to multiple consumers. Evolution 53:292–295 Bondy-Denomy J, Davidson AR (2014) To acquire or resist: the complex biological effects of CRISPR–Cas systems. Trends Microbiol 22:218–225 Bondy-Denomy J, Pawluk A, Maxwell KL, Davidson AR (2012) Bacteriophage genes that inactivate the CRISPR/Cas bacterial immune system. Nature 493:429–432 Brockhurst MA, Koskella B (2013) Experimental coevolution of species interactions. Trends Ecol Evol 28(6):367–375 Brockhurst MA, Morgan AD, Rainey PB, Buckling A (2003) Population mixing accelerates coevolution. Ecol Lett 6:975–979 Brockhurst MA, Rainey PB, Buckling A (2004) The effect of spatial heterogeneity and parasites on the evolution of host diversity. Proc R Soc B Biol Sci 271:107–111 Brockhurst MA, Morgan AD, Fenton A, Buckling A (2007) Experimental coevolution with bacteria and phage the Pseudomonas fluorescens – phi 2 model system. Infect Genet Evol 7:547–552 Brockhurst MA, Chapman T, King KC, Mank JE, Paterson S, Hurst GD (2014) Running with the red queen: the role of biotic conflicts in evolution. Proc R Soc B Biol Sci 281:20141382 Buckling A, Rainey PB (2002) Antagonistic coevolution between a bacterium and a bacteriophage. Proc R Soc B Biol Sci 269:931–936 Buckling A, Rainey PB (2003) The role of parasites in sympatric and allopatric host diversification (vol 420, pg 496, 2002). Nature 421:294–294 Buckling A, Wei Y, Massey RC, Brockhurst MA, Hochberg ME (2006) Antagonistic coevolution with parasites increases the cost of host deleterious mutations. Proc R Soc B Biol Sci 273:45–49 Bushell M, Sarnow P (2002) Hijacking the translation apparatus by RNA viruses. J Cell Biol 158:395–399 Capparelli R, Nocerino N, Lanzetta R, Silipo A, Amoresano A, Giangrande C, Becker K, Blaiotta G, Evidente A, Cimmino A, Iannaccone M, Parlato M, Medaglia C, Roperto S, Roperto F, Ramunno L, Iannelli D (2010) Bacteriophage-resistant Staphylococcus aureus mutant confers broad immunity against staphylococcal infection in mice. PLoS One 5:e11720 Chao L, Levin BR, Stewart FM (1977) Complex Community in a Simple Habitat – experimentalstudy with bacteria and phage. Ecology 58:369–378 Chibeu A, Lingohr EJ, Masson L, Manges A, Harel J, Ackermann H-W, Kropinski AM, Boerlin P (2012) Bacteriophages with the ability to degrade uropathogenic Escherichia coli biofilms. Viruses 4:471–487 Chopin M-C, Chopin A, Bidnenko E (2005) Phage abortive infection in lactococci: variations on a theme. Curr Opin Microbiol 8:473–479 Clokie MRJ, Millard AD, Letarov AV, Heaphy S (2011) Phages in nature. Bacteriophage 1:31–45

Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy

247

Daegelen P, Studier FW, Lenski RE, Cure S, Kim JF (2009) Tracing ancestors and relatives of Escherichia coli B, and the derivation of B strains REL606 and BL21(DE3). J Mol Biol 394:634–643 Danovaro R, Corinaldesi C, Dell'Anno A, Fuhrman JA, Middelburg JJ, Noble RT, Suttle CA (2011) Marine viruses and global climate change. FEMS Microbiol Rev 35:993–1034 Dennehy JJ (2012) What can phages tell us about host-pathogen coevolution? Int J Evol Biol 2012:396165 Destoumieux-Garzón D, Duquesne S, Peduzzi J, Goulard C, Desmadril M, Letellier L, Rebuffat S, Boulanger P (2005) The iron-siderophore transporter FhuA is the receptor for the antimicrobial peptide microcin J25: role of the microcin Val11-Pro16 beta-hairpin region in the recognition mechanism. Biochem J 389:869–876 Doulatov S, Hodes A, Dai L, Mandhana N, Liu M, Deora R, Simons RW, Zimmerly S, Miller JF (2004) Tropism switching in Bordetella bacteriophage defines a family of diversity-generating retroelements. Nature 431:476–481 Dupuis M-È, Villion M, Magadán AH, Moineau S (2013) CRISPR-Cas and restriction–modification systems are compatible and increase phage resistance. Nat Commun 4:2087 Escobar-Páramo P, Gougat-Barbera C, Hochberg ME (2012) Evolutionary dynamics of separate and combined exposure of Pseudomonas fluorescens SBW25 to antibiotics and bacteriophage. Evol Appl 5:583–592 Fenton A, Antonovics J, Brockhurst MA (2009) Inverse-gene-for-gene infection genetics and Coevolutionary dynamics. Am Nat 174:E230–E242 Fenton A, Antonovics J, Brockhurst MA (2012) Two-step infection processes can lead to coevolution between functionally independent infection and resistance pathways. Evolution 66:2030–2041 Filippov AA, Sergueev KV, He Y, Huang X-Z, Gnade BT, Mueller AJ, Fernandez-Prada CM, Nikolich MP (2011) Bacteriophage-resistant mutants in Yersinia pestis: identification of phage receptors and attenuation for mice. PLoS One 6:e25486 Fineran PC, Blower TR, Foulds IJ, Humphreys DP, Lilley KS, Salmond GPC (2009) The phage abortive infection system, ToxIN, functions as a protein–RNA toxin–antitoxin pair. Proc Natl Acad Sci USA 106:894–899 Fineran PC, Gerritzen MJH, Suárez-Diez M, Künne T, Boekhorst J, van Hijum SAFT, Staals RHJ, Brouns SJJ (2014) Degenerate target sites mediate rapid primed CRISPR adaptation. Proc Natl Acad Sci USA 111:E1629–E1638 Forde S, Thompson J, Bohannan BM (2007) Gene flow reverses an adaptive cline in a coevolving host-parasitoid interaction. Am Nat 169:794–801 Forde SE, Beardmore RE, Gudelj I, Arkin SS, Thompson JN, Hurst LD (2008a) Understanding the limits to generalizability of experimental evolutionary models. Nature 455:220–223 Forde SE, Thompson JN, Holt RD, Bohannan BJM (2008b) Coevolution drives temporal changes in fitness and diversity across environments in a bacteria-bacteriophage interaction. Evolution 62:1830–1839 Gaba S, Ebert D (2009) Time-shift experiments as a tool to study antagonistic coevolution. Trends Ecol Evol 24:226–232 Gandon S, Buckling A, Decaestecker E, Day T (2008) Host-parasite coevolution and patterns of adaptation across time and space. J Evol Biol 21:1861–1866 Glonti T, Chanishvili N, Taylor P (2010) Bacteriophage-derived enzyme that depolymerizes the alginic acid capsule associated with cystic fibrosis isolates of Pseudomonas Aeruginosa. J Appl Microbiol 108:695–702 Gomez P, Buckling A (2011) Bacteria-phage antagonistic coevolution in soil. Science 332:106–109 Gu E, Nguyen D, Shah N (2011) Capsular polysaccharide has a minor role on streptomycin-induced reduction of T7 phage adsorption to Escherichia coli. J Exp Microbiol Immunol (JEMI) 15:47–51

248

M. A. Brockhurst et al.

Gu J, Liu X, Li Y, Han W, Lei L, Yang Y, Zhao H, Gao Y, Song J, Lu R, Sun C, Feng X (2012) A method for generation phage cocktail with great therapeutic potential. PLoS One 7:e31698 Guerrero-Ferreira RC, Viollier PH, Ely B, Poindexter JS, Georgieva M, Jensen GJ, Wright ER (2011) Alternative mechanism for bacteriophage adsorption to the motile bacterium Caulobacter crescentus. Proc Natl Acad Sci USA 108:9963–9968 Hall AR, Scanlan PD, Morgan AD, Buckling A (2011) Host-parasite coevolutionary arms races give way to fluctuating selection. Ecol Lett 14:635–642 Hill C, Miller L, Klaenhammer T (1991) In vivo genetic exchange of a functional domain from a type II A methylase between lactococcal plasmid pTR2030 and a virulent bacteriophage. J Bacteriol 173:4363–4370 Hoskisson PA, Smith MCM (2007) Hypervariation and phase variation in the bacteriophage ‘resistome’. Curr Opin Microbiol 10:396–400 Hurst LD, Forde SE, Beardmore RE, Gudelj I, Arkin SS, Thompson JN (2008) Understanding the limits to generalizability of experimental evolutionary models. Nature 455:220–U244 Hyman P, Abedon ST (2010) Bacteriophage host range and bacterial resistance. Adv Appl Microbiol 70:217–248 Jessup CM, Bohannan BJ (2008) The shape of an ecological trade-off varies with environment. Ecol Lett 11:947–959 Killmann H, Braun V (1992) An aspartate deletion mutation defines a binding site of the multifunctional FhuA outer membrane receptor of Escherichia coli K-12. J Bacteriol 174:3479–3486 Killmann H, Videnov G, Jung G, Schwarz H, Braun V (1995) Identification of receptor binding sites by competitive peptide mapping: phages T1, T5, and phi 80 and colicin M bind to the gating loop of FhuA. J Bacteriol 177:694–698 Kim M, Ryu S (2012) Spontaneous and transient defence against bacteriophage by phase-variable glucosylation of O-antigen in Salmonella enterica serovar Typhimurium. Mol Microbiol 86:411–425 Kim J-W, Dutta V, Elhanafi D, Lee S, Osborne JA, Kathariou S (2012) A novel restrictionmodification system is responsible for temperature-dependent phage resistance in Listeria monocytogenes ECII. Appl Environ Microbiol 78:1995–2004 King G, Murray NE (1995) Restriction alleviation and modification enhancement by the Rac prophage of Escherichia coli K-12. Mol Microbiol 16:769–777 Koskella B (2014) Bacteria-phage interactions across time and space: merging local adaptation and time-shift experiments to understand phage evolution. Am Nat 184:S9–S21 Koskella B, Brockhurst MA (2014) Bacteria-phage coevolution as a driver of ecological and evolutionary processes in microbial communities. FEMS Microbiol Rev 38:916–931 Koskella B, Lin DM, Buckling A, Thompson JN (2012) The costs of evolving resistance in heterogeneous parasite environments. Proc R Soc B Biol Sci 279:1896–1903 Krüger D, Bickle TA (1983) Bacteriophage survival: multiple mechanisms for avoiding the deoxyribonucleic acid restriction systems of their hosts. Microbiol Rev 47:345 Krüger D, Barcak G, Smith H (1987) Abolition of DNA recognition site resistance to the restriction endonuclease EcoRII. Biomed Biochim Acta 47:K1–K5 Krylov V, Shaburova O, Krylov S, Pleteneva E (2012) A genetic approach to the development of new therapeutic phages to fight Pseudomonas aeruginosa in wound infections. Viruses 5:15–53 Labrie SJ, Moineau S (2007) Abortive infection mechanisms and prophage sequences significantly influence the genetic makeup of emerging lytic Lactococcal phages. J Bacteriol 189:1482–1487 Labrie SJ, Samson JE, Moineau S (2010) Bacteriophage resistance mechanisms. Nat Rev Microbiol 8:317–327 Le S, Yao X, Lu S, Tan Y, Rao X, Li M, Jin X, Wang J, Zhao Y, Wu NC et al (2014) Chromosomal DNA deletion confers phage resistance to Pseudomonas aeruginosa. Sci Rep 4:4738. https:// doi.org/10.1038/srep04738 Lenski RE, Levin BR (1985) Constraints on the coevolution of bacteria and virulent phage – a model, some experiments, and predictions for natural communities. Am Nat 125:585–602

Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy

249

Levin BR, Bull JJ (2004) Population and evolutionary dynamics of phage therapy. Nat Rev Microbiol 2:166–173 Lindberg AA (1973) Bacteriophage receptors. Ann Rev Microbiol 27:205–241 Liu M, Deora R, Doulatov SR, Gingery M, Eiserling FA, Preston A, Maskell DJ, Simons RW, Cotter PA, Parkhill J (2002) Reverse transcriptase-mediated tropism switching in Bordetella bacteriophage. Science 295:2091–2094 Lively CM (2010) A review of red queen models for the persistence of obligate sexual reproduction. J Hered 101:S13–S20 Lopez Pascua L, Hall AR, Best A, Morgan AD, Boots M, Buckling A (2014) Higher resources decrease fluctuating selection during host–parasite coevolution. Ecol Lett 17:1380–1388 Lopez-Pascua LDC, Buckling A (2008) Increasing productivity accelerates host-parasite coevolution. J Evol Biol 21:853–860 Luckey M, Wayne R, Neilands J (1975) In vitro competition between ferrichrome and phage for the outer membrane T5 receptor complex of Escherichia coli. Biochem Biophys Res Commun 64:687–693 Luijckx P, Fienberg H, Duneau D, Ebert D (2013) A matching-allele model explains host resistance to parasites. Curr Biol 23:1085–1088 Mahony J, McGrath S, Fitzgerald GF, van Sinderen D (2008) Identification and characterization of lactococcal-prophage-carried superinfection exclusion genes. Appl Environ Microbiol 74:6206–6215 Maillou J, Dreiseikelmann B (1990) The sim gene of Escherichia coli phage P1: nucleotide sequence and purification of the processed protein. Virology 175:500–507 Makarova KS, Wolf YI, Koonin EV (2013) Comparative genomics of defense systems in archaea and bacteria. Nucleic Acids Res 41:4360–4377 McGrath S, Seegers JF, Fitzgerald GF, van Sinderen D (1999) Molecular characterization of a phage-encoded resistance system in Lactococcus lactis. Appl Environ Microbiol 65:1891–1899 Meaden S, Koskella B (2013) Exploring the risks of phage application in the environment. Front Microbiol 4:358 Meyer JR, Dobias DT, Weitz JS, Barrick JE, Quick RT, Lenski RE (2012) Repeatability and contingency in the evolution of a key innovation in phage lambda. Science 335:428–432 Molineux IJ, Panja D (2013) Popping the cork: mechanisms of phage genome ejection. Nat Rev Microbiol 11:194–204 Munsch-Alatossava P, Alatossava T (2013) The extracellular phage-host interactions involved in the bacteriophage LL-H infection of Lactobacillus delbrueckii ssp. lactis ATCC 15808. Front Microbiol 4:408 Nechaev S, Severinov K (2008) The elusive object of desire – interactions of bacteriophages and their hosts. Curr Opin Microbiol 11:186–193 Nordström K, Forsgren A (1974) Effect of protein a on adsorption of bacteriophages to Staphylococcus aureus. J Virol 14:198–202 Nordström K, Forsgren A, Cox P (1974) Prevention of bacteriophage adsorption to Staphylococcus aureus by immunoglobulin G. J Virol 14:203–206 Nuismer SL, Doebeli M, Browning D (2005) The coevolutionary dynamics of antagonistic interactions mediated by quantitative traits with evolving variances. Evolution 59:2073–2082 Paez-Espino D, Morovic W, Sun CL, Thomas BC, Ueda K, Stahl B, Barrangou R, Banfield JF (2013) Strong bias in the bacterial CRISPR elements that confer immunity to phage. Nat Commun 4:1430 Parma DH, Snyder M, Sobolevski S, Nawroz M, Brody E, Gold L (1992) The Rex system of bacteriophage lambda: tolerance and altruistic cell death. Genes Dev 6:497–510 Paterson S, Vogwill T, Buckling A, Benmayor R, Spiers AJ, Thomson NR, Quail M, Smith F, Walker D, Libberton B, Fenton A, Hall N, Brockhurst MA (2010) Antagonistic coevolution accelerates molecular evolution. Nature 464:275–U154

250

M. A. Brockhurst et al.

Poullain V, Gandon S, Brockhurst MA, Buckling A, Hochberg ME (2008) The evolution of specificity in evolving and coevolving antagonistic interactions between a bacteria and its phage. Evolution 62:1–11 Refardt D, Bergmiller T, Kümmerli R (2013) Altruism can evolve when relatedness is low: evidence from bacteria committing suicide upon phage infection. Proc R Soc B Biol Sci 280:20123035 Reyes-Cortés R, Martínez-Peñafiel E, Martínez-Pérez F, de la Garza M, Kameyama L (2012) A novel strategy to isolate cell-envelope mutants resistant to phage infection: bacteriophage mEp213 requires lipopolysaccharides in addition to FhuA to enter Escherichia coli K-12. Microbiology 158:3063–3071 Roach DR, Sjaarda DR, Castle AJ, Svircev AM (2013) Host exopolysaccharide quantity and composition impact Erwinia amylovora bacteriophage pathogenesis. Appl Environ Microbiol 79:3249–3256 Samson JE, Magadán AH, Sabri M, Moineau S (2013) Revenge of the phages: defeating bacterial defences. Nat Rev Microbiol 11:675–687 Sasaki A (2000) Host-parasite coevolution in a multilocus gene-for-gene system. Proc R Soc B Biol Sci 267:2183–2188 Scanlan PD, Hall AR, Lopez-Pascua LDC, Buckling A (2011) Genetic basis of infectivity evolution in a bacteriophage. Mol Ecol 20:981–989 Scott AE, Timms AR, Connerton PL, Loc Carrillo C, Adzfa Radzum K, Connerton IF (2007) Genome dynamics of campylobacter jejuni in response to bacteriophage predation. PLoS Pathog 3:e119 Seed KD, Faruque SM, Mekalanos JJ, Calderwood SB, Qadri F, Camilli A (2012) Phase variable O antigen biosynthetic genes control expression of the major protective antigen and bacteriophage receptor in Vibrio cholerae O1. PLoS Pathog 8:e1002917 Seed KD, Lazinski DW, Calderwood SB, Camilli A (2013) A bacteriophage encodes its own CRISPR/Cas adaptive response to evade host innate immunity. Nature 494:489–491 Sørensen MCH, van Alphen LB, Harboe A, Li J, Christensen BB, Szymanski CM, Brøndsted L (2011) Bacteriophage F336 recognizes the capsular phosphoramidate modification of Campylobacter jejuni NCTC11168. J Bacteriol 193:6742–6749 Sørensen MCH, Van Alphen LB, Fodor C, Crowley SM, Christensen BB, Szymanski CM, Brøndsted L (2012) Phase variable expression of capsular polysaccharide modifications allows Campylobacter jejuni to avoid bacteriophage infection in chickens. Front Cell Infect Microbiol 2:11 Srinivasiah S, Bhavsar J, Thapar K, Liles M, Schoenfeld T, Wommack KE (2008) Phages across the biosphere: contrasts of viruses in soil and aquatic environments. Res Microbiol 159:349–357 Sun CL, Barrangou R, Thomas BC, Horvath P, Fremaux C, Banfield JF (2013) Phage mutations in response to CRISPR diversification in a bacterial population. Environ Microbiol 15:463–470 Thompson JN (1994) The coevolutionary process. University of Chicago Press, Chicago Thompson JN (2005) The geographic mosaic of coevolution. University of Chicago Press, Chicago Thompson JN (2013) Relentless evolution. University of Chicago Press, Chichago and london Thompson JN, Burdon JJ (1992) Gene-for-gene coevolution between plants and parasites. Nature 360:121–125 Van Valen L (1973) A new evolutionary law. Evol Theory 1:1–30 Vasu K, Nagamalleswari E, Nagaraja V (2012) Promiscuous restriction is a cellular defense strategy that confers fitness advantage to bacteria. Proc Natl Acad Sci USA 109:E1287–E1293 Westra ER, Swarts DC, Staals RH, Jore MM, Brouns SJ, van der Oost J (2012) The CRISPRs, they are a-changin': how prokaryotes generate adaptive immunity. Annu Rev Genet 46:311–339 Williams HT (2013) Phage-induced diversification improves host evolvability. BMC Evol Biol 13:17 Wilson GG, Murray NE (1991) Restriction and modification systems. Annu Rev Genet 25:585–627 Wommack KE, Colwell RR (2000) Virioplankton: viruses in aquatic ecosystems. Microbiol Mol Biol Rev 64:69–114

Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy

251

Yan J, Mao J, Xie J (2014) Bacteriophage polysaccharide Depolymerases and biomedical applications. BioDrugs 28:265–274 Yoon SH, Han MJ, Jeong H, Lee CH, Xia XX, Lee DH, Shim JH, Lee SY, Oh TK, Kim JF (2012) Comparative multi-omics systems analysis of Escherichia coli strains B and K-12. Genome Biol 13(5):R37 Young R (2013) Phage lysis: do we have the hole story yet? Curr Opin Microbiol 16:790–797 Zaleski P, Wojciechowski M, Piekarowicz A (2005) The role of Dam methylation in phase variation of Haemophilus influenzae genes involved in defence against phage infection. Microbiology 151:3361–3369 Zhang Q-G (2014) Exposure to phages has little impact on the evolution of bacterial antibiotic resistance on drug concentration gradients. Evol Appl 7(3):394–402 Zhang Q-G, Buckling A (2011) Antagonistic coevolution limits population persistence of a virus in a thermally deteriorating environment. Ecol Lett 14:282–288 Zhang Q-G, Buckling A (2012) Phages limit the evolution of bacterial antibiotic resistance in experimental microcosms. Evol Appl 5:575–582

Bacteriophage Ecology John J. Dennehy and Stephen T. Abedon

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Presence in Nature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Viable Phage Counts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Total Counts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nucleic Acid Sequence-Based Determinations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Biogeography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Population Ecology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Life Cycles as Ecological Phenomena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Virion Durability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Virion Attachment Affinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Evasion of Host Defenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Latent Periods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Burst Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Community Ecology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Predators, Parasites, or Parasitoids? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Killing the Winner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Exploitation of Biofilm-Associated Bacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Ecosystem Ecology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

254 255 256 264 266 267 271 273 273 275 276 277 278 279 280 281 282 284 286 287 287

Abstract

Ecology is the study of the interactions of organisms with their environments. Environments consist of biotic components (e.g., cellular organisms such as J. J. Dennehy Biology Department, Queens College and The Graduate Center of the City University of New York, New York, NY, USA e-mail: [email protected] S. T. Abedon (*) Department of Microbiology, The Ohio State University, Mansfield, OH, USA © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_8

253

254

J. J. Dennehy and S. T. Abedon

bacterial hosts) and abiotic components (e.g., temperature, pH, or ultraviolet radiation). Bacteriophage ecology thus is the study of the interactions of bacteriophages (phages) with biotic as well as abiotic aspects of their environments. The fundamental focus of ecology tends to be on populations, that is, genetically coherent groupings of organisms (species). Often this also involves considerations of the ecology of communities (interactions among multiple species in the same area) as well as the ecology of ecosystems (which include, conceptually, both communities and abiotic aspects of environments). Also within the purview of ecology are evolutionarily relevant interactions between organisms such as antagonistic coevolution, which here would be “arms races” between phages and bacteria. Ecology can be studied either observationally or empirically. Metagenomics, the study of the collective genetic material acquired from environmental samples, currently is the most prominent means of observational study in environmental microbiology, the latter an aspect of microbial ecology and thus (especially as viromics) of phage ecology as well. Empirical studies, by contrast, commonly use “models” to explore ecology, including ones exploring phage ecology. These models range from broth culture-based experiments in the laboratory to experimental microcosms located in situ, the latter such as in a pond but still separated from the pond. Particularly as theoretical studies, models also can be mathematical or computer-based. All of these approaches can be combined toward comprehending how individual organisms, groups of organisms, ecosystems, or even groups of ecosystems function in nature. In this chapter, we consider the ecology of phages as well as the ecological aspects of their evolutionary biology. As what is addressed can be considered to be the bulk of phage biology as found outside of well-controlled laboratory systems, the chapter covers substantial ground. As complementary material in this volume, we point the reader to the biology of phage productive infections, lysogeny, bacterial resistance to phages, and, as a form of applied phage ecology, phage therapy pharmacology.

Introduction The emphasis of early microbiology – as practiced prior to roughly the 1950s – was by necessity less molecular than it is today. Microorganisms were often considered as whole organisms, cultures of organisms, or primarily in terms of their impact on other organisms such as ourselves. This perspective was superseded by molecular approaches that tended to view microorganisms as collections of highly interactive molecules. Molecular approaches to microbiology can be very powerful, succeeding in elucidating, for example, the biochemical basis of genetics, ultimately giving rise to today’s emphasis on genome sequencing. Bacteriophages, as one of the primary model systems of biology’s molecular revolution (Cairns et al. 1966; Wasik et al. 2015), consequently have been described in exquisite molecular detail, and increasingly this is being done using ever more sophisticated technologies. In addition to being somewhat less molecular than this in its emphases, early microbiology also

Bacteriophage Ecology

255

often was more ecological in its perspective than it is today, e.g., with substantial consideration in its first decades of what grew on what or where or how rapidly, etc. During phage biology’s molecular revolution, however, ecological considerations – though often deemphasized (Abedon 2018, 2020c) – nevertheless have not been lost, and indeed have been strengthened by our increasing appreciation of the specific molecular basis of phage phenotypes. Ecology, though involves especially the interaction of organism phenotypes with environments. As too with microbiology, the emphases of ecology can be differentiated into a number of distinct perspectives. We can consider ecophysiology, which is study of the adaptations that individual organisms employ to deal with abiotic, which is study of nonliving aspects of environments. This approach can include study of the biology of phages in terms of the durability of their virions. We can consider as well the ecology of organism populations, such as the study of rates of phage population growth. We also can consider the interactions of organisms with other species, i.e., biotic components of environments, with such interspecific interactions collectively described as community ecology. This consists for phages especially the study of interactions with bacterial hosts including, for example, the impact of phages on bacterial densities within environments. Ecosystem ecology considers the contribution of organisms to the cycling of nutrients and flow of energy within and through environments. This, worldwide, turns out to be substantially microorganism driven with a significant phage impact (Fuhrman 1999; Wagg et al. 2014). In addition, there is mathematical ecology, which includes explorations of ecological principles by using model systems that are so basic that they can exist on paper or in computers. Furthermore, there is evolutionary ecology, which considers especially the evolutionary relevance of ecological adaptations, such as phage genetic adaptation to new hosts. Additional aspects of ecology along with evolutionary biology, and their various connections, are presented in Fig. 1. In this chapter, we provide a more whole organism, population, community, ecosystem, and also evolutionary adaptation emphasizing perspectives on phage biology. This is rather than a highly molecular orientation, though keep in mind that all phage molecular aspects can be considered, in an ultimate sense, to be products of ecological and evolutionary forces. We start, though, with consideration of what can be described as phage environmental microbiology, i.e., phage presence in nature in terms of types, numbers, and locations. Definitions of key terms can be found in Table 1.

Phage Presence in Nature In this section, we consider the environmental microbiology of phages, addressing questions such as what types of phages exist in environments, where they exist, and in what numbers. Despite typical equating of these endeavors with the concept of phage ecology, environmental microbiology in fact places less emphasis on the interactions of microorganisms within environments, instead concentrating on microorganism identification and enumeration as well as biogeography. The latter

256

J. J. Dennehy and S. T. Abedon

Environmental Microbiology

Ecosystem Ecology

Macroevolution

Horizontal Gene Transfer (HGT)

Speciation Community Ecology

Organism Relatedness

Ecology

Evolution

Population Ecology

Coevolution Evolutionary Ecology

Experimental Ecology

Ecophysiology

Microevolution

Experimental Evolution

Fig. 1 Various perspectives on ecology and relationships of ecology with evolutionary biology. Arrows indicate relatively prominent connections. Dashed lines are meant to separate relatively unrelated concepts, which nevertheless are found in close proximity as drawn. This figure is derived from as presented in Abedon (2009b)

is study of the spatial locations of organisms within the biosphere as well as their genetic (evolutionary) relationships to organisms found in different locations. Phage numbers in specific environments can be explored via phage viable counts as typically employ plaque assays (chapter ▶ “Detection of Bacteriophages: Phage Plaques”), total or direct counts (i.e., microscopically, though also as can be determined via flow cytometry), in terms of nucleic acid sequence (e.g., metagenomics, qPCR, fluorescent in situ hybridization, aka, FISH), or in terms of phage genome association with specific host bacteria.

Viable Phage Counts Viable count determinations are the oldest means of exploring phage occurrence within environments (d'Hérelle and Smith 1926). Here the presence of phages is determined particularly as plaques (chapter ▶ “Detection of Bacteriophages: Phage Plaques”). Emphasis only on plaque-forming ability, however, significantly underestimates phage densities in environments, as discussed more fully below. Viable count-based studies consequently historically had an effect of discouraging further study of phage ecology. Specifically, environmental phage densities as determined

Bacteriophage Ecology

257

Table 1 Definitions of key terms Abiotic

Adaptation (evolutionary)

Allele Antagonistic coevolution

Antagonistic pleiotropy

Autotroph

Biogeography

Biosphere Biotic

Burst size

Nonliving aspects of environments, including mineralized nutrients, even if those nutrients were sourced from a living thing, e.g., dissolved organic carbon Refers to evolutionary changes resulting from natural selection. That is, adaptations are aspects of organism phenotypes that result in greater reproductive, survival, or competitive abilities for organisms within specific habitats Variant of a gene as found at a given locus within an organism’s genome Arms races between interacting species, as most notably seen with phages in terms of evolution of bacterial resistance to specific phages, with phage populations countering by evolving to overcome those resistance mechanisms. With antagonistic coevolution, the adaptations by different organisms, in other words, serve to counter or neutralize each other Allelic changes in an organism that improve the evolutionary fitness associated with some phenotypes while reducing the evolutionary fitness associated with others or that improve fitness within one environmental context while reducing fitness within a different environmental context. For example, changes to phage tail fibers may improve virion adsorption affinities for one host type while reducing virion adsorption affinities for a different host type Organism capable of fixing environmental carbon dioxide, that is, converting an inorganic form of carbon to an organic form. Cyanobacteria, algae, and plants are autotrophic organisms (collectively, photoautotrophs), but autotrophs include as well bacteria and archaea that obtain their energy from inorganic sources and their organic carbon by fixing environmental carbon dioxide, that is, which exist as chemoautotrophs or chemolithoautotroph. The so-called mixotrophs are capable of obtaining carbon from both organic and inorganic sources, so serve as autotrophs as well Study of the distribution of organisms in space, i.e., organisms are not evenly present over the surface of the earth, and biogeography attempts to catalog that unevenness, provide explanations as to why it exists, also suggest consequences of such unevenness on the trajectories of organism evolution Totality of ecosystems present both in and on the earth Living entities found in environments, potentially including notyet-mineralized aspects of dead organisms. Prey organisms, which have been killed but not yet eaten, are generally still considered to be biotic components of environments. Functioning viruses, even as virions, generally are considered to be biotic components of environments Average number of phage particles (virions) produced per phageinfected bacterium, particularly as associated with lytic productive phage infections (continued)

258

J. J. Dennehy and S. T. Abedon

Table 1 (continued) Coevolution

Community

Community ecology Conspecific

Consumer

Decomposer

Decomposition

Direct count Dispersion

Dissolved organic carbon (DOC) Ecology Ecophysiology

Ecosystem

Interdependent evolution of two or more species that have ecological relationships, that is, as found in the same environment and impacting each other’s functioning. Coevolution for example can result from beneficial interactions, e.g., such as between prophages and bacterial hosts. In the context of phage-host relations, however, coevolution typically is studied instead in terms of antagonistic adaptations, i.e., see antagonistic coevolution Organisms found in the same location consisting of more than one species. A single phage population and a single bacterial population thus together make up a simple community but so too will more than one phage species or more than one bacterial species together make up a community Study of the interactions between different species within an environment Member of the same species as a focus individual, e.g., two phages together making up the same species are conspecifics. Note, though, that the concept of species among phages is itself not always well defined An organism that feeds on other organisms or the products of other organisms such as dissolved organic carbon. Heterotrophic bacteria by definition are consumers. Phages, as utilizers of bacteria-associated organic building blocks and energy supplies, too are consumers An organism that feeds by degrading otherwise dead organic matter, particularly organic matter associated with somewhat intact organisms. Consumers of only dissolved organic carbon at best can only arguably be considered to be decomposers The degradation of organic matter ultimately into simple organic as well as inorganic compounds, particularly as released directly into the environment (and thereby contrasting simply digestion). The impact of lysis on bacteria can be viewed as an aspect of decomposition Determination of microorganism numbers in terms of their recognizable presence, typically as via microscopy Related to the concept of “dispersal,” the dispersion of a population is the pattern of its location in space, i.e., such as clumped, random, or even. Clumping comes from either attraction of individuals to each other or instead a failure or progeny to leave their site of formation. Even dispersions are a consequence of active repulsion of individuals from each other Solubilized environmental organic materials as derived via organism decomposition. One consequence of lysis of bacteria is the generation of DOC Study of the interaction of organisms with their environments Study of the physiology of organisms and their associated evolutionary adaptations under different abiotic environmental conditions, e.g., such as different temperatures or different nutrient levels A combination of the biotic and abiotic components of an environment (continued)

Bacteriophage Ecology

259

Table 1 (continued) Ecosystem ecology

Effective burst size

Environment Environmental microbiology

Epifluorescence microscopy (EFM) Evolution

Evolutionary ecology Feast or famine

Fitness

Fixation (of alleles)

Fixation (of carbon)

Flow cytometry

Focus of infection

Study of the flow of energy through and cycling of nutrients within environments. Such study by necessity includes consideration of both biotic environmental components and abiotic environmental components Fraction of a burst size that survives to initiate successful new infections of their own, i.e., as affected by a combination of a phage’s actual burst size and the likelihood of individual phages not being inactivated prior to successfully infecting a suitable bacterium either productively or lysogenically The physical, chemical, and biological surroundings of an organism or community Study of microorganisms as observed within natural environments, including in terms of assessing what organisms are present, in what quantities, and with what biochemical characteristics. Metagenomic studies by and large are environmental microbiology studies Optical microscopy technique that employs ultraviolet irradiation to induce fluorescent materials to emit visible light, such as labeled nucleic acid of virus particles Changes in the frequency of what alleles are found within populations, where such changes can occur as a consequence of mutation, genetic migration (movement of alleles between populations), natural selection, or sampling error (the latter, i.e., genetic drift) Study of the ecological context of organism evolutionary adaptation Somewhat binary concept of organism exploitation of environments where resources are envisioned to be patchy over either time or space. Organisms therefore either are in the process of searching or waiting for resources (famine) or instead have found resources which they therefore are exploiting (feast) Measure of the anticipated short-term evolutionary success of a genotype. Thus, a given organism with a given set of alleles is expected, particularly as based on past measurements, to successfully produce a given average number of offspring over some length of time An allele associated with a given genetic locus that has come to be effectively the only allele found at that locus within a gene pool. Not to be confused with carbon fixation Conversion by autotrophic organisms of inorganic forms of carbon (particularly carbon dioxide) to organic forms of carbon such as sugars. Not to be confused with fixation of alleles Technique where fluorescently labeled cells or particles can be counted or characterized by suspending them in a thin stream of fluid that passes through a laser beam detector Localized region of phage exploitation of spatially constrained, especially clonal bacterial populations. In natural environments, such bacteria are usually bacterial microcolony or biofilm associated. In the laboratory, a phage focus of infection is exemplified especially as a phage plaque. A focus of infection is exploited in the course of gradual outward migration of that phage population starting at a single, initially phage-infected bacterium (continued)

260

J. J. Dennehy and S. T. Abedon

Table 1 (continued) Genome Genotype

Habitat Heterotroph

Historical contingency

Horizontal gene transfer (HGT) Host range Indicator

Killing the winner

Latent period Life-history trait

Lysis

Lysis inhibition

An organism’s complete set of genetic material, including all of its genes and allied noncoding regions Genetic makeup of an organism as distinguished particularly in terms of what alleles are present. Genotypes classically were distinguished based on upon how a single trait or set of traits differs within a group of individuals or a species. With the advent of DNA sequencing, genotypes instead are typically distinguished in terms of what nucleotide bases are present and in what order The local environment occupied by an organism Organism capable of assimilating environmentally sourced organic carbon. All animals and fungi, all non-photosynthetic protists, most bacteria, and many archaea are heterotrophic organisms. Heterotrophic bacteria are primary consumers of dissolved organic carbon, particularly within marine environments Limitation on the ability of an organism to evolve that is based on the number of mutational steps required to go from a currently successful genotype (the “history”) to a new reasonably successful genotype Movement of hereditary material from one organism and subsequent integration into another organism, other than from parent to offspring All of the types of bacteria, strains as well as species, that a specific phage isolate can infect Here, a bacterium that can support especially phage population growth toward plaque formation, i.e., the bacterial strain used for lawn formation Description of the vulnerability of especially highly abundant or alternatively rapidly replicating bacteria (i.e., winners) to substantial exploitation by co-localized phages (i.e., killing). Also described as kill the winner Duration of a phage lytic infection, starting with phage adsorption and ending with phage-induced bacterial lysis Specific feature that helps to define an organism’s life cycle, especially with reference to strategies that influence survival and reproduction, such as phage latent period length, burst size, and host range Destruction of a bacterial cell via the disruption of the plasma membrane and cell wall. For lytic phages, lysis is employed to release newly formed virions into the extracellular environment Mechanism of lytic phage latent period extension and burst size increase that is induced by adsorption of phages of the same time to already phage-infected bacteria. Specifically, a delay of minutes is required between the two (or more) adsorptions, with the first adsorbing phage described as a primary phage (primary adsorbing or infecting phage) and the second described as a secondary phage (secondarily adsorbing phage) (continued)

Bacteriophage Ecology

261

Table 1 (continued) Lysogeny

Lytic infection

Metagenome

Microcolony

Macroevolution

Microevolution

Mineralization Natural selection

Parasite Parasitoid

Condition in which a phage genome is found latently infecting a bacterium. Lysogenic phage infections are also described as lysogenic cycles and bacteria hosting phage lysogenic cycles are described as lysogens Phage infection beginning either with virion adsorption or prophage induction and ending with phage-induced bacterial lysis, especially as involving the maturation and release of new virion particles (phage progeny) The collective genetic material of especially all microorganisms inhabiting a particular location, as often characterized via highthroughput sequencing techniques (the latter, i.e., metagenomics or metagenomic analyses) Group of spatially associated cells which are clonally related, having been derived from a single founder cell, but which consist of an insufficient number of cells to be easily distinguished from their background by the naked eye. Biofilms often consist of multiple, individual microcolonies, though unlike individual microcolonies, biofilms themselves are not necessarily either clonal or consisting of just a single bacterial species Evolution of species lineages (typically as depicted as phylogenies), extinction of species, formation of new species, and relatedness of different species, that is, evolution as it occurs above the level of species. Macroevolution in phages can be greatly affected by horizontal gene transfer both toward generation of new phage varieties and in terms of making it difficult or impossible to trace lines of decent of phage lineages from common ancestors Evolution as it occurs below the level of the species, i.e., as consisting of the familiar mutation, genetic migration, natural selection, and genetic drift Conversion of materials to inorganic forms, such as conversion of organic carbon to carbon dioxide The nonrandom differential reproductive success of individual organisms through loss of unfavorable variants. This is a result especially of differences, in organism phenotypes, that lead to the perpetuation of certain underlying genotypes over others Organism that harms, but does not kill another organism, while at the same time obtaining various benefits from that other organism Organism that harms but does not kill another organism in the course of obtaining benefits from that other organism, but nevertheless ultimately does kill the hosting organism. This term typically is used to describe the placing of eggs within a living host organism, e.g., wasp parasitoids of beetles, with the juvenile wasp then consuming the beetle from within in the course of wasp development, and with the wasp eventually killing and then emerging from the beetle in a matured form. The description nevertheless may also be applied to the actions of phages during lytic infections in which infected bacteria are kept physiologically “alive” by phages in the course of maturation of new phage virions, all prior to killing of the host bacterium to release those virions into the extracellular environment (continued)

262

J. J. Dennehy and S. T. Abedon

Table 1 (continued) Patch (ecological)

Phage susceptibility type

Phagotrophic Phenotype Pleiotropy

Plaque

Polyphyletic Population Population ecology Predator Primary productivity Productive infection

Prophage

Secondary productivity

Sloppy eating

Description of a specific location within an environment. Environments, particularly given spatial structure, thus consist of multiple specific locations, each possessing distinct properties and characteristics such as in terms of what species are present and in what densities, or in terms of nutrient densities, etc. Those bacterial types, not necessarily all of the same species, that are susceptible to a specific phage type, thus as found within that phage’s host range. Typically that susceptibility is measured in terms of a phage’s ability to replicate, but also can be just a phage’s ability to kill a bacterium. This concept can be used within a context of the concept of killing the winner Referring to a means of nutrient acquisition by cells that involves the process of phagocytosis All observable aspects of an organism, except the sequence of nucleotides making up its genome Association of more than one phenotype with a given allele. Pleiotropies can place limitations on the improvement of adaptations as increases in benefits from improvement of one phenotype (as an adaptation) often can be associated with decreases in benefits associated with another phenotype, a concept known as antagonistic pleiotropy Especially a macroscopically visible region of localized phage population growth and reduced bacterial concentrations, particularly as seen in the laboratory in association with bacterially inoculated agar found in petri dishes Derived from two or more distinct ancestral lineages All organisms of a species that are found in a particular geographical area Study of the interactions between members of the same species Organism that kills another organism prior to consuming that other, prey organism New organism mass (biomass) generated in association with the conversion of abiotic sources of carbon into organic compounds Phage infection that results in the release of progeny phage, either by lysis of the host cell or by chronic release that does not kill the host Phage genome found in association with a host bacterium in a lysogenic infection or cycle. Prophages are able to produce new phage virions if activated in a process known as prophage induction New organism mass (biomass) generated by heterotrophic organisms, i.e., in the course of consumption of other organisms including in association with consumption of dissolved organic carbon When consumer organisms fail to fully exploit or ingest the nutrients associated with a given prey organism but instead as a consequence of the consumption process release many of these nutrients into the environment. The dissolved organic carbon released in the course of phage-induced lysis of bacteria arguably can be described as a consequence of sloppy “eating” (continued)

Bacteriophage Ecology

263

Table 1 (continued) Spatial structure

Speciation Species

Strictly lytic Strictly productive

Synteny

Temperate

Total count

Tradeoff

Transduction Transmission electron microscopy (TEM) Trophic Trophic level

Viable count

Virion Virus-like particle (VLP)

Limitations on organism movement within an environment such that, as a consequence, environments are likely to be spatially heterogeneous, e.g., with some locations displaying higher densities or certain bacteria or their phages, while other locations possess lower densities Formation of one or more new species from a previously existing species Population whose members are sufficiently similar as to be relatively “reproductively” isolated from other such populations, here meaning unlikely to recombine with other phages or produce viable recombinant progeny, even if located sympatrically (i.e., found within the same environment). Like bacteria, phage species tend to be defined, if they are defined at all, based on sequence similarities between phage genomes Phages capable of lytic infections but incapable of lysogenic infections Description of phages that are unable to lysogenize bacteria, usually described more narrowly as strictly lytic though chronically releasing phages can be strictly productive as well. Temperate phages by definition are not strictly productive Physical association of two or more loci within two or more genomes, i.e., similar gene orders between similar but not identical isolates Description of phages that can both productively infect and lysogenize their hosts, though not at the same time, as opposed to strictly productive, e.g., strictly lytic phages, either of which cannot lysogenize Determination of total numbers of organism present, without effort made to distinguish between viable and not viable individuals. Often this is accomplished via microscopic direct counts A situation where increases in the performance of one trait are associated with a reduction in the performance of another trait or traits. See, for example, the concept of antagonistic pleiotropy Virion-mediated horizontal gene transfer of genetic material from one cell to another Microscopy technique that uses an electron beam to image features or objects that are smaller than the resolution limit of light microscopy Refers to nutrient acquisition Refers to consumption of organisms by other organisms, where organisms that are found at higher trophic levels (e.g., predators) consume organisms found at lower trophic levels (e.g., prey) Determination of population numbers in terms of organism potential to individually reproduce. For phages, plaque counts are a typical viable count A complete, infective form of a phage, aka, a virus particle As used in virus ecology, these are entities that are virion-like in appearance as viewed microscopically

264

J. J. Dennehy and S. T. Abedon

by viable counts did not appear to be large enough to substantially impact bacterial populations. Phage ecology consequently was initially studied primarily by scientists with an intrinsic interest in phages (Campbell 1961), an interest in phages as environmental components or fecal indicators (Goyal et al. 1987), or an interest in using phages as models for the study of broader ecological principles (Chao et al. 1977; Levin et al. 1977), but not because phages were thought to be substantial contributors to ecosystem properties. Interest in phage ecology increased dramatically with development of total count methods of phage in situ enumeration (next subsection). Still, it is an important point to keep in mind that viable count determinations of phage numbers within environments are not lacking in meaning. Instead, they are suggestive that often the number of phages present, especially those that can form plaques on specific indicator bacteria, tends to not be dramatically high. This suggests that individual phage populations frequently are “struggling” for existence rather than undergoing substantial population growth and otherwise contributing to the devastation of specific bacterial populations. Viable count determinations, though “imperfect,” nevertheless may thus point to a partial “truth” about actual phage prevalence within environments. There are at least three reasons that phage viable counting can underestimate total phage prevalence within environments. First, a prerequisite for plaque assays is the culturing of permissible bacteria. Unfortunately, only a small minority of bacteria are readily culturable in laboratory media (Staley and Konopka 1985; Lloyd et al. 2018), and phages that cannot infect a plating host cannot be counted by viable counts. Related to this shortcoming, choices of indicator bacteria may be inappropriate, such as use of laboratory bacterial strains rather than whatever bacterial population or populations are supporting phage populations in situ. Second, not all phages form plaques even if those phages are otherwise viable and able to productively infect a given indicator host. This lack of plaquing ability is often the case with filamentous phages (Bradley et al. 1982; Waldor and Mekalanos 1996; Gonzalez et al. 2002), but phages that fail to plaque can also include tailed phages (Bradley et al. 1982; Loftus and Delisle 1995; Los et al. 2008) as well as tailless phages (Horiuchi and Adelberg 1965; Molnar and Lawton 1971; Bradley et al. 1982). Third, not all phages are present within environments as virions, but may exist instead as prophages, and not all prophages are readily induced or propagated (Howard-Varona et al. 2017), though prophages also will not contribute to virion total counts. An additional issue, technical but still relevant, is that during storage of environmental samples, viable counts often will decline due to the inactivation of viable phages over time. These issues together give rise to what can be described as “the great plate count anomaly” (or “great plaque count anomaly”). That is, viable count determinations of phages within environments seem to inevitably result in undercounts of total phage or virion numbers (Weinbauer 2004).

Phage Total Counts With the availability of modern microscopic and flow cytometric techniques, it is relatively easy to obtain total or direct counts of viruses from environmental

Bacteriophage Ecology

265

samples. Transmission electron microscopy (TEM) was employed almost immediately following its invention to characterize phage morphology (Peankuch and Kausche 1940; Ruska 1940), but it wasn’t until the 1970s that researchers used TEM to assess phage abundance in environmental isolates (Proctor 1997). It was through TEM observations, especially in the late 1980s and early 1990s, that aquatic environments were found to harbor vast numbers of phages (Bergh et al. 1989; Wommack and Colwell 2000). These observations were later extended to terrestrial habitats (Williamson et al. 2017). It would not be an exaggeration to claim that TEM studies revolutionized phage ecology. Nevertheless, and though effective, TEM is time-consuming and expensive. Sample preparation and observation are exacting tasks, and acquiring sufficient data from samples can be challenging. Moreover, automation of data acquisition by TEM is limited. Furthermore, and contrasting viable counts, TEM suffers from a tendency to overestimate phage numbers, as discussed in more detail below. Epifluorescence microscopy (EFM) is also employed to directly count phages in environmental samples (Williamson et al. 2003, 2005). Here phages can be observed and counted using light microscopes after virions have been labeled with fluorochromes that are specific for nucleic acids. DAPI (40 ,60 -diamino-2-phenylindole), a fluorochrome specific for double-stranded DNA, was the first fluorochrome used for labeling and counting of aquatic phages (Porter and Feig 1980). Though EFM counts can be more precise than TEM counts, they also suffer from a tendency to overestimate phage numbers (Bettarel et al. 2000; Forterre et al. 2013). EFM like TEM also is somewhat laborious, operator-skill dependent, and low in throughput, though does employ microscopes which are less expensive. Another method of procuring total environmental phage counts is flow cytometry. Similar to EFM, viruses are labeled with fluorophores, but instead of direct observation, a thin stream of fluid containing fluorophore-conjugated phages is passed through a laser detector, and the resulting data is analyzed computationally. Flow cytometry has mainly been employed to count aquatic phages (Marie et al. 1999), but can also be used to count phages extracted from porous materials and suspended in buffer (Williamson et al. 2013). Though high-throughput and mostly operatortime-independent, flow cytometry also tends to overestimate phage numbers in samples and do so to an even greater degree than even EFM counts. For example, in one study, viral abundance as measured by flow cytometry was 350- to 1,400-fold higher than viral abundance as measured by EFM for the same soil (Williamson et al. 2013). Such overestimations are, as noted, a general problem facing total counting methods. Particularly, there can be difficulties in distinguish phages from non-phage viruses or particles that in various ways look like but nevertheless are not viruses (Forterre et al. 2013). For this reason, researchers tend to classify all total count observations as “viruses-like particles” (VLPs) rather than as “phages,” despite the often numerical dominance of the latter. Furthermore, not all VLPs are viable even if they are phages, e.g., as due to subtle virion deformities (proteinlevel inactivations) or other damages (nucleic-acid-level inactivations), which preclude successful infection. Thus, total counts of phage numbers within

266

J. J. Dennehy and S. T. Abedon

environments, even as high-precision direct counts (i.e., TEM), are not as robust as might be hoped. Total counts also cannot distinguish among phages in terms of basic traits such as host range. Indeed, this includes phenotypically in any way except in terms of presence in a given environment or, except via electron microscopy, even in terms of virion morphology. Ecologically, it is crucial to keep in mind that the total counts of VLPs under most environmental circumstances cannot be equated with viable counts. In addition, a count of 107 VLPs is not necessarily a large number of phages as far as their impact on specific, individual bacterial types is concerned, since at best only a fraction of those VLPs present will tend to possess host ranges – typically as measured in terms of phage viable counts – that include specific bacterial types of interest. Thus, as noted in the previous section, phage total counts typically will greatly exceed phage viable counts and especially so when using only a single, specific bacterial indicator.

Nucleic Acid Sequence-Based Determinations The most powerful technique to ascertain phage abundance in natural habitats is direct environmental genomic sampling, i.e., metagenomics (Breitbart et al. 2002; Casas and Rohwer 2007; Hayes et al. 2017; Gregory et al. 2019), and particularly as has come to be described as viromics (Trubl et al. 2020). With viromics, encapsidated nucleic acid is isolated from environmental samples (e.g., water, soil, plants, feces) via a series of purification steps (e.g., centrifugation, filtration, digestion of unencapsidated nucleic acid) to remove debris and cellular components as well as bacterial, archaeal, and eukaryotic genetic material. Although there are a number of different high-throughput sequencing platforms, all generate random sequence reads that, given the small size of phage genomes, can provide hundreds-fold or even thousands-fold genome coverage of more abundant phage types. Reads are trimmed for quality, computationally aligned, and then assembled into contigs, which are consensus DNA sequences assembled from overlapping reads (Simpson et al. 2009). Phage abundance, distribution, and diversity can be estimated from contig spectra by counting the number of sequences that fall into each contig (Angly et al. 2005, 2009). While general metagenomics methodology is not difficult, there are many complications that can make the identification of phages from metagenomic sequence data difficult. Unlike cellular organisms, phages likely are polyphyletic. Collectively, that is, they probably do not share a universal common ancestor. This lack of common evolutionary heritage means that there are no universal genome elements shared by all or even by most phages. This issue makes it impossible to identify phages in an environmental sample based on a single genetic marker, such as the 16S rRNA genes of bacteria and archaea (Tringe and Hugenholtz 2008). With no genetic marker common to all phages, researchers are obliged to compare newly obtained sequences with existing sequences in data repositories, such as NCBI’s GenBank, using homology-based search tools, such as BLAST (Altschul et al. 1990), to distinguish phages from non-phage VLP nucleic acid. The limitation of this

Bacteriophage Ecology

267

approach is that it can only identify phages with some genetic resemblance to existing phages. Unfortunately, however, the majority, of environmentally acquired sequences have no significant similarity to any known sequence and therefore cannot be attributed specifically to phages (Mokili et al. 2012; Hurwitz et al. 2016; PaezEspino et al. 2016). Phage metagenomic studies also suffer from biases by nucleic acid type (Szekely and Breitbart 2016). Specifically, some common protocols are biased to dsDNA, and others biased to ssDNA (Kim and Bae 2018; Roux et al. 2016). Biases also can stem from differences in phage buoyant densities, which factors into some phage isolation strategies, or from whether phage genomes are linear or circular. In addition, RNA phages are commonly ignored due to the fact that many sequencing approaches specifically exclude RNA molecules (Greninger 2018). No one metagenomic sequencing strategy therefore is able to cover the full breadth of phage diversity within specific habitats, thereby limiting knowledge of what phages are present in any given environment.

Linking Phages to Bacteria via Metagenomics Through the analysis of metagenomic data, it is possible to link phages with their host bacteria even if neither entity has been physically isolated in pure culture. Perhaps the most obvious example is the computational identification of prophages in bacterial genomes (Lima-Mendez et al. 2008; Akhter et al. 2012; Roux et al. 2015). These prophage predictors rely on the detection of sequence similarities between regions of the bacterial genome and known phage genes as well as other “phage-like” genomic features (e.g., AT and GC skew, protein length, transcription strand directionality). Related to this, the sequencing of bacterial CRISPR loci can provide clues as to what phages have infected a bacterial lineage in the past, since CRISPR “spacers” consist of DNA sequence obtained from infecting phage (chapter ▶ “Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy”) (Edwards et al. 2016; Paez-Espino et al. 2016). Alternatively, the identification of bacterial genes in a phage may indicate that a particular bacterial type is susceptible to that phage (Touchon et al. 2017). This means of identification is complicated, though, by the presence of phage genes in bacterial genomes or bacterial genes in phage genomes (Hurwitz et al. 2016) for reasons other than phage infection, e.g., such as potentially due to DNA transformation. Definitive assignment of phage host range, however, generally requires culturing of phage and host, i.e., by means other than environmental sequencing.

Biogeography Biogeography is the study of the geographical distribution of organisms, that is, where they are located, why they are located where they are, and also what consequences there can be of their presence in one location but not another. For example, placental mammals are not indigenous to Australia, are absent at least in part due to sea barriers to placental mammal movement, and one consequence was a

268

J. J. Dennehy and S. T. Abedon

filling of niches in Australia by nonplacental mammals (both monotremes and marsupials), niches that in other locations are filled by placental mammals. Here we consider biogeography in the broadest of strokes, specifically in terms of phage numbers as typically seen in various environments, but also the phage potential to move over fairly substantial distances, i.e., more than as can be accomplished via local diffusion alone.

Phage Prevalence Within Environments Total counts indicate truly astounding numbers of phages in the biosphere, i.e., about 1031 (Cobian Guemes et al. 2016; Mushegian 2020), with these phages found everywhere bacteria are found. In many habitats, phages also outnumber bacteria by approximately tenfold (Cobian Guemes et al. 2016; Parikka et al. 2017). Actual phage numbers vary considerably from habitat to habitat, however. The Earth’s ocean waters (marine water columns) are likely to be the largest reservoir of phages given that the oceans both cover 70% of the Earth’s surface and can be very deep (Gregory et al. 2019), although it is possible that the atmosphere also contains substantial numbers of phages (Whon et al. 2012; Reche et al. 2018; Rosario et al. 2018). Though phage numbers in both environments can quite large (Fuhrman 1999; Wommack and Colwell 2000; Weinbauer 2004; Whon et al. 2012; Hurwitz and Sullivan 2013; Reche et al. 2018), nevertheless phages in these environments are somewhat diffuse compared to within soils and sediments (Pratama and van Elsas 2018). Williamson and others reported greater than 109 viruses per gram of forest and wetland soils in Delaware (Williamson et al. 2005; Williamson 2018). Marine sediments have been found to harbor up to 1010 viruses per gram (Danovaro et al. 2001; Breitbart et al. 2004a). These habitats, of course, represent one extreme of a distribution; some habitats, such as hot deserts, may have less than 104 phages per gram (Williamson et al. 2017). The relative paucity of phages in hot deserts may be an exception among global environments because even the cold deserts of Antarctica harbor virus numbers approaching or exceeding 108 viruses per gram (Williamson et al. 2007). Phages populate biotic aspects of environments as well. Perhaps best studied is the human microbiome, where phages have been observed in the mouth, gastrointestinal tract, bladder, skin, and lungs (Edlund et al. 2015; Hannigan et al. 2015; Tariq et al. 2015; Manrique et al. 2016; Barr 2017; Miller-Ensminger et al. 2018). Phage numbers in organismal microbiomes likely reflect those of their hosts, with highly populated regions like the gut containing the highest numbers of phages. By contrast, low host-abundant habitats, such as the skin, tend to have fewer phages (Hannigan et al. 2015). These phage populations tend to be relatively stable, and there exists a core “phageome” that appears to be consistent across many different individuals (Reyes et al. 2010; Minot et al. 2011; Dutilh et al. 2014; Manrique et al. 2016). Less well studied, but no less important, are the phages populating the microbiomes of other organisms. These phages interact with microbiomes with far-reaching effects on host organism physiology (Chatterjee and Duerkop 2018). Illustrative studies have found phages in great numbers, for example, in the tomato phyllosphere (Morella et al. 2018), the murine gut (Kim and Bae 2018), the termite

Bacteriophage Ecology

269

gut (Tikhe and Husseneder 2017), and in association with corals (Marhaver et al. 2008; Mahmoud and Jose 2017), but efforts to characterize such organismal viromes are nascent. We should expect nonetheless that phages will occur everywhere bacteria are found, and that phage numbers, as noted, will to some degree reflect those of their hosts.

Phage Movement Closer to the concept of phage ecology vs. that of phage environmental microbiology is the issue of phage movement both within and between environments. Phages can move either relative to their immediate environments (i.e., as via diffusion) or in association with the movement of their immediate environments (i.e., as via bulk flow) (chapters ▶ “Adsorption: Phage Acquisition of Bacteria”). Phage association with less fluid but still mobile aspects of environments, such as with animal bodies or in association with aerosolized particles, also can result in their movement. Animals, for example, can serve essentially as mechanical vectors or particles in the air can serve as vehicles for phage transmission between environments. Those mobility-enhancing associations would need to be reversible, however, in order for phages to be deposited in sites other than those directly associated with whatever they are moving with. Such movement relative to environments need not be over large distances or over long periods to effect significant phage dissemination. For example, movement of only a centimeter such as in soils could be viewed as a fairly substantial distance for a phage particle in search of a new host to infect, e.g., such as following hitchhiking on an earthworm. Movement in association with dust in the wind or with birds, on the other hand, could easily move phages many kilometers. The swallowing of phages such as found in food or water, or even in the course of dosing for phage therapy, also serves as examples of phage movement both within bodies and, ultimately, to the extent that these phages survive passage through gastrointestinal tract, to new locations, i.e., as upon defecation. The movement of phages within bodies in the course of phage therapy generally represents a component of what is known as their pharmacokinetics (Dabrowska and Abedon 2019) (chapter ▶ “Bacteriophage Pharmacology and Immunology”). Phage movement can occur as well while in association with a bacterium. Here the phage would be infecting the bacterium, resulting in the bacterium serving essentially as a biological vector. Substantial local movement could occur should that host display motility. This movement could occur over somewhat longer distances given lysogeny since lysogenic cycles (chapter ▶ “Temperate Phages, Prophages, and Lysogeny”) typically will last substantially longer than lytic cycles. Lysogeny thus could be viewed as an aid to phage dissemination and especially so to the extent that bacteria can be actively motile whereas free phages are not. Also in association with bacteria, though potentially less specific and consisting predominantly of phage absence of movement, is phage association with biofilm extracellular polymeric substances (EPS), which are thought to be able to serve as virus reservoirs (Abedon 2011). Phage virions in addition can specifically bind to the mucus produced by animals (Barr et al. 2013). Phages also may become reversibly

270

J. J. Dennehy and S. T. Abedon

absorbed to abiotic aspects of environments that are somewhat stationary, such as organic aspects of soils or sediments. Phage movement of course also occurs more locally within environments via diffusion. Diffusion occurs predominantly within more aqueous milieus, though also can involve phage movement within colloids, such as within biofilm EPS (Fig. 2). The latter movement in fact may be aided by phage- or infection-associated enzymes known as EPS depolymerases (Abedon 2011; Pires et al. 2016; Oliveira et al. 2018) (chapter ▶ “Biofilm Applications of Bacteriophages”), which can break down EPS, thereby increasing phage potential for diffusion through EPS such as to or away from host bacteria. EPS depolymerases, however, tend to be highly substrate specific and EPS molecules tend to be highly diverse, thereby limiting the utility of such adaptations.

Phage Distributions Within Environments Phages are found wherever their hosts are found. Consequently, phage distributions mirror the habitat requirements of their hosts. A good example of this phenomenon is the distribution of phages infecting Vibrio parahaemolyticus. One study reported that vibriophages isolated from coastal waters are genetically diverse, but this Bulk Water (that overlies biofilm) Free Phage

A Extracellular Polymeric Substance (EPS) Biofilm EPS Biofilm

B Lysis

F Biofilm

EPS

C

E D

EPS

Biofilm

Lysis

Biofilm EPS

EPS Biofilm

Microcolony

Fig. 2 Phage virion movement into, within, and out of a biofilm as aided by virion expression of EPS depolymerases. Envisaged is the formation of EPS-reduced corridors through which phage virions can more readily diffuse. Note that if EPS depolymerases are soluble rather than virion associated then only movement away from phage-lysed bacteria may be aided by EPS depolymerase expression (C and E in the figure, shown in green) rather than also movement explicitly toward to-be-infected bacteria (B and D in figure, shown in orange). Shown also is bacterial movement while not in association with biofilm, i.e., A and F in the figure (shown in black). This figure is derived from that of Abedon (2011)

Bacteriophage Ecology

271

diversity is not correlated with geography, e.g., as associated with areas tens of kilometers apart (Comeau et al. 2006). Instead, diversity was correlated with the source from which the vibriophages originated. Vibriophages isolated from oysters were similar no matter where they were found geographically, but different from vibriophages isolated instead from the water column or sediments. These results, along with others finding genetically similar phages or phage genes great distances apart (Short and Suttle 2002; Breitbart et al. 2004b; Angly et al. 2006; Ceyssens et al. 2009; Huang et al. 2010; Eggleston and Hewson 2016; Kalatzis et al. 2017), suggest that there are no spatial limitations to phage dispersal. One of us (JJD, personal communication), for example, isolated a Pseudomonas aeruginosa phage from a junkyard puddle in New York City that is 98.89% genomically identical to the Pseudomonas phage SM1 isolated in Malaysia, a distance of over 15,000 km. On the other hand, some phage populations appear to be geographically restricted, i.e., as over kilometer scales (Tucker et al. 2011; Marston et al. 2013; Huang et al. 2015; Hanson et al. 2016). For example, one study showed almost no overlap in cyanophage diversity between isolates obtained from Southern New England and Bermuda (Marston et al. 2013). Isolation-by-distance analyses of phage sequences, which assess how genetic sequence differences among organisms relate to how far they are apart, show positive correlations between geographic and genetic distances (Angly et al. 2006; Tucker et al. 2011). Strong patterns of isolation by distance are expected to occur only if geographic dispersal is limited. Why some phages but not others show global vs. much more local distributions is a mystery. There is some speculation, however, that life history, habitat, or indeed host differences may explain distributions (Chow and Suttle 2015). The high burst size, short latent period phage HMO-2011, for example, is observed in 10–25% of all genome sequences obtained from ocean surface metagenomes (Kang et al. 2013). Other factors governing phage distributions may include host range, native environmental milieu (e.g., marine vs. soil habitats), and durability.

Phage Population Ecology Populations consist of all of the individuals of a single species inhabiting a particular geographic area at the same time. The concept of species however can be difficult to define, particularly for phages, but at a minimum individuals making up a single species are more similar to each other than individuals sourced from two different species. In the ecological concept of species, this similarity is considered not just from a genetic perspective but also in terms of the degree to which two individuals are likely to ecologically compete for the same resources. Thus, two phages which are genetically similar, i.e., possess substantial gene synteny and DNA homology, and also are similar in virion morphology and host range, are more likely to be members of the same phage species. Population ecology is the study of those attributes of a species that govern how a population changes over time, including the events relating to a species growth, development, reproduction, and survival (phage population ecology, along with its

272

J. J. Dennehy and S. T. Abedon

relationship to phage community ecology and ecosystem ecology, is illustrated in Fig. 3). Such attributes are known as life history traits. In this section, we explore various phage life history traits and how phenotypic distinctions between phages of the same species can lead to differences in competitive ability, including faster phage population growth, greater virion durability, greater ease of exploiting hosts, or greater potential to resist extinction. See chapters ▶ “Adsorption: Phage Acquisition of Bacteria,” ▶ “Phage Infection and Lysis,” and ▶ “Temperate Phages, Prophages, and Lysogeny” for further consideration of phage phenotypes particularly as associated with various aspects of phage infection, survival, and population growth. Included in these discussions is the issue of tradeoffs, where improvement in one phenotypic trait in terms of its impact on organism functioning (particularly as affecting evolutionary fitness) has the effect of reducing organism functionality in

Interspecific Competition

Population Ecology

Exploitative Competition

Survival and Reproduction

Burst Size and Latent Period

Survival and Reproduction

Ecosystem Ecology

Recombination

Other Phage Species Antagonistic Interactions

Ecosystem Ecology

Abiotic Environment

Intraspecific Competition

Focus Phage

Ecophysiology Movement of Phages

Antagonistic Interactions

Bacteria (esp. host)

Nutrient Cycling and Energy Flow Lysogenic Conversion

Eukaryotes (esp. via exotoxins)

Community Ecology

Productive and Latent Cycles

Nutrient Cycling and Energy Flow

Phage-Induced Bacterial Lysis

Fig. 3 Relationships between different aspects of environments (pink and violet boxes), different aspects of phage ecology (yellow boxes), and various ecological phenomena (unboxed, brown text). A “Focus Phage” is simply that phage which is under primary consideration in a given study. Population ecology is covered over roughly the upper-left quadrant of the figure. Community ecology is covered over roughly the right half of the figure (including “Survival and Reproduction,” as also can be grouped under population ecology). Ecosystem ecology, particularly phage interaction with abiotic components of environments, is covered roughly in the lower-left quadrant of the figure. “Lysogenic Conversion” as grouped also under community ecology can also be associated with phage-mediated nutrient-generation within environments, i.e., as by phage-encoded exotoxins lysing eukaryotic cells (and hence “Lysogenic Conversion” is placed next to “Eukaryotes” in the figure). This figure is derived from that of Abedon (2009b)

Bacteriophage Ecology

273

terms of a different feature or context. For additional consideration of these sorts of issues, see (Abedon 2006; Bull 2006; Abedon 2008; Goldhill and Turner 2014).

Phage Life Cycles as Ecological Phenomena A successful organism must not only survive and reproduce but also do so better than other members of its population. These “struggles” on a population scale result in a refining of organism adaptations, which are products of mutation or recombination as acted upon by natural selection. Evolutionary ecology is the study of these adaptations from both evolutionary and ecological perspectives, including how they improve an organism’s fit to its environment along with how further improvements in fitness might be constrained. The latter can be due to historical contingencies (i.e., what adaptations or mutations are already present) or biogeographical constraints (limitations on the ability of organisms to reach potentially favorable environments). It is possible to subdivide various adaptations into specific organismal properties, generally as impacting either reproduction or survival. For phages, these phenotypes can be measured as often multifactorial aspects of phage life cycles (Fig. 4). Organismal properties, however, largely must be determined experimentally as so-called emergent properties, i.e., ones that can be difficult to predict simply from underlying biochemical aspects. In addition, often these properties will vary as a function of environmental conditions. Such phage organismal phenotypes include virion durability, host range, latent period, and burst size (chapter ▶ “Phage Infection and Lysis”). In addition, there are various aspects of phage lysogenic cycles (chapter ▶ “Temperate Phages, Prophages, and Lysogeny”). As these phenomena can occur within the context of phage life cycles, they generally can be studied from a perspective of phage population ecology, including in terms of their contribution to a phage’s ability to compete with conspecific phages.

Virion Durability It likely is preferable for phage virions to display greater extracellular durability. There likely are diminishing returns, however, such that increases in virion durability may not always result in commensurate increases in phage success. Phage virion durability also is less important the faster phages are able to find new hosts to adsorb (chapter ▶ “Adsorption: Phage Acquisition of Bacteria”), since then the duration over which virion durability is able to contribute to phage fitness should be shorter. Consequently, the greater the density of adsorbable host bacteria that are available in a phage’s immediate environment, the less that any increases in virion durability may be relevant to a phage’s fitness, since the rate which phage virions can encounter new host bacteria will tend to be a function of how many new host bacteria are present within a given environment. Alternatively, the longer that a virion must wait before finding a host to adsorb, then likely the more relevant virion durability is for phage

274

J. J. Dennehy and S. T. Abedon

fitness. Indeed, some phages may persist for decades or longer as virions before finding a new host to infect (Breitbart and Rohwer 2005). For an earlier version of this latter idea, see Williams et al. (1987). It is not necessarily cost free for a phage to display increases in virion durability. Declines in phage burst size or perhaps decreases in virion adsorption rates, for example, could be seen as antagonistic pleiotropies (tradeoffs) that could be associated with increases in phage durability (Conley and Wood 1975; de Paepe and Taddei 2006). On the other hand, fewer progeny phages may need to be produced to achieve a given level of fitness – as measured in terms of progeny survival and subsequent reproduction – given the production of more durable but fewer phages vs. greater numbers of more fragile phages. That is, a phage’s “effective burst size,” which is the number of virions which survive to reproduce (below), should be a function not just of burst size but also of their post-release virion durability.

Evolutionary Feedback

Life-History Traits Adsorption Molecular biochemistry physiology genetics

virion diffusivity affinity for receptors attachment to host

Infection evasion of host defenses eclipse duration virion maturation

Ecological populations communities ecosystems

Effective Burst Size maturation rate host range virion durability

Environmental Feedback Fig. 4 Phage organismal characteristics are their life history characteristics. Indicated is how these traits are underlain by molecular characteristics on one side (to the left) and are the means by which phages interact with their environments, i.e., ecologically, on the other (right side). Various life history characteristics are indicated (center) where “effective burst size” refers to not just how many virions are produced per phage-infected bacterium (burst size) but also how many of these virions survive to successfully infect new bacteria. Burst size more strictly tends to be a function not just of virion maturation rates during infections but also durations of periods of virion maturation as well. Indicated also are various feedback mechanisms, i.e., as modify individual organisms phenotypically over ecological time scales (such as physiological adaptation in response to environmental changes or feedback) or instead as modify organisms genotypically over evolutionary time scales (e.g., as due to mutation or natural selection, representing evolutionary feedback). This figure has been redrawn from that of Abedon (2009b)

Bacteriophage Ecology

275

A possible instance of reduced adsorption ability given increased virion durability is seen with coliphage T4 (family Myoviridae). Here increased virion durability could be a consequence of phage tail-fiber “retraction,” as in turn is associated with decreased adsorbability (Conley and Wood 1975). In this case, though, this is a phenotypic plasticity rather than genetically fixed durable state and is a characteristic that may be observed given cold temperatures, suboptimal pHs, or environments that are relatively lacking in adsorption cofactors such as relatively low concentrations of monovalent cations or free tryptophan (Kutter et al. 1994). The suggestion is that greater virion durability in this case might be observed in environments in which host adsorption is less desirable (extracolonic environments), whereas greater adsorbability but also lower virion durability should be seen in environments where host adsorption is more desirable (colonic lumens). Perhaps similar to the T4 adsorption story, a phenotypic variant of phage ΦX174 (family Microviridae) can be generated by lysing phage-infected bacteria at cold (4 °C) temperatures (Bleichrodt and van Abkoude 1967). This variant, once released, is not able to adsorb to its host at this temperature, unlike ΦX174 virions lysed at 37 °C, which are able to adsorb to hosts at 4 °C. These cold-lysed variants are convertible to the more adsorbable form given incubation at 37 °C. The cold-lysed variant also resists inactivation at 65 °C better than the warm-lysed variant (thereby suggesting greater virion durability). Interestingly, the transition from cold-lysed to warm-lysed forms is constrained at low pH. The implication may be that these phages, given lysis outside of the body, especially given colder temperatures, might be predisposed to an adsorption-resisting state that perhaps is more durable, with this state then reversible particularly under body conditions of mammals (or birds). This phenomenon was demonstrated with a number of other “small” ssDNA phages (family Microviridae), but not with two tested ssRNA phages (family Leviviridae).

Virion Attachment Affinity The “job” of a phage virion is to deliver its virus-genome payload to an infectionpermissive bacterium (chapter ▶ “Adsorption: Phage Acquisition of Bacteria”). Phage affinity for these bacteria impacts the rate of phage adsorption by affecting the likelihood that a given phage encounter with a host cell will result in attachment (Stent 1963; Abedon 2020a). Thus, in principle, a phage that adsorbs with 100% likelihood given encounter with a host cell will, on average, adsorb twice as fast as a phage that adsorbs with only 50% likelihood (this is so particularly given a wellmixed environment in which phages become rapidly separated from adsorbable cells should attachment fail). In an environment containing sufficiently high host densities, however, small increases in phage affinity may be of diminishing utility as a desorbing virion may be able to quickly find another bacterium to infect (Shao and Wang 2008; Abedon 2009a). Host attachment affinity represents only part of the phage attachment affinity story. Phages not only interact with their receptors on host cells but also may interact with extracellular nonhost entities (Dennehy et al. 2007). These nonhosts can be

276

J. J. Dennehy and S. T. Abedon

infection-impermissive bacteria or other components of the biotic and abiotic environment, such as animal tissues or soils to which virions might attach (Schijven and Hassanizadeh 2000). Therefore, phages ideally would display high affinities for specific host-cell receptor molecules and low affinities for nonhost molecules, as binding the latter would presumably preclude or at least delay binding to phagepermissive hosts. Note though that an exception to this rule may exist in cases where binding to nonhost molecules aids phage persistence in a favored habitat, e.g., phage binding to mucus to maintain a presence in association with animal tissues (Barr et al. 2013). Too high host specificity, however, can come with a cost. A high specificity to a particular host may result in a narrower adsorption range such that other potentially permissive hosts will not become phage infected (note: this tradeoff is another example of an antagonistic pleiotropy). The fewer potential hosts that a phage can adsorb, then the lower the densities of hosts available to a phage, either within a given environment or across multiple environments, and therefore the longer the delays between host adsorptions (Chan and Abedon 2012). Longer delays between adsorption events would tend to increase requirements for phage virion durability as well as otherwise extend phage generation times. Further complicating these issues, higher phage affinity for specific hosts, even if adsorption to those hosts normally would be desirable, is not necessarily always desirable. For example, in spatially structured environments, such as phage plaques (chapter ▶ “Detection of Bacteriophages: Phage Plaques”) or biofilms (chapter ▶ “Biofilm Applications of Bacteriophages”), reduced host affinity may be beneficial because it allows phages to diffuse further before adsorbing to host bacteria (Gallet et al. 2009; Roychoudhury et al. 2014), thereby potentially reducing the likelihood of multiple phage adsorptions to individual bacteria (Abedon 2017b). A related phenomenon can be described as an avoidance of the effects of bacteria “selfshading” (Boots and Mealor 2007; Tromas et al. 2014), i.e., in which already phageinfected bacteria or perhaps instead the debris of lysed bacteria (Rabinovitch et al. 2003; Aviram and Rabinovitch 2008; Bull et al. 2018) may prevent passage of other virions by them by serving as what can be described as sorptive scavengers (Abedon 2017b; Abedon 2020b). Indeed, lower attachment affinity may increase the potential for phages to leave biofilms altogether toward acquiring new biofilms to exploit. Thus, while greater virion affinity for host bacteria to a first approximation may seem as though it would always be desirable, in fact there are a number of counter arguments to the universal applicability of such a claim.

Evasion of Host Defenses Bacteria very commonly display antiphage defenses (Hyman and Abedon 2010; Labrie et al. 2010; Dy et al. 2014; Doron et al. 2018; Azam and Tanji 2019) (chapter ▶ “Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy”), and an ability to avoid being inactivated by these mechanisms is of obvious benefit to a phage. Such avoidance ability, however, likely does not come without cost. For example, specific phage adaptations that contribute to evasion of host defenses could require that phages encode additional genes, which could result in

Bacteriophage Ecology

277

metabolic burdens and/or require larger genomes (Maxwell 2016; Pawluk et al. 2018). If new genes are not required, nevertheless an ability to evade specific host defenses mutationally could possibly come at the cost of an ability to evade other host defenses (Zhang and Buckling 2011), which in turn could represent another example of an antagonistic pleiotropy. Changes in phage affinity for receptor molecules found on the surfaces of host bacteria, as one sees in the course of phage host-range switches, also is a situation in which phage adaptation toward successful infection of one otherwise resistant host can come at the expense of successful phage infection of other hosts (Ford et al. 2014). A phage specifically could incur burst size, latent period, or adsorption rate costs that are associated with bypassing host defenses. For example, a phage that displays a relatively wide host range – implying an ability to infect hosts displaying a diversity of active or passive antiphage defenses – may incur costs associated with being a generalist (Poullain et al. 2008; Hall et al. 2011). That is, adaptation to infecting multiple host types could come at the expense of an ability to more effectively infect any one specific host type such as might result in a reduced phage burst size while infecting any given host (which can be viewed as yet another example of an antagonistic pleiotropy). More narrowly, phage acquisition of mechanisms that allow for infection of any one host, such as mutations that permit evasion of specific host restriction endonuclease defenses, could result in fitness costs due to otherwise useful nucleotides being replaced. Not all mutations are purely beneficial or even purely neutral in terms of their impact on host fitness, though nonetheless net fitness gains should still occur in association with evasion of restriction endonuclease action due simply to phage survival upon infection of expressing hosts. Bacteria, by contrast, should be under frequency-dependent selection for diversity in their antiphage defenses. Specifically, different bacteria, even within the same bacterial species, ideally would display different levels of resistance to different phage types (Levin 1988). Thus, no one phage should be able to wipe out an entire bacterial species given a diversity of bacteria-encoded phage-resistance mechanisms, and the very process of a phage replicating at the expense of a bacterial population will select for those bacterial variants that are resistant to that phage (giving rise to a concept often described as killing the winner, as considered in greater detail below). As no single phage type is likely able to keep up with the simultaneous evolution by multiple potential hosts, and given costs associated with the display of specific mechanisms of phage evasion of host defenses, the result, seemingly, is a tendency for many phages to display relatively narrow host ranges, i.e., less than full coverage of individual bacterial strains found within a given species, much less abilities to infect multiple species (Ross et al. 2016; Hyman 2019).

Phage Latent Periods Latent periods, as a measure of the infection period of phage lytic infections, can vary as a function of allelic differences in a single gene. This life history characteristic is thereby arguably more a directly molecular rather than organismal property of

278

J. J. Dennehy and S. T. Abedon

a phage infection (Young 2014). Lysis timing, however, is still not predictable from gene sequence alone (Wang 2006), plus this timing varies with environmental conditions (Hadas et al. 1997). Furthermore, transplantation of lysis-timing genes between phages does not result in identical infection durations (Zheng et al. 2008). Factors such as cell size, cell physiological state, and membrane characteristics may also play a role in determining lysis timing (JJD, unpublished data). Lysis timing likely is therefore determined by a more complicated route than the expression of just one gene (chapter ▶ “Phage Infection and Lysis”). In addition, phage latent period in some cases is modifiable in the course of infection, resulting particularly in a phenomenon of lysis inhibition as seen especially in certain larger phages (Abedon 2019), or in deviations away from a direct path toward lysis as seen with lysogeny (Howard-Varona et al. 2017). Thus, latent period may be considered as a relatively complex, emergent, organismal property. Lastly, phage latent period directly impacts phage generation time, along with phage per infection productivity (next paragraph). Shorter phage latent periods will result in shorter phage generation times as a phage virion’s infection cycle consists of a combination of its search for a new host to infect and then the resulting infection. Barring tradeoffs, shorter latent periods therefore should be more beneficial to a phage than longer latent periods. Shorter phage latent periods, however, can come at the expense of phage burst size, as yet another example of an antagonistic pleiotropy. This is particularly to the extent that shorter latent periods are attained via reductions in the length of the period of progeny phage maturation, as is typically the case. Shorter latent periods nevertheless should still evolve under circumstances where the benefits associated with displaying shorter generation times outweigh the costs of burst size reductions, as likely is seen especially at higher host densities, when the rates at which phages acquire new bacteria can be substantially extended in relative terms by only small increases in phage latent periods, e.g., 30 min vs. 20 min (Abedon et al. 2001, 2003). There also exist circumstances in which longer latent periods, or more generally longer infections periods, likely are more beneficial. Longer latent periods presumably are favored under circumstances in which free phages, that is, extracellular phages, are more vulnerable to inactivation (Abedon 1990) and/or less capable of replicating than phages that are currently infecting bacteria. Such circumstances include when lower host densities extend the time of virion extracellular searches, i.e., so-called hard times (Stewart and Levin 1984). Examples of circumstances in which latent period extension may be so selected include the above-noted lysogeny (Erez et al. 2017; Abedon 2017a) and lysis inhibition (Abedon 2019).

Phage Burst Size There is an expectation that larger phage burst sizes will generally be more beneficial to phages than smaller ones. Larger burst sizes can, as noted, require longer latent periods to generate, which in turn will lead to longer phage generation times. Latent period (1) nevertheless is only one of three key determinants of phage burst size (chapter ▶ “Phage Infection and Lysis”), with the others being (2) the length of

Bacteriophage Ecology

279

eclipse period (i.e., the period elapsing between the entry of the phage’s genome into the host cell and the appearance of the first assembled progeny virion within the phage-infected bacterium), and (3) the progeny accumulation rate once the first progeny virion has been assembled. That is, generally burst size should be equal to the rate of post-eclipse virion maturation multiplied by the difference between the duration of a phage’s latent period and its eclipse period. The determinants of eclipse length as well as post-eclipse rates of phage accumulation are multifactorial, however, and little understood. Effective burst size (Abedon and Thomas-Abedon 2010; Chan and Abedon 2012; Abedon 2020c) is a concept that is somewhat equivalent to the basic reproductive number (R0) from epidemiology. Basic reproductive number is the number of new infections that occur, on average, starting from previous, single infections. For phages, if all released virions were to initiate new infections – as can be the case given low multiplicity broth growth under favorable conditions – then the phage burst size and the phage basic reproductive number would be essentially the same. Given inactivation of phages either during their search for new bacteria to infect or in the course of adsorbing to otherwise non-phage-permissive bacteria, then the basic reproductive number would decline from that of the original burst size, indeed likely by a lot in the real world. Phages also can be lost to adsorption of bacteria already infected by the same phage type, an issue that might be especially relevant during phage replication in association with bacterial biofilms (Abedon 2017b). Thus, the number of phages from a single phage-infected bacterium which survive to give rise to infections of their own, i.e., the effective burst size, would decline to less than the original burst size. The number of phages which survive to reproduce per infection (effective burst size) thus should be a function of (1) a phage’s latent period, (2) its eclipse period, and (3) rates of virion accumulation post the eclipse period and prior to bacterial lysis, i.e., collectively the phage’s actual burst size but also (4) the durability of the produced virion particles and (5) the length of time between virion release from phage infections and when those virions find new bacteria to infect.

Phage Community Ecology Interactions between phages and bacteria may be studied within a context of what can be described as phage-bacterial community ecology (Fig. 3). That is, these relationships are interspecific where at a minimum they involve a phage on the one hand (one species) and its bacterial host on the other (the other species), and groups of organisms consisting of more than one species are described as communities. In terms of phage-bacterial community ecology, one can consider how prophages can increase bacterial fitness due to lysogenic conversion (chapter ▶ “Temperate Phages, Prophages, and Lysogeny”) or how bacteria can be modified due to phage-mediated horizontal gene transfer, i.e., transduction (chapter ▶ “Bacteriophage-Mediated Horizontal Gene Transfer: Transduction”). Crucial also to phage-bacterial community interactions is what is described as antagonistic coevolution where bacterial mechanisms of phage resistance are overcome by phage host-

280

J. J. Dennehy and S. T. Abedon

range evolution in response to that resistance, which in turn motivates further bacterial evolution of resistance, and so on (chapter ▶ “Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy”). As these various issues are addressed in detail in other chapters of this volume, in this section, we focus instead on various ecological issues that stem from the phage role as exploiters of bacteria within environments. These issues include just how to label the specific ecological role of phages in environments (i.e., predators, parasites, or parasitoids?), the concept of kill the winner or killing the winner, and phage exploitation of bacteria as they are found in biofilms (chapter ▶ “Biofilm Applications of Bacteriophages”). Note that these various issues are closely tied as well to phage therapy, as an applied ecology phenomenon, and especially to phage therapy pharmacology (chapter ▶ “Bacteriophage Pharmacology and Immunology”).

Predators, Parasites, or Parasitoids? Part of community ecology is understanding the nature and utilities of direct interactions between different species such as are seen in predator-prey or parasitehost relationships. A parasite is an organism that lives on or in another organism and that benefits from its host’s resources, particularly at the host’s expense. Predators also harm other organisms. For predators, however, the death of its victim is beneficial, or at least not harmful to the predator, and furthermore that death, in a physiological sense, typically will take place prior to the predator achieving that benefit. Parasites instead usually benefit from the ongoing survival of their host or at least can accrue benefits as long as the parasite continues to survive in association with a metabolically intact host (vs., e.g., immune system-mediated elimination of the parasite). A parasitoid is an organism that behaves like a parasite but, like a predator, ultimately kills its host as part of its normal life cycle. Despite these differences in underlying characteristics, bacteriophages have been variously described as predators, parasites, or parasitoids. A case has been made, however, that lytic phages act as parasitoids during their lytic cycles, rather than as strictly parasites or predators (Forde et al. 2004). This perspective seems reasonable given that the productive cycles of lytic phages involve ongoing phage presence within a physiologically still-living host, but nevertheless that presence ultimately and intentionally ends with host death. Chronically infecting phages (chapter ▶ “Phage Infection and Lysis”), by contrast, are clearly parasitic since they appear neither to strive to kill their hosts nor directly benefit should that host death occur. These phages nonetheless place at least a metabolic burden on their hosts and therefore generally should be considered to be parasites rather than mutualists (the latter being symbiotic organisms whose presence benefits rather than harms their hosts). Chronically infecting phages, however, may nevertheless benefit from excessively exploiting their hosts, and this is even if such exploitation results in premature host demise if that should lead to greater rates of new-virion production, at least over the short term (Abedon 2006). Such excessive exploitation, though, does not imply that chronically infecting phages should necessarily be viewed as parasitoids.

Bacteriophage Ecology

281

Temperate phages vary between being parasitoids during their lytic cycles and at least potentially serving as mutualists during lysogenic cycles, the latter a consequence especially of lysogenic conversion (chapter ▶ “Temperate Phages, Prophages, and Lysogeny”). Notwithstanding the above arguments, lytic phages often are depicted at least metaphorically as predators of bacteria. There are at least three flaws in that reasoning, however. First, as noted, phages, unlike actual predators, do not physiologically kill bacterial “prey” prior to exploiting these hosts. Second, phages for the most part do not consume their hosts to the same extent as predators often do, that is, assimilate much of the host’s carbon and energy into phage “bodies” (Thingstad et al. 2008). Alternatively, however, phages might instead be described as “sloppy eaters,” that is, consumers which leave behind much of the consumed organism for exploitation by other organisms (vs. light eaters which leave much of the body of prey organisms intact for other organisms to consume). Third, unlike predators, the phage life cycle spans the exploitation of only a single prey individual, vs. predators that tend to exploit multiple prey individuals over the course of a single life span. Indeed, the fact that we habitually describe bacteria as phage “hosts” rather than as phage “prey” is suggestive that descriptions of phages as predators, rather than as parasitoids, parasites, or mutualists, do not necessarily seriously consider the reality, or at least the complexity, of phage-bacterial community ecological interactions.

Killing the Winner If a phage is acting as a bacterial predator, parasite, or parasitoid, then generally it is acting as an exploiter of bacterial victims. At a minimum, such activity has the effect of reducing the densities within environments of affected bacteria. Since phages can display substantial specificity in terms of which bacteria they impact, as due to narrow host ranges, then the presence of these exploiters within environments can result in an interspecific variation on frequency-dependent selection (Levin 1988). The result is a process often described as “kill the winner” or “killing the winner” (Thingstad 2000; Weinbauer 2004; Rodriguez-Brito et al. 2010; Winter et al. 2010; Diaz-Munoz and Koskella 2014; Jacquet et al. 2018). The idea of killing the winner is based on the population dynamics of lytically infecting phages and their bacterial hosts. Specifically, the more bacteria within an environment that become infected with a given phage type and the more metabolically active those bacteria, the – all else held constant – the more phage progeny of that type that will be produced. The more of these progeny phages that are produced, then the higher the resulting density of these specific phage types. Therefore, the greater the fraction of the bacteria that those phages are targeting that will become infected (again, holding all other variables constant). Since infections by especially obligately lytic phages are bactericidal, the expected result is a reduction in the densities of so-exposed phage-susceptible bacteria. “Winner bacteria” therefore are strains that are present prior to phage exposure at higher densities (and/or which display higher metabolic activities) and, given the presence of phages of suitable

282

J. J. Dennehy and S. T. Abedon

host range, the kill of “kill the winner” is a resulting phage-mediated decline in the densities of those bacteria. On a community-wide basis, the result can be greater bacterial diversity at least in terms of phage susceptibility or, equivalently, killing the winner can result in niches that tend to not be dominated by only one or a few bacterial strains. That is, rather than one or a few bacterial strains dominating an environment, or dominating an environmental niche, because phages will be expected to reduce the numbers of especially those bacteria that dominate numerically, there will be a potential for the diversity of bacteria within environments to be greater, at least in terms of phage susceptibility types. In addition, tradeoffs can exist in which phage resistance by bacteria is associated with reduced bacterial fitness costs, e.g., slower growth rates, such that in the presence of phages, phage-resistant bacteria may come to dominate environments, whereas in the absence of phages, environments instead may be dominated by otherwise phage-sensitive bacteria (Bohannan and Lenski 2000; Koskella et al. 2012). Lastly, the impact of killing phage populations is not forever since as their hosts decline in numbers, then through inevitable virion decay so too will the titers of these phages decline. Thus, winning bacteria beget winning phages while resulting “losing” bacteria beget as well “losing” phages.

Phage Exploitation of Biofilm-Associated Bacteria Ecologically, the dispersion of bacteria generally can either be random or clumped, the latter, aka, patchy. While random dispersion can correspond especially to wellmixed broth bacterial cultures, bacterial patches can consist instead of single-species biofilms, clonal bacterial microcolonies within a multispecies biofilm, localized single-species aggregates, or bacterial arrangements such as streptococci. In fact, in nature, bacteria are at least as likely, if not more so, to be found growing within biofilms, that is, as conglomerations of bacteria associated with bacteria-secreted extracellular polymeric substances and often as associated with surfaces (Flemming et al. 2016; Nadell et al. 2016). Consequently, many phage-bacterial interactions likely occur within the context of bacterial biofilms (chapter ▶ “Biofilm Applications of Bacteriophages”). Given such clustering of bacterial targets, the concept of phage productivity presumably becomes one of generating essentially what are localized phage epidemics, i.e., localized individual phage foci of infection. The more phage virions that are so produced per phage focus of infection, then the more phages which may then diffuse to new locations to equivalently initiate new localized epidemics (Abedon 2011, 2012b, 2020c). In comparing biofilm bacteria with planktonic bacteria, biofilm bacteria are likely less susceptible to phage attack. This reduced susceptibility may be due to greater bacterial physiological heterogeneity (i.e., fewer bacteria that are in the equivalent of exponential phase), to the presence of bacterially secreted extracellular polymeric substance (EPS) barriers, or to other factors that might impact phage susceptibility of biofilm bacteria vs. planktonic bacteria, e.g., differential display of receptor molecules (Sutherland et al. 2004; Abedon 2015, 2016, 2017b, 2020c). Though biofilm

Bacteriophage Ecology

283

bacteria may thus be less able to support the growth of phage populations due to reasons as mentioned above, we also note that there is no real understanding of the extent to which phage ability to exploit biofilm bacteria may be reduced relative to phage ability to exploit planktonic bacteria. In part this uncertainty stems from extensive evidence from the phage-mediated biocontrol of bacteria literature that phages in fact are able to exploit biofilm bacteria or at least substantially remove bacterial biofilms from surfaces (Abedon 2015, 2020c) (chapter ▶ “Biofilm Applications of Bacteriophages”). In addition, we note that many phages carry degradative enzymes, i.e., EPS depolymerases, that allow them to digest and penetrate bacterial EPS barriers (Sutherland et al. 2004; Azeredo and Sutherland 2008; Chan and Abedon 2015; Latka et al. 2017; Abedon 2011; Pires et al. 2016; Oliveira et al. 2018), which suggests that there are phages that might explicitly target biofilm-

(6c) Reversibly bound to environment-defining molecular pattern, ‘sitting and waiting' for bacterium (5) Diffusion to distant bacteria

Constrained diffusion

(6a) Bacterial microcolony somewhat distantly located from previous focus of infection

(6b) Planktonic bacterium

or

or

(1) Initial adsorption

Adsorption

(3) Diffusion to adjacent microcolony Bacterium

Adsorption

Adsorption

Adsorption (2) Withinmicrocolony substantial phage propagation

Killing the (local) winner

By chance diffusion path

Microcolony

(4) Diffusion to non-adjacent microcolony

Fig. 5 Phage-biofilm interactions across more-local and more-distant scales. The model combines that of Abedon (2012b), which is based on a model presented by Abedon (2011), and that of Barr et al. (2013), the latter which considers phage virion reversible interaction with mucus (see also Abedon (2020c) for a general overview of phage-biofilm ecological interactions). Phage exploitation of biofilms begins with phage adsorption to a single bacterium, which gives rise to new phages that can infect immediately adjacent as well as, though with less likelihood and rapidity, more distant bacteria found within the same biofilm. In addition, virions can diffuse out of their parental biofilm to “search” for new biofilms to exploit. The more virions produced locally in the course of exploiting a given biofilm, within a single “focus of infection,” then likely the more virions released into the extra-biofilm environment. Thereby, the more virions searching for new hosts, then the greater the potential for at least one of those virions to acquire a new biofilm to exploit. Bacterial microcolonies are assumed to consist of clonally related bacteria. This figure has been redrawn from that of Abedon (2015)

284

J. J. Dennehy and S. T. Abedon

forming bacteria. Indeed, Harper et al. (2014) report that 2-day-old biofilms can potentially support phage replication better than bacterial lawns can support phage plaque formation, though biofilms may become less susceptible to phages as they mature such as due to changes in bacterial physiology or EPS (Abedon 2016). At a minimum, it is likely, therefore, that phage exploitation of such non-planktonic bacteria is at least more complex than phage exploitation of planktonic bacteria, and more challenging to both phages and researchers as well. Given these various ideas, we can envisage a phage “life cycle” in association with phage-susceptible biofilm bacteria that consists of long searches by phage virions for bacteria to infect, with these searches followed by local “epidemics” of phage replication. Since phage virions are not infinitely durable, we can expect potentially substantial phage attrition during these searches, i.e., virion inactivation prior to finding new patches of bacteria to exploit. Thus, an alternation between phage “feast” and “famine” can be expected, with localized bacterial exploitation (“feast”) ideally producing sufficient phage numbers that at least one phage will survive their passage between bacterial patches (“famine”) to succeed in locating a new bacterial patch to exploit (“feast”). This scenario conceptually is little different from phage exploitation of planktonic bacteria, with “feast” associated with bacterial infection and “famine” associated with movement between bacteria. In the case of exploitation of clustered bacteria, however, the “feast” will be potentially richer, as well as more complex, involving multiple bacterial infections, whereas the “famine” for a given overall environmental density of bacteria may be greater due to greater distances between biofilms (Abedon 2012a, 2017b, 2020c). See Fig. 5 for summary of this basic ecological scenario of phage-biofilm ecological interactions.

Phage Ecosystem Ecology Ecosystem ecology is the study of nutrient cycling within environments (i.e., within ecosystems) and energy flow through environments. Other than in terms of their impact directly on host bacteria, lytic phages impact environments primarily through releasing nutrient – and chemical energy – via their partial decomposition of infected bacteria. This works in three ways. First, there is leakage of the cytoplasmic contents into the extracellular environment, including of released phage virions. Second is the decomposition or disruption of bacterial polymers by phage enzymes in the course of adsorption, infection, and subsequent lysis. Most notable is the partial degradation of bacterial membrane and peptidoglycan (Young 2014). Many phages degrade the bacterial genome as well, mining this DNA for nucleotides, though these are not necessarily fully exploited prior to lysis (Carlson et al. 1994). Last, the killing as well as partial decomposition of bacteria upon lysis has the effect of making the bacterium’s constituent molecules more susceptible to other environmental agents such as exoenzymes (Nelson et al. 2012; Rodriguez-Brito et al. 2010). From the perspective of nutrient release, we can view this phage impact as somewhat equivalent to the as-noted “sloppy eating.” That is, phages incorporate only a fraction of the infected bacterium into new phage particles, with the rest of the

Bacteriophage Ecology

285

Consumers of Protists (Small Animals) ML

Lysis & Partial Decomposition

Cellular Respiration

Bacterial Grazers (Protists) ML

Heterotrophic Bacteria Viral Shunt (VS)

2’ Productivity

Microbial Loop (ML)

ML

Dissolved Organic Carbon (DOC) VS

Autotrophs (e.g., algae, cyanobacteria) 1’ Productivity

Sunshine & Other Abiotic Energy Sources

Carbon Dioxide

Fig. 6 Microbial loop or microbial food web refers to food chains or webs, that are based at least in part on the assimilation of dissolved organic carbon (DOC) by heterotrophic bacteria (dashed orange arrows). DOC is generated in the course of organism decomposition, with some of that decomposition a result of virus-induced cellular lysis. The latter is described as what is known as the virus shunt (dashed red arrows). This DOC is then available to heterotrophic bacteria. Thus, the viral shunt consists of viral-induced lysis of bacteria (including cyanobacteria but also archaea, algae, and protists), which has the effect of short-circuiting the microbial loop by converting easily engulfed cells into not easily engulfed acellular molecules, molecules that instead have now become available especially to phage-susceptible heterotrophic bacteria

bacterium then becoming available to other organisms. Upon conversion of lysed bacteria into truly soluble materials, the resulting DOC (for dissolved organic carbon) can be “recycled” back into other heterotrophic bacteria. Phage infection thus can have the perverse effect of benefitting physiologically the very bacterial communities that phages are exploiting. A consequence of this recycling is that material within ecosystems that is acquired by bacteria to a degree can become trophically trapped within bacterial communities (a phenomenon commonly known as the virus shunt; see Fig. 6) rather than becoming more readily assimilated into higher trophic levels, such as to protists or animal predators of bacteria (Wilhelm and Suttle 1999; Kuzyakov and Mason-Jones 2018). With soils or sediments, the primary ecological function is one of decomposition. That is, organisms as well as parts of organisms die and become associated as dead bodies or parts of bodies with these environments (organisms also can die outside of or above these environments then fall to soils from above, e.g., as from trees, or fall to sediments from the water column). These soil- or sediment-associated dead organisms or their parts are then consumed by decomposers including by heterotrophic bacteria. The heterotrophic bacteria can serve as hosts for lytic phages, which lyse the bacteria, thereby releasing DOC, which is consumed by heterotrophic bacteria. The function of phages in terms of these trophic interactions is to contribute to the overall trophic function of these environments, one of decomposition, though

286

J. J. Dennehy and S. T. Abedon

because of phages there are likely fewer nutrients and less energy available to organisms other than heterotrophic bacteria or, alternatively, other than non-phagotrophic protists or fungi (Jacquet et al. 2018). Within aquatic water columns the same processes occur, but with a twist. Specifically, cyanobacteria contribute to primary productivity, that is, the conversion of sunlight and carbon dioxide into organic carbon. Through trophic interactions, in which consumer organisms literally consume other organisms, the carbon and energy associated with organic carbon is able to move within and through ecosystems (respectively). The greater the overall primary productivity, that is, the greater the amount of organic carbon produced by these primary producer organisms, then the greater numbers of consumers that an environment can sustain. By lysing cyanobacteria, phages “short circuit” the movement of organic carbon and energy up food chains, contributing to the conversion of cyanobacteria into DOC, which is then available to phytoplankton (Weinbauer 2004; Peduzzi 2016; Kuzyakov and Mason-Jones 2018). Since energy conversions are always less than 100% efficient, the acquisition of primary productivity by higher trophic levels is less efficient in the presence of cyanophages than it is in their absence. See Fig. 6 for illustration.

Summary Though the division of biology into molecular vs. ecological is artificial, the distinction nevertheless exists in practice between the study of organisms as whole entities vs. the study of their genes or proteins, i.e., study of their molecules. As microorganisms are normally found in situ – even if that in situ is a component of another organism’s microbiome – the vast majority of microbiological phenomena possess an ecological component, that is, representing some sort of interaction, either directly or indirectly, with some environmental component that is external to the organism itself. Indeed, except for certain housekeeping functions that are relevant potentially only internally to microorganisms (e.g., such as the function of singlestranded DNA binding proteins), nearly all molecular aspects of microorganisms, except under the most artificial of circumstances, or when considered purely from the standpoint of bioengineering, either affect or are affected by environments, and therefore have some relevance to ecology. This is true even if the ecological aspects are not explicitly considered. Though it is possible to do substantial amounts of biology without taking ecology into account, that does not lessen the relevance of ecological thinking. Indeed, ignorance of the ecological context of organism evolutionary adaptations, all of which have underlying molecular bases, can hinder even molecular approaches to understanding organism biology. Though phage ecology is a vast subject, and we have only superficially covered its breadth and depth, here we have provided some of that ecological context that can be relevant to better understanding phages as molecular entities, or as more than just molecular entities. Such improved ecological understanding also has relevance toward better understanding phages as more

Bacteriophage Ecology

287

applied yet still ecological entities, such as in the context of phage use as antibacterial agents (phage therapy).

Cross-References ▶ Adsorption: Phage Acquisition of Bacteria ▶ Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy ▶ Bacteriophage Pharmacology and Immunology ▶ Bacteriophage-Mediated Horizontal Gene Transfer: Transduction ▶ Biofilm Applications of Bacteriophages ▶ Detection of Bacteriophages: Phage Plaques ▶ Phage Infection and Lysis ▶ Temperate Phages, Prophages, and Lysogeny

References Abedon ST (1990) Selection for lysis inhibition in bacteriophage. J Theor Biol 146:501–511 Abedon ST (2006) Phage ecology. In: Calendar R, Abedon ST (eds) The bacteriophages. Oxford University Press, Oxford, pp 37–46 Abedon ST (2008) Ecology of viruses infecting bacteria. In: Mahy BWJ, Van Regenmortel MHV (eds) Encyclopedia of virology, 3rd edn. Elsevier, Oxford, pp 71–77 Abedon ST (2009a) Kinetics of phage-mediated biocontrol of bacteria. Foodborne Pathog Dis 6:807–815 Abedon ST (2009b) Phage evolution and ecology. Adv Appl Microbiol 67:1–45 Abedon ST (2011) Bacteriophages and biofilms: ecology, phage therapy, plaques. Nova Science Publishers, Hauppauge Abedon ST (2012a) Spatial vulnerability: bacterial arrangements, microcolonies, and biofilms as responses to low rather than high phage densities. Viruses 4:663–687 Abedon ST (2012b) Thinking about microcolonies as phage targets. Bacteriophage 2:200–204 Abedon ST (2015) Ecology of anti-biofilm agents II. Bacteriophage exploitation and biocontrol of biofilm bacteria. Pharmaceuticals (Basel) 8:559–589 Abedon ST (2016) Bacteriophage exploitation of bacterial biofilms: phage preference for less mature targets? FEMS Microbiol Lett 363:fnv246 Abedon ST (2017a) Commentary: communication between viruses guides lysis-lysogeny decisions. Front Microbiol 8:983 Abedon ST (2017b) Phage "delay" towards enhancing bacterial escape from biofilms: a more comprehensive way of viewing resistance to bacteriophages. AIMS Microbiol 3:186–226 Abedon ST (2018) Phage therapy: various perspectives on how to improve the art. Meth Mol Biol 1734:113–127 Abedon ST (2019) Look who's talking: T-even phage lysis inhibition, the granddaddy of virus-virus intercellular communication research. Viruses 11(10):E951 Abedon ST (2020a) Phage therapy: killing titers, multiplicity of infection, adsorption theory, and passive versus active treatments. In: Kurtboke DI, Aminov R (eds) Advances on the applications of bacteriophages. Nova Science Publishers, Hauppauge Abedon ST (2020b) Phage-phage, phage-bacteria, and phage-environment communication. In: Witzany G (ed) Biocommunication of phages. Springer, New York Abedon S.T. (2020c) Bacteriophage-mediated biocontrol of wound infections, and ecological exploitation of biofilms by phages. In: Shiffman M, Low M. (eds) Biofilm, Pilonidal Cysts

288

J. J. Dennehy and S. T. Abedon

and Sinuses. Recent Clinical Techniques, Results, and Research in Wounds, vol 1. Springer, Cham, pp 121–158 Abedon ST, Thomas-Abedon C (2010) Phage therapy pharmacology. Curr Pharm Biotechnol 11:28–47 Abedon ST, Herschler TD, Stopar D (2001) Bacteriophage latent-period evolution as a response to resource availability. Appl Environ Microbiol 67:4233–4241 Abedon ST, Hyman P, Thomas C (2003) Experimental examination of bacteriophage latent-period evolution as a response to bacterial availability. Appl Environ Microbiol 69:7499–7506 Akhter S, Aziz RK, Edwards RA (2012) PhiSpy: a novel algorithm for finding prophages in bacterial genomes that combines similarity- and composition-based strategies. Nucl Acids Res 40:e126 Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local alignment search tool. J Mol Biol 215:403–410 Angly F, Rodriguez-Brito B, Bangor D, McNairnie P, Breitbart M, Salamon P, Felts B, Nulton J, Mahaffy J, Rohwer F (2005) PHACCS, an online tool for estimating the structure and diversity of uncultured viral communities using metagenomic information. BMC Bioinfor 6:41 Angly FE, Felts B, Breitbart M, Salamon P, Edwards RA, Carlson C, Chan AM, Haynes M, Kelley S, Liu H, Mahaffy JM, Mueller JE, Nulton J, Olson R, Parsons R, Rayhawk S, Suttle CA, Rohwer F (2006) The marine viromes of four oceanic regions. PLoS Biol 4:e368 Angly FE, Willner D, Prieto-Davo A, Edwards RA, Schmieder R, Vega-Thurber R, Antonopoulos DA, Barott K, Cottrell MT, Desnues C, Dinsdale EA, Furlan M, Haynes M, Henn MR, Hu Y, Kirchman DL, McDole T, McPherson JD, Meyer F, Miller RM, Mundt E, Naviaux RK, Rodriguez-Mueller B, Stevens R, Wegley L, Zhang L, Zhu B, Rohwer F (2009) The GAAS metagenomic tool and its estimations of viral and microbial average genome size in four major biomes. PLoS Comput Biol 5:e1000593 Aviram I, Rabinovitch A (2008) Dynamical types of bacteria and bacteriophages interaction: shielding by debris. J Theor Biol 251:121–136 Azam AH, Tanji Y (2019) Bacteriophage-host arm race: an update on the mechanism of phage resistance in bacteria and revenge of the phage with the perspective for phage therapy. Appl Microbiol Biotechnol 103:2121–2131 Azeredo J, Sutherland IW (2008) The use of phages for the removal of infectious biofilms. Curr Pharm Biotechnol 9:261–266 Barr JJ (2017) A bacteriophages journey through the human body. Immunol Rev 279:106–122 Barr JJ, Auro R, Furlan M, Whiteson KL, Erb ML, Pogliano J, Stotland A, Wolkowicz R, Cutting AS, Doran KS, Salamon P, Youle M, Rohwer F (2013) Bacteriophage adhering to mucus provide a non-host-derived immunity. Proc Natl Acad Sci U S A 110:10771–10776 Bergh O, Børsheim KY, Bratbak G, Heldal M (1989) High abundance of viruses found in aquatic environments. Nature (London) 340:467–468 Bettarel Y, Sime-Ngando T, Amblard C, Laveran H (2000) A comparison of methods for counting viruses in aquatic systems. Appl Environ Microbiol 66:2283–2289 Bleichrodt JF, van Abkoude ER (1967) The transition between two forms of bacteriophage phi-X174 differing in heat sensitivity and adsorption characteristics. Virology 32:93–102 Bohannan BJM, Lenski RE (2000) Linking genetic change to community evolution: insights from studies of bacteria and bacteriophage. Ecol Lett 3:362–377 Boots M, Mealor M (2007) Local interactions select for lower pathogen infectivity. Science (New York, N Y ) 315:1284–1286 Bradley DE, Sirgel FA, Coetzee JN, Hedges RW, Coetzee WF (1982) Phages C-2 and J: IncC and IncJ plasmid-dependent phages, respectively. J Gen Microbiol 128 (Pt 10):2485–2498 Breitbart M, Rohwer F (2005) Here a virus, there a virus, everywhere the same virus? Trends Microbiol 13:278–284 Breitbart M, Salamon P, Andresen B, Mahaffy JM, Segall AM, Mead D, Azam F, Rohwer F (2002) Genomic analysis of uncultured marine viral communities. Proc Natl Acad Sci U S A 99: 14250–14255

Bacteriophage Ecology

289

Breitbart M, Felts B, Kelley S, Mahaffy JM, Nulton J, Salamon P, Rohwer F (2004a) Diversity and population structure of a near-shore marine sediment viral community. Proc R Soc Lond B Biol Sci 271:565–574 Breitbart M, Miyake JH, Rohwer F (2004b) Global distribution of nearly identical phage-encoded DNA sequences. FEMS Microbiol Lett 236:249–256 Bull JJ (2006) Optimality models of phage life history and parallels in disease evolution. J Theor Biol 241:928–938 Bull JJ, Christensen KA, Scott C, Jack BR, Crandall CJ, Krone SM (2018) Phage-bacterial dynamics with spatial structure: self organization around phage sinks can promote increased cell densities. Antibiotics (Basel) 7(1):8 Cairns J, Stent G, Watson JD (1966) Phage and the origins of molecular biology. Cold Spring Harbor Laboratory Press, Cold Spring Harbor Campbell A (1961) Conditions for the existence of bacteriophages. Evolution; Int J Organic Evolut 15:153–165 Carlson K, Raleigh EA, Hattman S (1994) Restriction and modification. In: Karam JD, Kutter E, Carlson K, Guttman B (eds) Molecular biology of bacteriophage T4. ASM Press, Washington, D.C., pp 369–370 Casas V, Rohwer F (2007) Phage metagenomics. Meth Enzymol 421:259–268 Ceyssens PJ, Miroshnikov K, Mattheus W, Krylov V, Robben J, Noben JP, Vanderschraeghe S, Sykilinda N, Kropinski AM, Volckaert G, Mesyanzhinov V, Lavigne R (2009) Comparative analysis of the widespread and conserved PB1-like viruses infecting Pseudomonas aeruginosa. Environ Microbiol 11:2874–2883 Chan BK, Abedon ST (2012) Bacteriophage adaptation, with particular attention to issues of phage host range. In: Quiberoni A, Reinheimer J (eds) Bacteriophages in dairy processing. Nova Science Publishers, Hauppauge, pp 25–52 Chan BK, Abedon ST (2015) Bacteriophages and their enzymes in biofilm control. Curr Pharm Des 21:85–99 Chao L, Levin BR, Stewart FM (1977) A complex community in a simple habitat: an experimental study with bacteria and phage. Ecology 58:369–378 Chatterjee A, Duerkop BA (2018) Beyond bacteria: bacteriophage-eukaryotic host interactions reveal emerging paradigms of health and disease. Front Microbiol 9:1394 Chow CE, Suttle CA (2015) Biogeography of viruses in the sea. Annu Rev Virol 2:41–66 Cobian Guemes AG, Youle M, Cantu VA, Felts B, Nulton J, Rohwer F (2016) Viruses as winners in the game of life. Annu Rev Virol 3:197–214 Comeau AM, Chan AM, Suttle CA (2006) Genetic richness of vibriophages isolated in a coastal environment. Environ Microbiol 8:1164–1176 Conley MP, Wood WB (1975) Bacteriophage T4 whiskers: a rudimentary environment-sensing device. Proc Natl Acad Sci U S A 72:3701–3705 Dabrowska K, Abedon ST (2019) Pharmacologically aware phage therapy: pharmacodynamic and pharmacokinetic obstacles to phage antibacterial action in animal and human bodies. Microbiol Mol Biol Rev 83:e00012–e00019 Danovaro R, Dell'Anno A, Trucco A, Serresi M, Vanucci S (2001) Determination of virus abundance in marine sediments. Appl Environ Microbiol 67:1384–1387 de Paepe M, Taddei F (2006) Viruses' life history: towards a mechanistic basis of a trade-off between survival and reproduction among phages. PLoS Biol 4:e193 Dennehy JJ, Friedenberg NA, Yang YW, Turner PE (2007) Virus population extinction via ecological traps. Ecol Lett 10:230–240 d'Hérelle F, Smith GH (1926) The bacteriophage and its behavior. Williams & Wilkins Co, Baltimore Diaz-Munoz SL, Koskella B (2014) Bacteria-phage interactions in natural environments. Adv Appl Microbiol 89:135–183 Doron S, Melamed S, Ofir G, Leavitt A, Lopatina A, Keren M, Amitai G, Sorek R (2018) Systematic discovery of antiphage defense systems in the microbial pangenome. Science (New York, N Y ) 359(6379):eaar4120

290

J. J. Dennehy and S. T. Abedon

Dutilh BE, Cassman N, McNair K, Sanchez SE, Silva GG, Boling L, Barr JJ, Speth DR, Seguritan V, Aziz RK, Felts B, Dinsdale EA, Mokili JL, Edwards RA (2014) A highly abundant bacteriophage discovered in the unknown sequences of human faecal metagenomes. Nat Commun 5:4498 Dy RL, Richter C, Salmond GP, Fineran PC (2014) Remarkable mechanisms in microbes to resist phage infections. Annu Rev Virol 1:307–331 Edlund A, Santiago-Rodriguez TM, Boehm TK, Pride DT (2015) Bacteriophage and their potential roles in the human oral cavity. J Oral Microbiol 7:27423 Edwards RA, McNair K, Faust K, Raes J, Dutilh BE (2016) Computational approaches to predict bacteriophage-host relationships. FEMS Microbiol Rev 40:258–272 Eggleston EM, Hewson I (2016) Abundance of two Pelagibacter ubique bacteriophage genotypes along a latitudinal transect in the North and South Atlantic Oceans. Front Microbiol 7:1534 Erez Z, Steinberger-Levy I, Shamir M, Doron S, Stokar-Avihail A, Peleg Y, Melamed S, Leavitt A, Savidor A, Albeck S, Amitai G, Sorek R (2017) Communication between viruses guides lysislysogeny decisions. Nature (London) 541:488–493 Flemming HC, Wingender J, Szewzyk U, Steinberg P, Rice SA, Kjelleberg S (2016) Biofilms: an emergent form of bacterial life. Nat Rev Microbiol 14:563–575 Ford BE, Sun B, Carpino J, Chapler ES, Ching J, Choi Y, Jhun K, Kim JD, Lallos GG, Morgenstern R, Singh S, Theja S, Dennehy JJ (2014) Frequency and fitness consequences of bacteriophage Φ6 host range mutations. PLoS One 9:e113078 Forde SE, Thompson JN, Bohannan BJM (2004) Adaptation varies through space and time in a coevolving host–parasitoid interaction. Nature (London) 431:841–844 Forterre P, Soler N, Krupovic M, Marguet E, Ackermann H-W (2013) Fake virus particles generated by fluorescence microscopy. Trends Microbiol 21:1–5 Fuhrman JA (1999) Marine viruses and their biogeochemical and ecological effects. Nature (London) 399:541–548 Gallet R, Shao Y, Wang I-N (2009) High adsorption rate is detrimental to bacteriophage fitness in a biofilm-like environment. BMC Evol Biol 9:241 Goldhill DH, Turner PE (2014) The evolution of life history trade-offs in viruses. Curr Opin Virol 8C:79–84 Gonzalez MD, Lichtensteiger CA, Caughlan R, Vimr ER (2002) Conserved filamentous prophage in Escherichia coli O18:K1:H7 and Yersinia pestis biovar orientalis. J Bacteriol 184:6050–6055 Goyal SM, Gerba CP, Bitton G (1987) Phage ecology. CRC Press, Boca Raton Gregory AC, Zayed AA, Conceicao-Neto N, Temperton B, Bolduc B, Alberti A, Ardyna M, Arkhipova K, Carmichael M, Cruaud C, Dimier C, Dominguez-Huerta G, Ferland J, Kandels S, Liu Y, Marec C, Pesant S, Picheral M, Pisarev S, Poulain J, Tremblay JE, Vik D, Babin M, Bowler C, Culley AI, de VC, Dutilh BE, Iudicone D, Karp-Boss L, Roux S, Sunagawa S, Wincker P, Sullivan MB (2019) Marine DNA viral macro- and microdiversity from pole to pole. Cell 177:1109–1123 Greninger AL (2018) A decade of RNA virus metagenomics is (not) enough. Virus Res 244:218–229 Hadas H, Einav M, Fishov I, Zaritsky A (1997) Bacteriophage T4 development depends on the physiology of its host Escherichia coli. Microbiology 143:179–185 Hall AR, Scanlan PD, Morgan AD, Buckling A (2011) Host-parasite coevolutionary arms races give way to fluctuating selection. Ecol Lett 14:635–642 Hannigan GD, Meisel JS, Tyldsley AS, Zheng Q, Hodkinson BP, SanMiguel AJ, Minot S, Bushman FD, Grice EA (2015) The human skin double-stranded DNA virome: topographical and temporal diversity, genetic enrichment, and dynamic associations with the host microbiome. MBio 6(5):e01578–e01515 Hanson CA, Marston MF, Martiny JB (2016) Biogeographic variation in host range phenotypes and taxonomic composition of marine cyanophage isolates. Front Microbiol 7:983 Harper DR, Parracho HMR, Walker J, Sharp R, Hughes G, Werthrén M, Lehman S, Morales S (2014) Bacteriophages and biofilms. Antibiotics 3:270–284

Bacteriophage Ecology

291

Hayes S, Mahony J, Nauta A, van SD (2017) Metagenomic approaches to assess bacteriophages in various environmental niches. Viruses 9(6):E127 Horiuchi K, Adelberg EA (1965) Growth of male-specific bacteriophage in Proteus mirabilis harboring F-genotes derived from Escherichia coli. J Bacteriol 89:1231–1236 Howard-Varona C, Hargreaves KR, Abedon ST, Sullivan MB (2017) Lysogeny in nature: mechanisms, impact and ecology of temperate phages. ISME J 11:1511–1520 Huang S, Wilhelm SW, Jiao N, Chen F (2010) Ubiquitous cyanobacterial podoviruses in the global oceans unveiled through viral DNA polymerase gene sequences. ISME J 4:1243–1251 Huang S, Zhang S, Jiao N, Chen F (2015) Marine cyanophages demonstrate biogeographic patterns throughout the global ocean. Appl Environ Microbiol 81:441–452 Hurwitz BL, Sullivan MB (2013) The Pacific Ocean virome (POV): a marine viral metagenomic dataset and associated protein clusters for quantitative viral ecology. PLoS One 8:e57355 Hurwitz BL, U'Ren JM, Youens-Clark K (2016) Computational prospecting the great viral unknown. FEMS Microbiol Lett 363:fnw077 Hyman P (2019) Phages for phage therapy: isolation, characterization, and host range breadth. Pharmaceuticals (Basel) 12(1):35 Hyman P, Abedon ST (2010) Bacteriophage host range and bacterial resistance. Adv Appl Microbiol 70:217–248 Jacquet S, Zhong X, Peduzzi P, Thingstad TF, Parikka KJ, Weinbauer MG (2018) Virus interactions in the aquatic world. In: Hyman P, Abedon ST (eds) Viruses of microorganisms. Caister Academic Press, Norwich, pp 115–141 Kalatzis PG, Rorbo NI, Castillo D, Mauritzen JJ, Jorgensen J, Kokkari C, Zhang F, Katharios P, Middelboe M (2017) Stumbling across the same phage: comparative genomics of widespread temperate phages infecting the fish pathogen Vibrio anguillarum. Viruses 9 Kang I, Oh HM, Kang D, Cho JC (2013) Genome of a SAR116 bacteriophage shows the prevalence of this phage type in the oceans. Proc Natl Acad Sci U S A 110:12343–12348 Kim M-S, Bae J-W (2018) Lysogeny is prevalent and widely distributed in the murine gut microbiota. ISME J 12:1127–1141 Koskella B, Lin DM, Buckling A, Thompson JN (2012) The costs of evolving resistance in heterogeneous parasite environments. Proc Biol Sci 279:1896–1903 Kutter E, Kellenberger E, Carlson K, Eddy S, Neitzel J, Messinger L, North J, Guttman B (1994) Effects of bacterial growth conditions and physiology on T4 infection. In: Karam JD, Kutter E, Carlson K, Guttman B (eds) The molecular biology of bacteriophage T4. ASM Press, Washington, DC, pp 406–418 Kuzyakov Y, Mason-Jones K (2018) Viruses in soil: Nano-scale undead drivers of microbial life, biogeochemical turnover and ecosystem functions. Soil Biol Biochem 127:305–317 Labrie SJ, Samson JE, Moineau S (2010) Bacteriophage resistance mechanisms. Nat Rev Microbiol 8:317–327 Latka A, Maciejewska B, Majkowska-Skrobek G, Briers Y, Drulis-Kawa Z (2017) Bacteriophageencoded virion-associated enzymes to overcome the carbohydrate barriers during the infection process. Appl Microbiol Biotechnol 101:3103–3119 Levin BR (1988) Frequency-dependent selection in bacterial populations. Philos Trans R Soc Lond Ser B Biol Sci 319:459–472 Levin BR, Stewart FM, Chao L (1977) Resource limited growth, competition, and predation: a model and experimental studies with bacteria and bacteriophage. Am Nat 111:3–24 Lima-Mendez G, Van Helden J, Toussaint A, Leplae R (2008) Prophinder: a computational tool for prophage prediction in prokaryotic genomes. Bioinformatics (Oxford, England) 24:863–865 Lloyd KG, Steen AD, Ladau J, Yin J, Crosby L (2018) Phylogenetically novel uncultured microbial cells dominate Earth microbiomes. mSystems 3:e00055–e00018 Loftus A, Delisle AL (1995) Inducible bacteriophages of Actinobacillus actinomycetemcomitans. Curr Microbiol 30:317–321 Los JM, Los M, Wegrzyn A, Wegrzyn G (2008) Role of the bacteriophage lambda exo-xis region in the virus development. Folia Microbiol 53:443–450

292

J. J. Dennehy and S. T. Abedon

Mahmoud H, Jose L (2017) Phage and nucleocytoplasmic large viral sequences dominate coral viromes from the Arabian Gulf. Front Microbiol 8:2063 Manrique P, Bolduc B, Walk ST, van der Oost J, de Vos WM, Young MJ (2016) Healthy human gut phageome. Proc Natl Acad Sci U S A 113:10400–10405 Marhaver KL, Edwards RA, Rohwer F (2008) Viral communities associated with healthy and bleaching corals. Environ Microbiol 10:2277–2286 Marie D, Brussaard CPD, Thyrhaug G, Bratbak G, Vaulot D (1999) Enumeration of marine viruses in culture and natural samples by flow cytometry. Appl Environ Microbiol 65:45–52 Marston MF, Taylor S, Sme N, Parsons RJ, Noyes TJ, Martiny JB (2013) Marine cyanophages exhibit local and regional biogeography. Environ Microbiol 15:1452–1463 Maxwell KL (2016) Phages fight back: inactivation of the CRISPR-Cas bacterial immune system by anti-CRISPR proteins. PLoS Pathog 12:e1005282 Miller-Ensminger T, Garretto A, Brenner J, Thomas-White K, Zambom A, Wolfe AJ, Putonti C (2018) Bacteriophages of the urinary microbiome. J Bacteriol 200:e00738–e00717 Minot S, Sinha R, Chen J, Li H, Keilbaugh SA, Wu GD, Lewis JD, Bushman FD (2011) The human gut virome: inter-individual variation and dynamic response to diet. Genome Res 21:1616–1625 Mokili JL, Rohwer F, Dutilh BE (2012) Metagenomics and future perspectives in virus discovery. Curr Opin Virol 2:63–77 Molnar DM, Lawton WD (1971) Growth of male-specific bacteriophage in Pasteurella harboring F-genotes derived from Escherichia coli. J Virol 7:24–28 Morella NM, Gomez AL, Wang G, Leung MS, Koskella B (2018) The impact of bacteriophages on phyllosphere bacterial abundance and composition. Mol Ecol 27:2025–2038 Mushegian AR (2020) Are there 1031 virus particles on Earth, or more, or less? J Bacteriol 202: e00052–e00020 Nadell CD, Drescher K, Foster KR (2016) Spatial structure, cooperation and competition in biofilms. Nat Rev Microbiol 14:589–600 Nelson DC, Schmelcher M, Rodriguez-Rubio L, Klumpp J, Pritchard DG, Dong S, Donovan DM (2012) Endolysins as antimicrobials. Adv Virus Res 83:299–365 Oliveira H, Sao-Jose C, Azeredo J (2018) Phage-derived peptidoglycan degrading enzymes: challenges and future prospects for in vivo therapy. Viruses 10 Paez-Espino D, Eloe-Fadrosh EA, Pavlopoulos GA, Thomas AD, Huntemann M, Mikhailova N, Rubin E, Ivanova NN, Kyrpides NC (2016) Uncovering Earth's virome. Nature (London) 536:425–430 Parikka KJ, Le RM, Wauters N, Jacquet S (2017) Deciphering the virus-to-prokaryote ratio (VPR): insights into virus-host relationships in a variety of ecosystems. Biol Rev Camb Philos Soc 92:1081–1100 Pawluk A, Davidson AR, Maxwell KL (2018) Anti-CRISPR: discovery, mechanism and function. Nat Rev Microbiol 16:12–17 Peankuch E, Kausche GA (1940) Isolierung und, übermikroskopische Abbildung eines Bakteriophagen. Naturwissenschaften 28:46 Pires DP, Oliveira H, Melo LD, Sillankorva S, Azeredo J (2016) Bacteriophage-encoded depolymerases: their diversity and biotechnological applications. Appl Microbiol Biotechnol 100:2141–2151 Porter KG, Feig YS (1980) The use of DAPI for identifying and counting aquatic microflora 1. Limnol Oceanogr 25:943–948 Poullain V, Gandon S, Brockhurst MA, Buckling A, Hochberg ME (2008) The evolution of specificity in evolving and coevolving antagonistic interactions between a bacteria and its phage. Evolution; Int J Organic Evolut 62:1–11 Proctor LM (1997) Advances in the study of marine viruses. Microsc Res Tech 37:136–161 Rabinovitch A, Aviram I, Zaritsky A (2003) Bacterial debris—an ecological mechanism for coexistence of bacteria and their viruses. J Theor Biol 224:377–383 Reche I, D'Orta G, Mladenov N, Winget DM, Suttle CA (2018) Deposition rates of viruses and bacteria above the atmospheric boundary layer. ISME J 12:1154–1162

Bacteriophage Ecology

293

Reyes A, Haynes M, Hanson N, Angly FE, Heath AC, Rohwer F, Gordon JI (2010) Viruses in the faecal microbiota of monozygotic twins and their mothers. Nature (London) 466:334–338 Rodriguez-Brito B, Li L, Wegley L, Furlan M, Angly F, Breitbart M, Buchanan J, Desnues C, Dinsdale E, Edwards R, Felts B, Haynes M, Liu H, Lipson D, Mahaffy J, Martin-Cuadrado AB, Mira A, Nulton J, Pasic L, Rayhawk S, Rodriguez-Mueller J, Rodriguez-Valera F, Salamon P, Srinagesh S, Thingstad TF, Tran T, Thurber RV, Willner D, Youle M, Rohwer F (2010) Viral and microbial community dynamics in four aquatic environments. ISME J 4:739–751 Rosario K, Fierer N, Miller S, Luongo J, Breitbart M (2018) Diversity of DNA and RNA viruses in indoor air as assessed via metagenomic sequencing. Environ Sci Technol 52:1014–1027 Ross A, Ward S, Hyman P (2016) More is better: selecting for broad host range bacteriophages. Front Microbiol 7:1352 Roux S, Enault F, Hurwitz BL, Sullivan MB (2015) VirSorter: mining viral signal from microbial genomic data. PeerJ 3:e985 Roux S, Solonenko NE, Dang VT, Poulos BT, Schwenck SM, Goldsmith DB, Coleman ML, Breitbart M, Sullivan MB (2016) Towards quantitative viromics for both double-stranded and single-stranded DNA viruses. PeerJ 4:e2777 Roychoudhury P, Shrestha N, Wiss VR, Krone SM (2014) Fitness benefits of low infectivity in a spatially structured population of bacteriophages. Proc Biol Sci 281:20132563 Ruska H (1940) Die Sichtbarmachung der bakteriophagen Lyse im Übermikroskop. Naturwissenschaften 28:45–46 Schijven JF, Hassanizadeh SM (2000) Removal of viruses by soil passage: overview of modeling, processes, and parameters. Crit Rev Environ Sci Technol 30:49–127 Shao Y, Wang I-N (2008) Bacteriophage adsorption rate and optimal lysis time. Genetics 180:471–482 Short SM, Suttle CA (2002) Sequence analysis of marine virus communities reveals groups of related algal viruses are widely distributed in nature. Appl Environ Microbiol 68:1290–1296 Simpson JT, Wong K, Jackman SD, Schein JE, Jones SJ, Birol I (2009) ABySS: a parallel assembler for short read sequence data. Genome Res 19:1117–1123 Staley JT, Konopka A (1985) Measurement of in situ activities of nonphotosynthetic microorganisms in aquatic and terrestrial habitats. Ann Rev Microbiol 39:321–346 Stent GS (1963) Molecular biology of bacterial viruses. WH Freeman and Co, San Francisco Stewart FM, Levin BR (1984) The population biology of bacterial viruses: why be temperate. Theor Pop Biol 26:93–117 Sutherland IW, Hughes KA, Skillman LC, Tait K (2004) The interaction of phage and biofilms. FEMS Microbiol Lett 232:1–6 Szekely AJ, Breitbart M (2016) Single-stranded DNA phages: from early molecular biology tools to recent revolutions in environmental microbiology. FEMS Microbiol Lett 363 Tariq MA, Everest FL, Cowley LA, de SA, Holt GS, Bridge SH, Perry A, Perry JD, Bourke SJ, Cummings SP, Lanyon CV, Barr JJ, Smith DL (2015) A metagenomic approach to characterize temperate bacteriophage populations from cystic fibrosis and non-cystic fibrosis bronchiectasis patients. Front Microbiol 6:97 Thingstad TF (2000) Elements of a theory for the mechanisms controlling abundance, diversity, and biogeochemical role of lytic bacterial viruses in aquatic systems. Limnol Oceanogr 45:1320–1328 Thingstad TF, Bratbak G, Heldal M (2008) Aquatic phage ecology. In: Abedon ST (ed) Bacteriophage ecology. Cambridge University Press, Cambridge, UK, pp 251–280 Tikhe CV, Husseneder C (2017) Metavirome sequencing of the termite but reveals the presence of an unexplored bacteriophage community. Front Microbiol 8:2548 Touchon M, de Sousa JAM, Rocha EP (2017) Embracing the enemy: the diversification of microbial gene repertoires by phage-mediated horizontal gene transfer. Curr Opin Mirobiol 38:66–73 Tringe SG, Hugenholtz P (2008) A renaissance for the pioneering 16S rRNA gene. Curr Opin Microbiol 11:442–446 Tromas N, Zwart MP, Lafforgue G, Elena SF (2014) Within-host spatiotemporal dynamics of plant virus infection at the cellular level. PLoS Genet 10:e1004186

294

J. J. Dennehy and S. T. Abedon

Trubl G, Hyman P, Roux S, Abedon ST (2020) Coming-of-age characterization of soil vuruses: a user's guide to virus isolation, detection within metagenomes, and viromics. Soil Sys 4:23 Tucker KP, Parsons R, Symonds EM, Breitbart M (2011) Diversity and distribution of singlestranded DNA phages in the North Atlantic Ocean. ISME J 5:822–830 Wagg C, Bender SF, Widmer F, van der Heijden MG (2014) Soil biodiversity and soil community composition determine ecosystem multifunctionality. Proc Natl Acad Sci U S A 111:5266–5270 Waldor MK, Mekalanos JJ (1996) Lysogenic conversion by a filamentous phage encoding cholera toxin. Science 272:1910–1914 Wang I-N (2006) Lysis timing and bacteriophage fitness. Genetics 172:17–26 Wasik BR, Bhushan A, Ogbunugafor CB, Turner PE (2015) Delayed transmission selects for increased survival of vesicular stomatitis virus. Evolution; Int J Organic Evolut 69:117–125 Weinbauer MG (2004) Ecology of prokaryotic viruses. FEMS Microbiol Rev 28:127–181 Whon TW, Kim MS, Roh SW, Shin NR, Lee HW, Bae JW (2012) Metagenomic characterization of airborne viral DNA diversity in the near-surface atmosphere. J Virol 86:8221–8231 Wilhelm SW, Suttle CA (1999) Viruses and nutrient cycles in the sea: viruses play critical roles in the structure and function of aquatic food webs. Bioscience 49:781–788 Williams ST, Mortimer AM, Manchester L (1987) Ecology of soil bacteriophages. In: Goyal SM, Gerba CP, Bitton G (eds) Phage ecology. Wiley, New York, pp 157–179 Williamson KE (2018) Viruses of microorganisms in soil ecosystems. In: Hyman P, Abedon ST (eds) Viruses of microorganisms. Caister Academic Press, Norwich, pp 77–93 Williamson KE, Wommack KE, Radosevich M (2003) Sampling natural viral communities from soil for culture-independent analyses. Appl Environ Microbiol 69:6628–6633 Williamson KE, Radosevich M, Wommack KE (2005) Abundance and diversity of viruses in six Delaware soils. Appl Environ Microbiol 71:3119–3125 Williamson KE, Radosevich M, Smith DW, Wommack KE (2007) Incidence of lysogeny within temperate and extreme soil environments. Environ Microbiol 9:2563–2574 Williamson KE, Corzo KA, Drissi CL, Buckingham JM, Thompson CP, Helton RR (2013) Estimates of viral abundance in soils are strongly influenced by extraction and enumeration methods. Biol Fertil Soils 49:857–869 Williamson KE, Fuhrmann JJ, Wommack KE, Radosevich M (2017) Viruses in soil ecosystems: an unknown quantity within an unexplored territory. Annu Rev Virol 4:201–219 Winter C, Bouvier T, Weinbauer MG, Thingstad TF (2010) Trade-offs between competition and defense specialists among unicellular planktonic organisms: the "killing the winner" hypothesis revisited. Microbiol Mol Biol Rev 74:42–57 Wommack KE, Colwell RR (2000) Virioplankton: viruses in aquatic ecosystems. Microbiol Mol Biol Rev 64:69–114 Young R (2014) Phage lysis: three steps, three choices, one outcome. J Microbiol 52:243–258 Zhang QG, Buckling A (2011) Antagonistic coevolution limits population persistence of a virus in a thermally deteriorating environment. Ecol Lett 14:282–288 Zheng Y, Struck DK, Dankenbring CA, Young R (2008) Evolutionary dominance of holin lysis systems derives from superior genetic malleability. Microbiology 154:1710–1718

Bacteriophage Pharmacology and Immunology Krystyna Dąbrowska, Andrzej Górski, and Stephen T. Abedon

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Safety Considerations in Phage Choice for Phage Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Avoiding Temperate Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Avoiding Phages Encoding Virulence Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Avoiding Transducing Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary: Safety Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Therapy Pharmacology Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pharmacokinetics and Pharmacodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pharmacokinetics in More Detail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conceptualizing Pharmacology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Passive Treatment Versus Active Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Passive Phage Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Active Penetration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Active Phage Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summarizing Phage Therapy Pharmacological Phenomena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

296 299 303 304 305 305 306 306 308 308 310 311 315 318 318

K. Dąbrowska Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wrocław, Poland e-mail: [email protected] A. Górski Bacteriophage Laboratory, Phage Therapy Unit, Hirszfeld Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wrocław, Poland Department of Clinical Immunology, Transplantation Institute, Medical University of Warsaw, Wrocław, Poland e-mail: [email protected] S. T. Abedon (*) Department of Microbiology, The Ohio State University, Mansfield, OH, USA e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_9

295

296

K. Dąbrowska et al.

Phage Interactions with Immune Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview of Immunity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adaptive Immunity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Innate Immunity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Effects of Immune Response on Phage Pharmacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Relation of Immune Response to Phage Pharmacodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

321 322 323 326 327 328 330 333 333

Abstract

The discovery of bacterial viruses approximately 100 years ago fairly quickly led to their use as antibacterial agents. For roughly two decades – early 1920s to early 1940s – bacteriophages represented the only means readily available to medicine by which many bacterial infections might be treated and cured. This near monopoly, however, came to a close as antibiotics became generally available. Antibiotics, especially as more broadly specific, selectively toxic antibacterials were both more easily developed and more easily used medicinals than phages. Phage therapy did not disappear from medical practice altogether, however, and increasingly is viewed as a viable alternative to antibiotics under circumstances where bacterial resistance to antibiotics is an issue. In addition are circumstances where a more selectively toxic antibacterial is desired, antibacterials that, for example, have less of a negative impact on nontarget members of a body’s microbiome. As for any drug, the successful development of phage therapeutics requires a pharmacological approach, whether implicit or, ideally, explicitly implemented. In this chapter, we consider pharmacokinetic and pharmacodynamic principles, body impact on drugs and drug impact on body, respectively, and both as they may be applied to the development of phagebased antimicrobials. As an important facet of both the pharmacokinetics and pharmacodynamics of phage therapy, we take a close look particularly at phage interactions with the mammalian immune system.

Introduction While the development of phages as medicinals has been strongly influenced by the parallel development of antibiotics, at the same time antibiotics have generally been studied in terms of a much more rigorous pharmacological tradition than has been the case for phage therapy. There exist perhaps four prominent reasons for that dissimilarity in development approaches. First is that phages for phage therapy in almost all instances have proven to be mostly safe to use as drug equivalents (Olszowska-Zaremba et al. 2012; McCallin et al. 2013; Abedon 2015d; Pirnay et al. 2015), and particularly so in purified forms (Gill and Hyman 2010; Łobocka et al. 2014; Fish et al. 2016; Schooley et al. 2017; Chan et al. 2018). Low drug toxicity has the effect of reducing the importance of

Bacteriophage Pharmacology and Immunology

297

secondary pharmacodynamic issues, meaning in turn that approaches to improving phage therapy efficacy often can be explored without overriding concern for potential negative impacts of those strategies on patient health. Prominently, the display by a drug of only low levels of toxicity allows for reduced need to rein in drug concentrations, that is, to devote substantial effort towards the avoidance of exceeding a drug’s minimum toxic density during the course of dosing. As a result, efforts in phage therapy development have tended to emphasize phage antibacterial effectiveness and delivery strategies, and this is rather than achievement of sufficient densities while simultaneously avoiding toxicity. Generally, that is, reaching bacteria with sufficient numbers of the right kind of phages, and over sufficient time spans, can be the emphasis of phage therapy development even to the point where in fact it can be helpful to remind researchers that explicit monitoring of potential toxicities – especially of treated animals during preclinical development, if only for the sake of building “up in the literature a record of phage toxicity testing” (Abedon 2012b) – can be important as well. A second reason for pharmacology having played less of a role in phage therapy development, versus development of small molecule antibiotics, stems from the potential for phages to increase in concentration during their action in situ. That is, phages as viruses possess an ability to replicate in the course of their antibacterial action, which in turn has the effect, in a number of clinical circumstances, of increasing the potential for specific phage delivery strategies to result in sufficient phage densities to achieve bacterial eradication. Generally, if bacterial infections by phages are robust enough – large enough burst sizes in combination with sufficiently fast virion adsorption and not excessively long latent periods – then simply delivering even relatively small quantities of phages to those bacteria often can result in the generation of sufficient phage densities, over time, to result in eradication of sensitive bacteria (i.e., active treatment). Furthermore, this typically can be achieved, as noted, without substantial concerns over phage-associated toxicity. The need to study the ability of a drug to reach its intended target in sufficient numbers therefore can be alleviated somewhat if an inherent property of the drug is one of increasing to sufficient numbers once it has reached its target. Third, though there is justified interest during phage therapy development in phage host range (Hyman and Abedon 2010; Chan and Abedon 2012a; Łobocka et al. 2014), in terms of phage spectrum of activity, there tends to be much less interest in what for antibiotics is a major consideration, that of minimum inhibitory concentration (MIC). Though at least in part this lack of interest in MIC determination is a result of conceptual difficulties in defining phage minimum inhibitory concentrations (Abedon 2011a), such as due to the single-hit killing characteristics of phages (Bull and Regoes 2006), the net result nonetheless is that there has been little tradition in phage therapy for studying the underlying bases of efficacy in terms of phage performance except to the extent that should one phage type prove to be ineffective in the clinic then another phage – from among of what typically is a diversity of possible, safe choices – may be employed instead (Pirnay et al. 2011; Chan et al. 2013). This latter point is changing to a degree, however, as researchers have begun to more formally link together more subtle aspects of phage host range, that is, infection performance in vitro, with their effectiveness during experimental

298

K. Dąbrowska et al.

phage therapy (Henry et al. 2013; Bull and Gill 2014; Lindberg et al. 2014); see also (Łobocka et al. 2014). Nevertheless, an important aspect of antibiotic pharmacological study, that of a concentration dependence of antibacterial activity, has played much less of a role in phage therapy development than it has for chemotherapies. Fourth, especially early development of the practice of phage therapy was a time during which substantial clinical experimentation was permissible, for example, Abedon (2015d, 2018a), but also Chanishvili (2012a). When alternative treatments are lacking, and especially when a patient’s survival is under threat, then principles of compassionate care can be applied including as emergency investigational new drugs (Międzybrodzki et al. 2012; Kutter et al. 2015; Fish et al. 2016; Schooley et al. 2017; Chan et al. 2018). Treatments thus may commence without consideration of possible subtleties of drug pharmacology beyond following standard protocols of drug application or infection preparation (e.g., wound debridement), and otherwise carefully monitoring patient health, as well as use of informed phage choice and purification (Gill and Hyman 2010; Łobocka et al. 2014). At the same time, additional pharmacological study, versus simply efforts towards efficacy enhancement (e.g., trying different phages or delivery strategies), may not be robust in the course of such care. By contrast, in modern medicine, where an emphasis is usually placed more on regulation and standards of care rather than extensive clinical trial and error, pharmacological considerations can and must have a much more prominent place in drug development. For a historical look at phage use clinically versus preclinically, see the review by Abedon (2015d) on the use of phages to treat lung-associated infections. Despite the potential for both the study of phage therapy and achievement of efficacy without strong consideration of phage therapy pharmacology, modern norms of drug development nonetheless provide a countering force to this tendency. Related to the previous point is the use of animal models for phage therapy development. Animal models typically do not perfectly mimic human disease. As a consequence, use of animals as an aspect of drug development requires consideration of how pharmacological characteristics may differ between these experimental systems and actual patients. Indeed, patient safety considerations along with the typical expense of clinical trials alone should drive such an interest. Furthermore, there exists an economic as well as moral argument for consideration of phage therapy pharmacology as an aid towards improving treatment protocols before, during, and after animal experimentation since this can lead to reductions in the total number of trials – clinical as well as preclinical – that are required for successful development. Included among phage properties during phage therapy that warrant such increased consideration – and consistently of interest to those familiarizing themselves with this technology – is the issue of phage virion interactions with mammalian immune systems. These interactions can be distinguished into three somewhat distinct facets (1) the potential for immune systems to inhibit phage therapy efficacy (a pharmacokinetic concern), (2) the potential for immune system reaction to phage presence to result in side effects (as distinct, it should be noted, from the potential for target bacteria to provoke negative immune reactions), and (3) the potential for phages to provide positive immunomodulatory effects, including as may occur independently of bacterial targeting during phage therapy.

Bacteriophage Pharmacology and Immunology

299

With such considerations in mind, in this chapter we walk through the basics of phage therapy pharmacology, pointing the reader to a growing literature on the subject. We begin with discussion of the relative safety that has been observed with phage therapy. We then provide a general discussion of phage therapy pharmacology. Lastly, we consider issues of phage-immune system interaction. For additional reading, note that Łobocka et al. (2014) provide an excellent primer on the criteria that can be employed in choosing a phage for phage therapy; see equivalently Gill and Hyman (2010). For discussion of routes of phage delivery into bodies, see Ryan et al. (2011). For practical considerations of phage therapy pharmacology, experimentation, and debugging of protocols, see Abedon (2012b, 2017c, 2018b) as well as the Appendix to Abedon (2017a). For consideration of the use of phage cocktails in phage therapy, see Chan and Abedon (2012b) and Chan et al. (2013). For discussion of the distinction between phage-mediated “biocontrol” and phage therapy, see Abedon (2009). For additional consideration of the history of phage therapy, see Summers (2005), Abedon et al. (2011), Chanishvili (2012b), Harper and Morales (2012), Summers (2012), and Abedon (2017b) (chapter “▶ The Discovery of Bacteriophages and the Historical Context”). And for volumes with a substantial phage therapy component, see Kutter and Sulakvelidze (2005), Sabour and Griffiths (2010), Abedon (2010), Hyman and Abedon (2012), and Borysowski et al. (2014) along with this volume. A glossary of terms relevant to phage therapy pharmacology can be found in Table 1.

Safety Considerations in Phage Choice for Phage Therapy As noted, a key reason for why phage therapy has tended to be developed without an accompanying robust pharmacological tradition is the relative safety of phages as antibacterials, which includes, historically, the phage therapy of fairly large numbers of people in Europe and the former Soviet Union (Abedon et al. 2011; Kutter et al. 2010; Chanishvili 2012a; Kutter et al. 2014); see also (Abedon 2015d). This safety is the result of a number of factors stemming from a combination of inherent phage properties and well-informed phage choice. In short, an important goal for phage choice in phage therapy is to choose those phages that are inherently safe (Łobocka et al. 2014; Pirnay et al. 2015), and this generally means (1) the avoidance of phages that can display lysogenic cycles (chapter ▶ “Temperate Phages, Prophages, and Lysogeny”), (2) avoidance as well of phages that may otherwise encode potentially toxic genes (chapter ▶ “Temperate Phages, Prophages, and Lysogeny,”), and (3) otherwise, at least to some degree, avoiding phages that are able to transduce nonphage genes between bacteria. To a first approximation, the phage potential to display all three of these properties can be reduced via the exclusive use of nontemperate and particularly professionally lytic phages for phage therapy purposes (chapter ▶ “Phage Infection and Lysis”). In this section, we consider these issues further. Note again the reviews by Łobocka et al. (2014) and Gill and Hyman (2010) covering the subject of phage choice.

300

K. Dąbrowska et al.

Table 1 Glossary of terms relevant to phage therapy pharmacology Term Abortive infection (broadly defined)

Abortive infection (narrowly defined)

Absorption

Active infection Active penetration

Active replication Active treatment (or therapy)

Adsorption

Auto dosing

Bactericidal phage infection Distribution

Excretion

Definition Bacterial infections by phages that result in bacterium death in combination with substantially reduced phage-infection productivity; abortive infections can be the result either of incompatibilities between infecting phages and bacteria or instead bacteria-encoded antiphage mechanisms Bacterial infection by phages that result in bacterial death in combination with inactivation of the infecting phage (i.e., a combination of resulting in neither a phage burst nor a latent phage latent infection) Uptake of medicinals into the blood; relevance during phage therapy is seen either given requirements for systemic phage application or, alternatively, should more localized phage administration result in phage movement into blood; contrast “Adsorption,” below Equivalent to productive infection Proposed scenario of phage antibiofilm activity that is dependent on phage-induced lysis of target bacteria, particularly as results either in more effective phage movement to underlying layers of bacteria and/or as allows for more effective phage infection of these underlying bacteria given successful adsorption The consequence of active infection Phage therapy or phage-mediated biocontrol of bacteria that requires productive phage infections to be successful; particularly, active treatment is required to the extent that abortive/bactericidal infections alone are insufficient to result in successful eradication of target bacteria Phage-virion acquisition of a new bacterial host; contrast “Absorption,” above. The process of phage adsorption generally ends with specific virion interactions with receptor molecules found on bacterial surfaces and is followed by translocation of virion-associated nucleic acid into the bacterial cytoplasm In situ generation of new phage particles as a consequence of productive phage infections. Auto dosing involves active phage replication and productive phage infections and is required for active treatment Typically associated with phage lytic cycles though can result from phage abortive infections as well Movement of medicinals out of the blood and (ideally) into target tissues; note that the term “penetration,” though not quite identical in meaning, nevertheless is often employed in a phagetherapy context rather than distribution Movement of medicinals out of the body in a chemically intact state; importance to phage therapy is seen especially when such phage movement aids phages in accessing target bacteria, for example, systemic phage dosing towards phage access to the urinary tract (continued)

Bacteriophage Pharmacology and Immunology

301

Table 1 (continued) Term Immunomodulation

Inundation therapy (or inundative treatmenta) Lysis from without

Lysogen Lysogenic conversion Lysogenic cycle (or infection)

Lytic phage

Metabolism

Mixed passive-active treatment (or therapy)

Neutralizing antibody Obligately lytic Parenteral Passive treatment (or therapy)

Penetration

Per os

Definition Regulation of elements of immune response by an external factor (here: phage), thus changing the mode of action of the immune system in the presence of this factor Equivalent to passive treatment, so named because phage numbers in excess of bacteria numbers must be supplied via standard dosing for successful bacterial eradication to occur Abortive interaction between certain phages and their hosts that strictly is associated with a combination of high multiplicity phage adsorption, truncated infections, and bacterial lysis; so far as is known, lysis from without is associated with only a relatively small subset of phage types, that is, T-even-type phages so should not be invoked indiscriminately Prophage-containing bacterium Expression of prophage-encoded genes that results in modification of the phenotype of a lysogenized bacterium Latent phage infection that does not, without induction, involve production of progeny phage virions and during which the phage genome exists as a prophage Bacterial virus capable of infecting lytically. Use of this descriptor ideally has no bearing on a phage’s potential to infect lysogenically, and while all virulent phages are lytic phages, not all lytic phages are virulent phages. Chemical modification of medicinals; for phages this includes activation of bactericidal activity, associated bacterial lysis, and production of virion progeny Phage therapy or phage-mediated biocontrol of bacteria that does not explicitly require productive phage infections to be successful, that does require bactericidal phage infections, but which nevertheless is enhanced in its effectiveness as a consequence of infection production of new phage virions Specific antibody that is capable of blocking the activity of its target Phage that is unable to display lysogenic cycles; equivalent to strictly lytic but see also professionally lytic Nonalimentary application of medicinals especially for the sake of systemic distribution Phage therapy or phage-mediated biocontrol of bacteria that does not require productive phage infections for treatments to be efficacious, but does require that phage infections are bactericidal Physical movement of phage virions especially within poorly mixed environments towards target bacteria, for example, as combining the pharmacokinetic aspects of absorption and distribution (and, in certain cases, also excretion), as well as phage movement into the matrix of bacterial biofilms Oral application of medicinals, for example, as can be followed by absorption into the blood and then movement (distribution) out of the blood (continued)

302

K. Dąbrowska et al.

Table 1 (continued) Term Phage therapy

Phage-mediated bacterial biocontrol Pharmacodynamics

Pharmacodynamics, primary Pharmacodynamics, secondary Pharmacokinetics

Pharmacologically emergent property Primary infection (of individual cells)

Primary infection (epidemiological sense)

Productive infection Professionally lytic (as used here)

Prophage Restrictive infection or restricted infection

Definition Application of bacterial viruses especially to bodies as a means of reducing numbers of unwanted bacteria, and particularly as observed within clinical, medicinal, or veterinary contexts Application of bacterial viruses especially other than bacteriainfected organisms as a means of reducing numbers of unwanted bacteria Impact of drugs on a body, that is, how drugs affect bodies, with bodies defined to include not just body tissues but also a body’s microbiome. Pharmacodynamics thus considers both the positive and negative impacts of drugs, on bodies, but not factors specifically affecting drug concentrations over time in the body – see instead pharmacokinetics for the latter. Pharmacodynamics can be distinguished into primary versus secondary effects Intended impact of drugs on a body Unintended impact of drugs on a body, including though not limited to in terms of toxicities and side effects Impact of a body on drugs, that is, how bodies chemically or spatially influence drugs following dosing; the pharmacokinetics of a drug helps to determine a drug’s concentration within the vicinity of drug targets within the body over time, where drug concentrations within the vicinity of drug targets within the body in turn will tend to determine the magnitude of pharmacodynamic effects; see also absorption, distribution, excretion, and metabolism as pharmacokinetic processes Ability of medicinals to display especially unexpected toxicities during drug preclinical as well as clinical development Phage infection of a previously phage-uninfected bacterium, though the concept can be applied to the active infection of a previously latently phage-infected bacteria as well, for example, phage adsorption followed by productive infection of a bacterial lysogen Phage infection of a previously not actively phage-infected bacterium where the infecting virion was not generated in situ but rather was supplied via dosing, that is, an infection that occurs by a virion after that virion has entered into a new environment Phage infection that directly results in the production as well as release of progeny phage virions Lytic phage that is both unable to display lysogenic cycles, that is, is obligately lytic, and is not otherwise closely related to or recently descended from phages that can display lysogenic cycles (i.e., not similar genetically to temperate phages) Term used to describe latently infecting phage genomes during lysogenic cycles Phage infection that results in phage inactivation in conjunction with survival of the host bacterium, for example, as mediated by bacterial restriction-modification systems or CRISPR-Cas systems (continued)

Bacteriophage Pharmacology and Immunology

303

Table 1 (continued) Term Reticuloendothelial system

Secondary infection (of individual cells)

Secondary infection (epidemiological sense)

Single-hit killing characteristics Strictly lytic Superinfection immunity

Temperate phage Transduction Translocation (phage infection) Translocation (pharmacokinetics)

Definition Mechanism of phagocytic removal of particles such as viruses from blood without requirement for antibody or complement, a. k.a., mononuclear phagocyte system Phage infection of a previously actively phage-infected bacterium (though the concept can be applied to the secondary infection of previously latently phage-infected bacteria as well); many authors use the phrase superinfection instead Phage infection of a previously not actively phage-infected bacterium where the infecting virion had been generated in situ rather than supplied via dosing; secondary infections in this sense are the hallmark of active treatments Phage potential to effect bactericidal effects on bacteria given the adsorption of only one phage; contrast with the multihit killing kinetics of bactericidal small-molecule antibiotics Equivalent to obligately lytic Prophage-expressed means by which phages of equivalent immunity type are prevented from successfully infecting (i.e.,, prevented from superinfecting), including prevented from displaying bactericidal activity Bacterial virus that is capable of infecting lysogenically; often incorrectly referred to instead as lysogenic phages Movement by phages especially of bacterial DNA from one bacterial host to another Movement of phage DNA into a bacterium’s cytoplasm following virion adsorption; more generally, this is nucleic acid translocation Movement of phage virions especially across the wall of the gastrointestinal tract into systemic circulation, as more or less equivalent to absorption; often referred to as phage translocation

a Note that the while the proper adjectival form of “inundate” in fact is “inundatory,” Payne et al. (2000) as well as Payne and Jansen (2003) refer to an “inundative dose” or “inundative doses” while “inundative biological control” is a legitimate term, of which ‘inundative treatment,” a form of biological control using phages as the control agent, could be viewed as an example (“inundatory biological control,” by contrast, is not a legitimate term). Though Payne et al. (2000) also speak of an “inundatory dose,” in fact, as determined via Google as well as Google Scholar, the use of “inundative treatment” or “inundative dose” is far more prevalent than the equivalent use of “inundatory treatment” or “inundatory dose”

Avoiding Temperate Phages Bacteriophages come in a variety of types. For purposes of phage therapy, or phagemediated biocontrol of bacteria, the vast majority are both lytic and tailed. Lytic refers to the means by which a phage is released from its bacterial host, involving the destruction of the bacterial cell envelope (chapter ▶ “Phage Infection and Lysis”). In the phage life cycle, this lysis allows the intracellularly produced phage progeny virions to exit the host cell. In terms of phage therapy, lysis also represents a prominent aspect of the phage bactericidal nature, though in fact bacterial genetic

304

K. Dąbrowska et al.

death, if not explicitly bacterial metabolic death, generally precedes phage-induced bacterial lysis. The phage tail by contrast is not explicitly required for phage therapy but nonetheless is very common among lytic phages (chapter “▶ Structure and Function of Bacteriophages”) and, as a consequence, is very common as well among those phages that are employed in phage therapy. Among tailed phages there are temperate phages versus those that are not temperate. The latter also are known as virulent or what has been variously described as “obligately lytic,” “professionally lytic,” or “strictly lytic” (Hobbs and Abedon 2016). Many authors also, though incorrectly, use simply the term “lytic” as a synonym for not temperate, though this latter practice should be discouraged since the productive cycle of all tailed phages is what is known as a lytic cycle. Temperate phages – which are common among tailed phages, for example, perhaps half (Ackermann 2005) – also with relatively few exceptions employ lytic cycles towards phage virion production but in addition are able to display lysogenic cycles (chapter ▶ “Temperate Phages, Prophages, and Lysogeny,”). The resulting lysogens tend to be resistant to hosting infections by subsequently adsorbing phages, particularly phages that are of the same type as those already lysogenically infecting the bacterium. This impact of phages displaying lysogenic cycles on subsequent infections by phages of the same type is described as superinfection immunity, homoimmunity, or simply immunity (Campbell 2006; Casjens and Hendrix 2015). As a consequence of this phage-mediated immunity, use of temperate phages in phage therapy can directly result – even in the absence of bacterial mutation – in the generation of bacterial pathogens that are resistant to the very phages employed as antibacterials to combat them. This is one reason that the use of temperate phages for phage therapy is frowned upon, and this is so even to the extent that temperate phages, upon bacterial infection, often will enter into lytic cycles rather than lysogenic cycles.

Avoiding Phages Encoding Virulence Factors An important second reason for avoiding the use of temperate phages in phage therapy is that, among phages, the encoding of bacterial virulence factors is almost if not always associated with temperate phages (Christie et al. 2012; Kuhl et al. 2012). Obligately lytic phages – or more explicitly, professionally lytic phages, defined as lytic phages that are not immediate descendants of temperate phages (Hobbs and Abedon 2016) – are as a consequence preferentially used for phage therapy purposes. For safety reasons, and even though the expectation is that professionally lytic phages will not carry potentially dangerous bacterial virulence factor genes, nevertheless it is common practice to fully sequence and then do bioinformatic analysis (chapter “▶ Genetics and Genomics of Bacteriophages”) on phages prior to their use for phage therapy purposes (Łobocka et al. 2014). If nothing else, such analysis can be helpful in determining whether or not such phages truly are professionally lytic, that is, not containing lysogeny-associated gene sequences, versus being temperate phages that under the conditions tested simply happen to not display lysogenic cycles.

Bacteriophage Pharmacology and Immunology

305

Despite the potential utility of fully sequencing and annotating phages prior to clinical use, in practice the utility and safety of phage therapy substantially predates the development of such bioinformatic analysis. As the costs of sequencing continue to decline, however, arguments against including routine bioinformatic analysis in phage characterization will become weaker. An additional issue that is related to phage bioinformatic analysis prior to use stems from potential regulatory approaches which, in principle, could favor the approval of general strategies of phage product development rather than the development simply of specific phages, for example, strategies as one sees with the yearly influenza vaccine (Sulakvelidze and Kutter 2005), a comparison which to date has been pointed out by a number of authors, that is, as listed in Abedon (2017b). To the extent that full-genome sequencing and bioinformatic analysis becomes incorporated into approved protocols, then such analysis would become inherent to the ongoing development of phage therapy schemes.

Avoiding Transducing Phages As noted, an additional issue is that of transduction (Łobocka et al. 2014) (chapter “▶ Bacteriophage-Mediated Horizontal Gene Transfer: Transduction”). Two types of phage-mediated transduction of bacterial genes exist, which are described as specialized transduction versus generalized transduction. Specialized transduction, as usually defined, explicitly is the carriage of bacterial genes by partially intact temperate phages. Specialized transduction as a concern to phage therapy consequently can be avoided via the exclusive use of nontemperate and especially professionally lytic phages for phage therapy. Generalized transduction is the carriage of bacterial genes by phage virions that have failed also to package phage genes. This property is seen especially in phages whose DNA packaging does not involve specific genome-packaging sequences and also for phages whose life cycles do not involve substantial destruction of the genome of the bacterial host in the course of infection. Notwithstanding the frequently expressed concern with the ability of some phages to readily transduce bacterial DNA, not all authors agree that this is of substantial concern regarding phage choice for phage therapy (chapter “▶ Bacteriophage-Mediated Horizontal Gene Transfer: Transduction”).

Summary: Safety Considerations To avoid negative secondary pharmacodynamic issues during phage therapy – that is, preventing phage-based antibacterials from displaying toxicity and side effects – it is thus crucial under most circumstances to show with reasonably high certainty that a phage is professionally lytic, that it lacks genes that could potentially encode bacterial virulence factors, and also, as some argue, it can be helpful as well to avoid for therapy purposes phages that are able to support generalized transduction. In addition to these issues of phage choice, and especially for parenteral delivery

306

K. Dąbrowska et al.

(Speck and Smithyman 2016), it can be crucial to purify phages away from lysisassociated bacterial toxins such as endotoxin (Boratynski et al. 2004; Gill and Hyman 2010; Łobocka et al. 2014; Pirnay et al. 2015; Szermer-Olearnik and Boratynski 2015). Thus, once phages have been deemed to be obligately lytic – and ideally professionally lytic – as well as free of bacterial virulence factors genes, unable to readily effect the transduction of bacterial genes, and have been appropriately purified, then they generally can be deemed to be safe for phage therapy use (Olszowska-Zaremba et al. 2012). Phages, contrasting novel chemotherapeutics, tend also to be relatively lacking in pharmacologically emergent properties (Curtright and Abedon 2011). Thus, once characterized in vitro there tend to be few emergent safety issues observed following the introduction of new phage types into either animals or the clinic.

Phage Therapy Pharmacology Basics In pharmacology, the body is considered to consist of both its own tissues and associated microbiota. Thus, in the course of treating an infection, the normal functioning of an antimicrobial agent is to affect the body by causing the elimination of a microbial parasite or pathogen. This action can be described as the antimicrobial’s primary pharmacodynamic behavior. Secondary pharmacodynamic behavior of an antimicrobial agent, that is, abnormal or at least unintended or undesired functioning, instead is associated with the disruption of normal body tissues, of body metabolism generally, or of normal microbiota, with the latter consisting of commensal or mutualistic microorganisms that contribute to the maintenance of body homeostasis. A major advantage of properly chosen phages as antibacterial agents – especially as delivered topically, in a purified form, or both – is that there is a low tendency for these agents to give rise to substantial disruptions of normal body tissues, metabolism, or normal microbiota beyond as caused by the bacterial infection itself. See Fig. 1 for an overview of these and additional pharmacological concepts.

Pharmacokinetics and Pharmacodynamics The body’s impact on a drug (pharmacokinetics) generally affects a drug’s impact on the body (pharmacodynamics), rather than the other way around. Specifically, a drug needs to reach its target before it can act, and the more drug that reaches its target then typically the stronger its effect, either positive or negative (primary or secondary). From a pharmacological perspective, then, the factors controlling a drug’s concentration in the vicinity of its target within a body are a function of the body’s impact on that drug. Pharmacokinetics thus impacts pharmacodynamics by impacting what a drug’s density will be within the immediate vicinity of target tissues or within the immediate vicinity of microorganisms. Thus: dosing ➔ pharmacokinetics ➔ local drug densities ➔ pharmacodynamics ➔ efficacy or toxicity

Bacteriophage Pharmacology and Immunology

307

Pharmacologically, a body consists of body tissues (e.g., liver cells) in combination with microbiota (i.e., “microbiome”); the body impacts the functioning of drugs (pharmacokinetics) and is impacted by the functioning of drugs (pharmacodynamics); each can act positively as well as negatively

Application, Dosing

Pharmacokinetics:

Pharmacodynamics:

Efficacy, Toxicity

Topical Per Os Parenteral “Auto”

Absorption Distribution Metabolism Excretion

Impact on body tissues and commensal microbiota

(these are components of pharmacodynamics)

Phage Replication

The impact of a drug is dependent on its concentration – as modified by its pharmacokinetics – as that concentration is found in the vicinity of both intended and unintended targets

Fig. 1 Overview of pharmacological basics with small emphasis on phage therapy. The latter is found in the lower-left of the figure, that is, “Phage Replication,” but the other factors discussed are applicable to phage therapy pharmacology as well as pharmacology more generally. “Application” and “Dosing” are to and of the body, while “Auto” refers to dosing that is generated automatically by the drug in the course of interaction with the body, that is, such as resulting from phage replication (auto dosing). These, along with pharmacokinetics, control medicinal concentrations in specific locations of the body, which in turn affects pharmacodynamics. Generally larger pharmacodynamic effects are observed, both positive and negative, the greater a drug’s concentration. In addition, as a general rule we can expect greater positive effects the higher that drug concentrations are found in the immediate vicinity of desired targets and greater negative effects the higher that drug concentrations are in the vicinity of undesired targets. Vis-à-vis “Phage Replication” in the figure, along with the curved arrow, we are explicitly describing “metabolism” as an aspect of phage therapy pharmacokinetics, that is, the chemical modification of a medicinal (phage interactions with host bacteria, in other words, results in chemical modifications of phages otherwise known as a phage infection)

(for the latter, positive/primary pharmacodynamics effects and negative/secondary pharmacodynamic effects, respectively). Towards distinguishing “pharmacodynamics” from “pharmacokinetics” mnemonically, consider that “dynamics” refers to “change” within a system. Thus, in pharmacology “dynamics” refers to the extent to which a drug impacts, that is, changes a body. Kinetics, by contrast, refers to rates. Pharmacokinetics thus addresses the rate at which effective drug concentrations are reached or, alternatively, are removed, where in pharmacology generally these rates are controlled by the action of the body on a drug. In the progression, dosing ➔ pharmacokinetics ➔ local drug densities ➔ pharmacodynamics ➔ efficacy or toxicity, the last entry reflects that pharmacodynamics can be differentiated into intended drug modifications of the body, that is, the above-noted primary (i.e., “principal”) pharmacodynamics, versus a drug’s less

308

K. Dąbrowska et al.

intended and occasionally toxic impact on the body, or secondary (“incidental”) pharmacodynamics. In either case, the degree of pharmacodynamic effects is a function of local drug densities and thus is dependent on a drug’s pharmacokinetics. It is important to keep in mind, however, that primary effects and secondary effects can result from drug densities that have built up in different locations within the body, with primary pharmacodynamic effects taking place in one location and secondary pharmacodynamic effects potentially taking place in another. The underlying physiological basis for a drug’s efficacy thus need not be identical to the underlying physiological basis for a drug’s toxicity. This is a situation which can “reward” high drug specificity since the result can be inherently lower occurrences of secondary pharmacodynamic effects, that is, to the extent that a drug interacts physiologically with as few targets within a body as possible and/or builds up in concentration in a minimal number of locations within the body.

Pharmacokinetics in More Detail Note that drug pharmacokinetics can be distinguished further into what traditionally are described as absorption (with a “b,” that is, not “adsorption” with a “d”), distribution, metabolism, and excretion. These, respectively, are drug entrance into the blood (where “absorb” means for something to move into something else), drug movement out of the blood into body tissues (i.e., “distribute” throughout the body), drug chemical modification (recall that “metabolism” refers to chemical reactions), and drug removal from the body (“excretion” meaning to expel or eliminate something as waste). These are important general pharmacological concepts. In the actual practice of phage therapy, however, they typically are referenced slightly differently, as we will consider. For additional discussion of pharmacokinetics and its application to phage therapy, see Abedon and Thomas-Abedon (2010) and Abedon (2014a, 2014b). See Table 2 for general summary of pharmacokinetics.

Conceptualizing Pharmacology Various concepts of pharmacokinetics (PK), pharmacodynamics (PD), and the impact of pharmacokinetics on pharmacodynamics we represent in a general, simplified form in Fig. 2. Indicated explicitly is the impact of pharmacokinetics on pharmacodynamics. This occurs via the various processes involved in pharmacokinetics, that is, absorption, distribution, metabolism, and excretion (Table 2). The arrows can be interpreted as indicating the PK impact on drug concentrations within specific regions of the body such that various PD effects may be realized. These PD effects can be primary (positive or efficacy) or, instead, secondary (generally negative). In the following subsections, we elaborate on Fig. 2 towards discussing phage therapy passive treatment, active penetration, and active treatment in greater pharmacological detail. Generally not considered will be specific dosing strategies including topical (which includes directly into the lungs), per os for treatment within the

Bacteriophage Pharmacology and Immunology

309

Table 2 Overview of different facets of pharmacokineticsa PK aspect Absorption

General effects Increase in drug concentrations within the blood

Distribution

Increase in drug concentrations outside of the blood (typically as reached from the blood)

Metabolism

Change in drug concentrations due to chemical modification of the drug; for phages this includes aspects of virion adsorption, gene expression, and also bactericidal, bacteriolytic, and reproduction activity Change in drug concentrations within the body due to removal of intact drug from the body

Excretion

Penetration

Combining phagevirion absorption, distribution, and to some extent metabolism; this term is not traditionally used as an aspect of pharmacokinetics but nevertheless is useful for discussion of phage therapy pharmacology

Positive effects Access of drugs to systemic circulation, thereby increasing drug concentrations both within the blood and systemically Access of drugs to targets as found beyond systemic circulation, thereby increasing drug concentrations within the body outside of the blood, though this can occur with varying degrees of specificity as well as efficiency depending on the drug, target tissues, and circumstances Though positive effects of metabolism on traditional drugs tend to be relatively rare, in certain cases drug activation within the body occurs through chemical modification, thereby increasing drug in situ concentrations; note, for example, phage reproduction Though rare as a positive effect, excretion can result, in certain cases, in drug movement to excretory organs as drug targets, for example, the urinary tract, thereby increasing local drug concentrations As an endpoint, penetration results in phage genome presence within the cytoplasm of target bacteria, with virion adsorption and genome translocation requiring phage metabolism; this is all towards initiation of further phage metabolism

Negative effects Exposure of body systemically to drugs; dilution of drug from that of original dose Exposure of tissues generally to potential drug toxicity; further dilution of drug concentration

Inactivation of drugs through chemical modification explicitly reduces drug quantities and thereby drug concentration; note, for example, consequences of phage-immune system interactions Removal of chemically intact drugs from the body explicitly reduces drug quantities within the body and thereby decreases drug concentration Generally this would be due to movement (penetration) away from target bacteria, that is, such that phage concentrations in the vicinity of target bacteria are not as large as they otherwise could be

a Recall that pharmacokinetics refers to the body’s impact on a drug, which can be viewed as affecting the rate at which effective drug concentrations are reached within the vicinity of drug targets – both primary and secondary – as well as the rate at which these concentrations will then tend to decline, both as following dosing

310

K. Dąbrowska et al.

Pharmacokinetics

Efficacy (1° PD)

Absorption Distribution Metabolism Excretion

Side Effects (2° PD) Fig. 2 Pharmacology basics. Pharmacokinetics is abbreviated as “PK” and pharmacodynamics as “PD.” Pharmacokinetics is distinguished into absorption, distribution, metabolism, and excretion, as defined in the main text as well as Table 2. Pharmacodynamics are distinguished further into primary (1 ) or efficacy-associated pharmacodynamic effects (i.e., intended consequences of drug treatment) and secondary (2 ), particularly body-harmful pharmacodynamic effects, that is, especially unintended consequences of drug treatment, i.e., side effects. It is usually the latter which represent pharmacologically emergent properties during drug development – pharmacologically emergent properties are difficult-to-predict properties of drugs that, as a consequence, tend to come to light only in the course of animal or clinical testing of drugs for safety. Unexpected toxicity can derail the development of a drug, though so too can in vitro or animal-testing-associated efficacy that fails to translate, for pharmacokinetic or pharmacodynamic reasons, into sufficient efficacy during clinical trials

gastrointestinal track, per os for systemic treatment, and various more direct means of introducing phages, or drugs, systemically including parenteral (directly into body tissue), where the latter includes intraperitoneally (IP, or into the peritoneum, that is, the abdominal cavity), intramuscular (IM), and subcutaneous (immediately below the skin). For further discussion of such details, see Abedon (2014a). An additional consideration is the potential to arm phages with, for example, homing peptides to enable their localization, following more systemic delivery, to infected tissues in densities which can better ensure efficient eradication of target bacteria (Górski et al. 2015).

Passive Treatment Versus Active Treatment In Figs. 1 and 2 we presented the interplay between (1) phage dosing, (2) phage therapy pharmacokinetics, and (3) phage therapy pharmacodynamics. In Fig. 3 we add the concepts of passive treatment, active treatment, and what can be described as active penetration. As considered in detail in the following subsections, phages often must penetrate through various barriers or into various compartments in order to reach bacteria. Ideally for phage therapy, this penetration is followed at

Bacteriophage Pharmacology and Immunology

311

the very least by bactericidal activity. Bactericidal activity typically is usually (though not necessarily) associated with bacterial lysis as well as the release of additional phage virions from the lysed bacteria. The result is some degree of auto dosing, that is, in situ increases in phage numbers due to in situ phage replication. Passive treatment depends solely on phage bactericidal effects. In other words, it requires only a combination of phage-virion penetration to bacteria and subsequent phage chemical modification (i.e., chemical modification as occurs in the course of phage adsorption and subsequent infection; these latter processes are aspects of phage metabolic activity and therefore, pharmacokinetically, of metabolism; Table 2). Active penetration, such as into bacterial biofilms, probably depends, minimally, on phage-induced bacterial lysis, another aspect of phage therapy pharmacokinetics and also as follows virion penetration (see chapter ▶ “Biofilm Applications of Bacteriophages”). Active treatment is dependent on the production of new virions, though also follows a combination of phage penetration and subsequent phage chemical modification. These various issues are discussed further in the following subsections. The terms “active” and “passive” can be semantically confusing (Abedon and Thomas-Abedon 2010). It may be useful therefore to think of these processes in the following terms: • Passive treatment is entirely dependent on phages that are supplied from outside of the site of treatment, that is, as via traditional dosing procedures such as by a clinician. It requires some degree of phage metabolic activity to achieve bactericidal effects but not so much metabolic activity that new virions necessarily are produced. It is “passive” in the sense that there is less metabolic activity on the part of infecting phages than is required to produce and release new phage virions. • Active treatment is dependent on the production of new phage virions as well as bactericidal activity. It is “active” in the sense that it requires “active” production of new phage virions as well as “active” lysis of phage-infected bacteria. • Active penetration is dependent, presumably, on the lysis of phage-infected bacteria but not necessarily dependent on the production of new phages by those infections. It is “active” because we at least assume that it requires “active” lysis of phage-infected bacteria. • Mixed passive-active treatment is dependent on bactericidal phage infections – which, as noted, can confusingly be thought of metabolically as more “passive” – but as augmented by “active” phage production by phage-infected bacteria. “Active” thus implies greater metabolic activity on the part of an infecting phage than “passive.” See Fig. 3 as well as Table 3 for summary.

Passive Phage Therapy Figure 4 serves as a modification of Fig. 2. These modifications make Fig. 4 more specifically a description of phage therapy. Here the term “Penetration” has been

A

•Phage Application/Dosing (to body) •Metabolism if due to auto dosing (PK)

B

•Absorption (into blood) (PK) •Penetration to vicinity of tissues (PK)

C

•Distribution (into tissues) (PK) •Penetration into biofilms (PK)

D

Sufficient phage #s: penetration consequence •Adsorption (to target bacterium) (PK) •Metabolism (phage chemical modification) (PK)

E

•Activation of Bactericidal Activity (PK) •Metabolism (phage chemical modification) (PK)

F

•Bactericidal Activity (PD) •Primary Pharmacodynamic Effect

G

•Activation of Bacteriolytic Activity (PK, PD) •Both Metabolism and Primary PD Effect

H

•Activation of Replication Activity (PK) •Metabolism (phage chemical modification) (PK)

Passive Treatment

Active Penetration

K. Dąbrowska et al.

Active Treatment

312

Return to “A”, “B”, or “C Ignored in figure are phage losses or therapy side effects. Fig. 3 Overview of phage therapy pharmacokinetics, primary pharmacodynamics, and different treatment approaches. From A through H are various sequential phenomena associated with the phage treatment of bacterial infections. Not all steps are seen in all treatment approaches nor are all steps always necessary for treatment success. Here efficacy is equated with eradication or at least killing of target bacteria and therefore, for phages and phage therapy, is equivalent to bactericidal effects (F). Both penetration (B and C) and metabolism (D and E) refer to various pharmacokinetic phenomena, with some overlap most notably at the point of phage adsorption. Passive treatment, at a minimum, requires only those steps necessary to effect bactericidal activity (A through F). Though in principle passive treatment involves penetration of sufficient numbers (#s) of phages that in situ production of new virions is not required for treatment success (H). Nevertheless, even passive treatment often can benefit from additional dosing (A). Active penetration, which is defined in terms of phage infection-mediated movement, such as into bacterial biofilms, is thought minimally to require bacteriolytic effects (G), but presumably can benefit as well from new virion production and release (H). Active penetration likely can benefit as well from subsequent phage dosing (A). Active treatment is dependent upon active, in situ phage replication (H) to supply sufficient phage numbers to result in treatment success, and it too can benefit from repeated dosing (A). Note that the absorption (B) and distribution (C) steps in particular are thought to contribute to primary pharmacodynamic effects only given systemic rather than more localized phage dosing

Bacteriophage Pharmacology and Immunology

313

Table 3 Summary of phage therapy treatment approaches Approach Passive treatment Active treatment Mixed passive-active treatment Active penetration a

Bactericidaldependent? Yes Yes Yes

Bacteriolysisdependent? No Yes No, but helps

Productiveadependent? No Yes No, but helps

Yes

Yes

Not necessarily

Productive refers to in situ phage-virion production, i.e., auto dosing

used to replace “Absorption” and “Distribution” (Table 2; Fig. 3). This has been done because both of these latter terms refer essentially to the movement of drugs from their point of administration, to patients, to their point of association with target tissues. For phages – whether to a localized bacterial infection, into the midst of a more systemic infection, or instead into bacterial biofilms – such movement often is described in terms that are equivalent to that of penetration, for example, Kutateladze and Adamia (2008). In addition, the term penetration potentially incorporates not just absorption and distribution but also an aspect of the pharmacokinetic concept of metabolism, that is, as associated with virion adsorption since this involves modification of the virion particle. The latter also represents the first step of phage activation towards bactericidal activity. Activation of bactericidal activity typically also involves phage-genome translocation into the bacterial cytoplasm, which at least arguably is also an aspect of phage penetration. Phage gene expression clearly, by contrast, is pharmacokinetically an aspect of metabolism rather than penetration. All three of these steps, however, are typically necessary for phages to effect bactericidal activity – adsorption, translocation, gene expression – and represent chemical or at least physical modification of the adsorbing or infecting phages. In Fig. 4, however, metabolism is not explicitly presented but instead is shown in terms of its consequences, that is, in the form of resulting “Activity,” keeping in mind that for passive treatment such activity must by definition be bactericidal. For a chemotherapeutic, metabolism in combination with excretion typically results in declines in drug presence within bodies. Hence, in Fig. 4 these two pharmacokinetic processes – phage inactivation due to chemical modification and phage removal from the body via excretion – have been replaced simply with “Losses” (note, though, that additional aspects of metabolism in phage therapy are also considered in this figure, as well as subsequent figures). More generally, pharmacokinetic processes control the concentrations of a drug that can reach targets, which is shown here for phages as “Sufficient #s.” This is sufficient numbers or densities of phages in the vicinity of target bacteria, particularly phages that have been metabolically “activated” (Abedon 2014b) in the course of phage adsorption of target bacteria in combination with subsequent phage infection of those bacteria. Sufficient numbers of phages is crucial for phage therapy success, or success of phage-mediated bacterial biocontrol more generally (Abedon 2008; Hagens and Loessner 2010; Abedon 2011a, c). Specifically, it is the density of biologically active

314

K. Dąbrowska et al.

Pharmacokinetics Penetration

Efficacy (1° PD)

Sufficient #s

Activity

Losses No Active Replication ≈ Passive Treatment (a.k.a., Inundative therapy)

Side Effects (2° PD) Fig. 4 Phage therapy pharmacology basics. Starting with Fig. 2, four modifications have been made to generate this figure. The first is that generally there is a relatively low association between those phages used in phage therapy and resulting treatment side effects. To explicitly acknowledge this lessening of secondary pharmacodynamic effects, the vertical arrow upon which the phage image is now superposed has been narrowed. More explicitly, this is a description of a lower tendency for well-characterized phages to display pharmacologically emergent properties, a.k.a., unexpected side effects. The second modification is actually a series which includes first the replacement of “Absorption” and “Distribution” simply with “Penetration,” which for phages is often a more apt term, and particularly so when referring to topical treatment as well as phage treatment of bacterial biofilms. In addition, “Metabolism” and “Excretion” have been replaced in part with “Losses,” that is, phage losses. Nonetheless, it is important to keep in mind that phage metabolism also can be associated in a pharmacokinetic sense with gains in phage function. Some of these gains are implicitly associated in the figure with the concept of penetration – as in, not only must phages reach target bacteria to have an impact on those bacteria, but they also must be metabolically activated by those bacteria, in the guise of phage-virion adsorption (which typically requires some degree of virion-particle rearrangement (chapter “▶ Structure and Function of Bacteriophages”)) and also phage-genome translocation into the bacterial cytoplasm. Third, the role of pharmacokinetics in impacting phage concentrations, as required for primary pharmacodynamic effects, is explicitly indicated as contributing to “Sufficient #s,” that is sufficient phage numbers, which can be interpreted also as sufficient numbers of bactericidal phage infections (sufficient phage numbers are also required for secondary pharmacodynamic effects, though this is not emphasized in the figure). Lastly, a column for phage “Activity” has been added, referring to the manner of impact of phages on target bacteria, particularly as mediated, pharmacokinetically, by metabolism. In this column, the concept of “passive treatment,” that is, “inundative therapy,” has been added. This is phage therapy in which phages act equivalently to chemotherapeutic antibacterials in the sense that there is no requirement for any phage activity beyond their antibacterial nature, which for lytic phages would inherently be bactericidal. Note nevertheless that phage replication in fact can occur within the context of passive treatment, though if such replication adds to phage antibacterial activity then the result may instead be referred to as a mixed passive-active treatment

agents, that is, titer for phages (Abedon 2016b), that tends to determine the magnitude of pharmacodynamic processes, whether primary or secondary. There are two additional changes going from Fig. 2 to Fig. 4 that involve pharmacodynamics. The simpler is a narrowing of the arrow pointing downward

Bacteriophage Pharmacology and Immunology

315

to secondary pharmacodynamics, indicating that toxicities/side effects tend to not be a substantial concern with phage therapy (to indicate that this is a phage-associated property, we have placed the image of a phage virion, that of phage T4, over this arrow). The second point, as also relevant to phage therapy pharmacodynamics, is that the extent of phage primary pharmacodynamic activity can be a function of the degree to which phages have been pharmacokinetically activated, particularly metabolically activated, hence addition of the term, “Activity,” as found in the box to the right in Fig. 4. Under this heading in this figure, the pharmacokinetically activated pharmacodynamic phage action is bactericidal. Phages thus are relatively safe (few secondary pharmacodynamic effects) and in fact that relative safety stems to a fair extent from this requirement that they be metabolically activated, particularly in the course of their intimate as well as irreversible interaction with target bacteria, prior to displaying cytotoxic activity. A “Productive” phage infection is one that does not result in either lysogeny, abortive infection, or lysis from without (Abedon 2011d). Confusingly, vis-à-vis phage therapy pharmacology, a productive phage infection also can be described as “Active” (above). That is, with an “Active” infection, phage progeny will be produced and virions released relatively soon after phage adsorption of a target bacterium. If active or productive phage infections are not required for successful phage therapy, then a therapy may be described as an inundative treatment or, equivalently, as a passive treatment (Payne et al. 2000; Payne and Jansen 2001). A complication on this latter idea, however, is that it is possible for passive treatment to take place even given active phage replication. Nevertheless, so long as that replication is not absolutely required to achieve therapeutic success, and in situ production of virions also does not otherwise improve the rapidity or likelihood of treatment success – and even if it is difficult to inundate target bacteria with phages using a single phage dose – then that treatment still can be described as passive/inundative. Passive treatment requires only a minimal degree of pharmacokinetically associated metabolism: activation of phage bactericidal activity.

Active Penetration In addition to phage action against individual bacteria, we can also consider phage penetration into bacterial biofilms (chapter “▶ Biofilm Applications of Bacteriophages”). This penetration at a minimum may require phage-induced bacterial lysis versus solely bacteria killing (chapter ▶ “Phage Infection and Lysis”). Lysis-induced stripping away of bacteria, such as from the surface of bacterial microcolonies or biofilms, can be described as at least one aspect of phage “active penetration” into those structures (Abedon and Thomas-Abedon 2010). This represents a pharmacodynamic impact (bactericidal activity) that is followed by yet another form of pharmacokinetic metabolism, in this case bacterium-associated biochemical modification of the phage “drug” that leads to a phage-induced bacterial lysis. In between these two pharmacokinetic processes is the pharmacodynamic phage killing of target bacteria. This phage-induced bacterial lysis is presented as

316

K. Dąbrowska et al.

Pharmacokinetics Penetration

Efficacy (1° PD)

Sufficient #s

Activity

Losses Bacterial Lysis ≈ Active Penetration

Side Effects (2° PD)

Bacterial Lysis

Fig. 5 Pharmacology of phage-therapy active penetration. Modifying Fig. 4, this figure incorporates the pharmacokinetic property of phage-induced bacterial lysis. Pharmacokinetically, this is a consequence of metabolism in that it represents chemical modification of phages as drug equivalents, in this case phage-directed but target bacterium-associated conversion of a bactericidal infection to a bacteriolytic one. The resulting lysis can contribute to what can be described as an active penetration, particularly of phages into bacterial biofilms, a phenomenon which at a minimum likely involves phage-induced bacterial lysis. This lysis might strip away outer layers of biofilm bacteria so that phages can either gain access to underlying bacteria or so that previously underlying bacteria can gain better access to nutrients and thereby change physiologically so that they are better able to support bactericidal and/or bacteriolytic phage infections. The added diagonal line indicates the potential for bacteriolytic phage infections to supply new virions, which could contribute to further phage penetration into biofilms. The diagonal line is dashed, however, to indicate that active penetration in principle may not require in situ phage production as phages might be supplied to previously underlying bacteria instead via standard dosing with additional phage virions. Lastly, bacterial lysis can impact the body more generally, particularly as a consequence of the release of potentially toxic bacterial lysis products. Consequently, a second horizontal arrow has been added (bottom), though for phage therapy purposes the resulting side effects in most instances generally have not been found to be severe, hence the narrow gauge of the resulting arrow

the vertical arrow found on the right of Fig. 5. Thus, pharmacokinetics gives rise to both phage bactericidal and lytic activity, while these processes in turn give rise, respectively, to bacteria killing and potentially also to further phage penetration into bacterial biofilms. The “active” aspect of “active penetration,” as noted, refers to phage-induced bacterial lysis as stemming from some approximation of “active” phage replication. Bacterial lysis may facilitate further phage penetration into biofilms, hence the dashed diagonal arrow added to Fig. 5, connecting the lower-right quadrant with the upper-left. This line is dashed because lysis would contribute only indirectly to further phage penetration into biofilms, or into bacterial microcolonies, should phage release happen to not follow phage-induced bacterial lysis, for example, as associated with abortive infections rather than phage-productive ones. Following bacterial

Bacteriophage Pharmacology and Immunology

317

death, bacteria underlying the biofilm surface may become more physically available for virion adsorption, and/or nutrients (or oxygen) may become more physically accessible, thereby contributing to improvement in the physiological potential for those bacteria to support phage infections. Nonetheless, unless lysed bacteria supply new phages then phage infections themselves are not contributing directly to subsequent phage penetration. That is, theoretically subsequently penetrating phage virions might instead be supplied exogenously via further phage dosing rather than supplied in situ via auto dosing. Thus, this process of active penetration can be viewed as requiring greater antibacterial action than solely bactericidal activity, but at a minimum this additional activity must be phage-induced bacterial lysis – we suggest – rather than explicitly requiring virion production by those same infections as well. Purely passive treatment mediated by bactericidal but not replicationcompetent phages, in other words, could conceivably be employed to destroy bacterial microcolonies or biofilms, but we speculate that this destruction may be more efficiently achieved if resulting phage infections are associated not just with bacterial killing but with lysis of targeted bacteria as well. To our knowledge, however, that hypothesis has not yet been tested. It is important to emphasize that phage induced bacterial lysis in the absence of virion release, though possible (e.g., as a form of abortive phage infection), in fact is probably not typical for phage infections during phage therapy. Instead we consider this possibility here predominately to illustrate the point that in principle phages during phage therapy may be able to actively penetrate into bacterial biofilms without necessarily productively infecting the bacteria that they are attacking (i.e., producing new phage virions). Successful active penetration in the absence of active virion production presumably could occur, however, only given further phage application, that is, repeated dosing in the course of treating biofilm-associated bacterial infections. Such repeated dosing may be necessary even with in situ phage production if burst sizes are small or, instead, if phages have a low potential to immediately productively infect underlying bacteria, that is, such as if those bacteria in fact are in a stationary phase-like state (Abedon 2015c, 2017d; Bryan et al. 2016). Phage-induced bacterial lysis could potentially result in side effects that are in addition to phage application alone. As a consequence, we have added the lower horizontal arrow to Fig. 5. In practice, however, substantial side effects have not been observed with modern phage therapy, hence the narrowness of that arrow as well as the superimposed image to indicate that this low potential for side effects is relatively phage specific. It is important as well to point out that phages are not the only bacteriolytic antibacterial agents that can be employed against bacterial infections, as cell wall disrupting and therefore lysis-inducing antibiotics are common. It is possible, however, that phages are better equipped in terms of their ecological properties to penetrate into biofilms than necessarily are naturally occurring antibiotics (Abedon 2015b), though phages likely still are limited in that ability (Abedon 2016a, 2017d), thereby necessitating multiple dosing in the actual practice of phage therapy against biofilms and/or the use of biofilm-disrupting agents (such as extracellular polymeric substance-disrupting depolymerase enzymes). For more on phage

318

K. Dąbrowska et al.

interactions with bacterial biofilms, microcolonies, or cellular arrangements including within a phage therapy context, see Abedon (2011b, 2012c), (2015c), (2016a), (2017a), Brussow (2013), Fan et al. (2013), Harper et al. (2014), Parasion et al. (2014), Sillankorva and Azeredo (2014), Chan and Abedon (2015), Gutierrez et al. (2016), Khalifa et al. (2016), and Motlagh et al. (2016) as well as (chapter “▶ Biofilm Applications of Bacteriophages”).

Active Phage Therapy Active penetration of phages into bacterial biofilms or microcolonies potentially involves – or indeed requires – phage-induced bacterial lysis, as presented in Fig. 5. In addition, an active treatment, as presented in Fig. 6, is also possible. With passive treatment, the necessary metabolism aspect of pharmacokinetics is limited to the transformation of a relatively inert phage particle into a bactericidal phage infection. With active penetration, metabolism at a minimum likely must involve the transformation of a phage virion into an infection possessing both bactericidal and bacteriolytic activities. With active treatment, by contrast, phage association with target bacteria must result not only in the metabolic development of both bactericidal and bacteriolytic activity, but productive phage infection as well, that is, the production of new phage virions. Active treatment, in other words, is phage treatment that requires in situ amplification of phage numbers, that is, auto dosing. This activity, pharmacokinetically, is also a consequence of metabolism, that is, further chemical modification of the initially dosed phage “Drug.” This aspect of phage therapy pharmacokinetics feeds back into penetrative aspects as well as to subsequent phage losses. This is indicated in Fig. 6 as a diagonal arrow, now solid as well as overlain with a phage image because by necessity, with active treatment, new phage virions are being supplied in situ. So long as the distances virions must travel to reach new target bacteria are small, however, then neither phage losses nor phage penetration, particularly in terms of absorption and distribution, may play large, further roles in impacting “Sufficient #s,” this despite auto dosing having a positive impact on phage densities and thereby on phage therapy efficacy at more local scales. In other words, active treatment, once it is initiated locally, will tend to continue to act predominantly locally, with local production of additional phages resulting in further local reductions in numbers of target bacteria, which in turn will give rise to further local phage production (Abedon 2017a). As in situ amplification of phage numbers is not thought to directly result in substantial increases in phage therapy side effects, the bottom, horizontal arrow in Fig. 6 remains thin in this figure as well.

Summarizing Phage Therapy Pharmacological Phenomena In terms of activity, a phage infection can, for instance, be not bactericidal. One example of nonbactericidal phage infections is restrictive infections, such as

Bacteriophage Pharmacology and Immunology

Pharmacokinetics Penetration

319

Efficacy (1° PD)

Sufficient #s

Activity

Losses Active Replication ≈ Acve Treatment

Side Effects (2° PD)

In Situ Amplification

Fig. 6 Pharmacology of active phage therapy. With active treatment, the “PK” found in the lower-right quadrant refers not just to the metabolic conversion (metabolism) of a particle to a lytic infection but to a phage-productive one as well. The resulting phage particles, as represented by the diagonal arrow, are then subject to distribution, to further metabolism (both activation and inactivation), and potentially also to excretion. Ideally, so far as phage therapy is concerned, this in situ amplification results in an increase in phage density that, in turn, results in a greater likelihood of phage adsorption to target bacteria and thereby greater phage therapy efficacy. This process is termed active treatment or active therapy and contrasts with passive treatment (Fig. 4), but can also work in conjunction with lysis towards an active penetration into bacterial biofilms (Fig. 5)

resulting from the action of restriction endonucleases or, alternatively, as mediated by bacterial CRISPR-Cas systems (Hyman and Abedon 2010; Labrie et al. 2010; Abedon 2012a; Dy et al. 2014; Seed 2015). In this case, phage therapy can fail not because there is insufficient phage penetration to target bacteria, or insufficient absolute phage numbers, but instead due to insufficient phage bactericidal activity. Alternatively, it is possible for phage infections to be nonproductive but still bactericidal, which can be described as forms of abortive infections. Such abortive infections, as they result in bacterial death, can support passive therapy/inundative treatment (Fig. 4). Abortive infections as so defined may or may not also result in bacterial lysis or, alternatively, can result in bacterial lysis without substantial infection, that is, as seen with the phenomenon known as lysis from without (Abedon 2011d). These infections, if they are bacteriolytic but not necessarily phage productive, may be able to support an active penetration of phages into biofilms (Fig. 5), though if so constrained then active penetration may occur only if additional phages are supplied via repeated phage dosing. An ability of phage infections to both lyse bacteria and release phage progeny, however, may result in more robust penetration into bacterial biofilms. Such a process, to the extent that it is dependent on in situ phage production, would qualify as both an active penetration and an active treatment (Fig. 6).

320

K. Dąbrowska et al.

We summarize in Fig. 7 these three levels of increasing phage activity that can play important though distinct roles in phage therapy success. Thus, there are bactericidal infections (a minimum level of activity necessary to achieve some degree of efficacy), bacteriolytic infections (a minimum level of activity necessary to achieve active penetration), and productive infections (a minimum level of activity necessary to support active treatment). To achieve any of these ends, insufficiencies in phage penetration, as phage movement towards target bacteria, Pharmacokinetics Penetration

Efficacy (1° PD)

Sufficient #s

Activity

Losses Bactericidal passive treatment

---------------------> Bactericidal active penetration

---------------------> Bacteriolytic active treatment

Side Effects (2° PD) Fig. 7 Comparing passive treatment, active penetration, and active treatment in terms of phage-infection activity. Bactericidal but not bacteriolytic phage activity (“Bactericidal” but not “> Bactericidal”) corresponds to passive treatment (see also Fig. 4; the “>” symbol in this figure means “greater than,” for example, “greater than just bactericidal activity” or “greater than just bacteriolytic activity”). Bacteriolytic but not phage-productive activity (“> bactericidal” but not “> Bacteriolytic”) corresponds to what can represent a minimum of activity necessary for phages to effect an active penetration into bacterial biofilms (see also Fig. 5). For bactericidal along with bacteriolytic activity to be effective in the course of phage therapy, in the absence of productive phage infection (i.e., lacking in active infection/auto dosing/secondary infection), then the phages required to infect and kill additional bacteria must be supplied from outside of the area under treatment, particularly via traditional approaches to dosing. That is, in the absence of in situ amplification in phage numbers or instead given only weak amplification, then multiple, repeated phage dosing may be required to achieve eradication of target bacteria. Fully phage productive infections (“> Bacteriolytic,” meaning more than simply bacteriolytic) form the basis for active treatment, which explicitly involves the generation of secondary phage infections, that is, phage infections by phage particles that have been generated in situ, i.e., as following active phage replication (auto dosing; see also Fig. 6). This generation of phage particles in active treatment leads directly to secondary infections, which in the figure are indicated via the diagonal arrow superimposed with a phage particle. Note that active penetration and even otherwise passive therapy, that is, inundative treatment, may still be aided by in situ phage amplification/secondary infection, so-called mixed passive-active treatment. Furthermore, active treatment is absolutely dependent on sufficient in situ phage amplification for treatment success to occur. Such treatment thus can fail if phage burst sizes are insufficient, if phage inactivation rates are too high, or instead if bacterial targets are not sufficiently plentiful to support in situ phage population growth to what essentially must be inundatory phage densities for phage therapy to be successful

Bacteriophage Pharmacology and Immunology

321

and otherwise insufficiencies in delivery of adequate phage numbers to target bacteria may potentially be augmented via multiple, repeated phage dosing. Note that phage infections that are a direct consequence of such dosing can be described as primary phage infections, whereas phage infections that occur only following in situ phage replication, a.k.a., active replication or auto dosing, instead can be described as secondary phage infections (Abedon 2015a). With secondary infection defined explicitly in this latter, epidemiological sense (Payne et al. 2000; Payne and Jansen 2001; Wei and Krone 2005), then active treatment in phage therapy can be defined as a form of treatment that, as a consequence of dosing with fewer phages than would be necessary to achieve passive treatment, therefore happens to require secondary phage infection.

Phage Interactions with Immune Systems Bacteriophage interactions with mammalian immune systems have both pharmacokinetic and pharmacodynamic consequences for phage therapy. There are two major reasons for this. First, in terms of pharmacokinetics, with therapeutic application of phages into a living organism, the immune system can serve as a crucial component of a phage’s environment (Górski et al. 2012; Hodyra-Stefaniak et al. 2015). Indeed, for the phage as essentially an “invading” microorganism, the immune system represents the primary means by which body tissues, contrasting microflora, impact phages – the “immune systems” of target bacteria are also relevant, though beyond the scope of this chapter; see instead Abedon (2012a) for discussion of the bacteria “immunity” against phages. These interactions with the body’s immune system, from a pharmacokinetic perspective, largely result in phage losses, leading to virion inactivation (pharmacokinetically a consequence of metabolism). Phage elimination from the body as whole virions, such as via the kidneys, represents by contrast a pharmacokinetic mechanism of excretion, but generally this occurs absent immune system function. The phage, in situ, thus is constantly in contact with elements that compose the body environment, and these elements, in our own bodies, are dominated – in terms of body-phage interactions – especially by the actions of the mammalian immune system. The second impact of phage interactions with immune systems during phage therapy stems from the importance of immunity for the health and homeostasis of organisms. This role involves multiple components and requires substantial flexibility to a wide range of possible reactions and interactions. The responses of immune systems, as a consequence, are not solely directed against foreign objects, but also towards the immune system itself, resulting in modifications of its own functions, so-called immune system modulation (immunomodulation). The immune system in particular can react to stimuli, including of the body’s own making, that can have positive as well as negative effects on homeostasis. Such stimuli can be provided by bacteriophages as well, thereby contributing to phage-associated pharmacodynamics. As with phage antibacterial activity, these pharmacodynamic effects can be both positive and negative with regard to overall body health, but explicitly are

322

K. Dąbrowska et al.

pharmacodynamic because they represent phage impact on the body rather than (strictly) body impact on the phage. These effects can also occur independently of phage interaction with target bacteria and thereby are to a degree independent of phage auto dosing, though phage amplification in situ can result in enhanced phage immunomodulatory effects, as too can phage modification of target bacteria, particularly in terms of bacterial lysis.

Overview of Immunity Although the first studies on the interactions between bacteriophages and the immune system were conducted by Felix d’Hérelle shortly after the discovery of bacteriophages (Górski et al. 2012), we still do not have a complete picture of all elements and factors that play roles in phage-mediated modulation of immunity. The incompleteness of this picture stems in part from immune responses to phages resulting from the action of a variety of elements of the immune system as well as both intraspecific and interspecific variation in immune systems functions, but also because different phages will interact with the different components of immune systems in different ways. The most general classification of those immune system elements, however, comprise innate versus adaptive immune responses, that is, innate immunity and adaptive immunity, with major differences existing between these two types both generally and in terms of phage-immune system interactions. Innate immunity represents nonspecific reactions to foreign objects that are usually recognized by universal molecular patterns associated with pathogens (Pathogen-Associated Molecular Patterns, or PAMPs), for example, bacterial DNA (CpG), peptidoglycan, or lipopolysaccharide. Innate immunity reaction to detected pathogens is related to inflammation, which is one of the first responses of the immune system to infection. This type of response engages a wide collection of leukocyte types, that is, white blood cells. These are mainly the phagocytes: monocytes, macrophages, neutrophils, tissue dendritic cells, and mast cells, but also eosinophils, basophils, and natural killer (NK) cells. Collectively, these cells are capable of recognizing and eliminating pathogens. They are also important mediators in the activation of the adaptive immune system. Acellular and therefore humoral elements of innate immunity are represented mainly by the serum complement system (Rus et al. 2005; Kawai and Akira 2006; Medzhitov 2007). Adaptive immunity responses are executed by specific elements of the immune system that are distinct from but nevertheless communicate extensively with innate aspects of the immune system. “Adaptive” generally refers to the requirement of adaptive immunity that it must first be “trained” in order to gain an ability to robustly recognize and, especially, respond to “foreign” invaders of the body. The adaptive immune system further asserts immunological memory, that is, retention of enhanced abilities to recognize and respond to these invaders. Adaptive immunity is specifically directed to selected objects, often involving what can be described as at least an approximation of lock-and-key matching. This matching involves spatial

Bacteriophage Pharmacology and Immunology

323

and chemical complementarity between the antigens (more precisely, the epitopes) of foreign substances, on the one hand, and specific immune system molecules on the other. The latter consist particularly of various highly diverse receptors associated with lymphocytes, which are the key adaptive-immunity effecting leukocytes, along with antibodies. As a consequence of the resulting specificity of interactions between antigens and these immune system molecules, adaptive immunity is also described as specific immunity. Major types of lymphocytes are B cells and T cells. B cells are involved in the specific humoral immune responses, that is, in the production of antibodies. T cells are involved in cell-mediated specific immune response (Janeway et al. 2005; Pancer and Cooper 2006). Importantly, innate immunity involving nonlymphocyte leukocytes and adaptive immunity elements involving B and T lymphocytes are tightly linked into one, consistent system: they cooperate, they stimulate, and/or they control each other, depending on individual situation and needs. Here we consider these complementary interactions with regard to the impact of bacteriophages on immune system functioning, especially with regard to phage virions.

Adaptive Immunity Specific antibodies to virions have been by far the most often investigated and acknowledged part of mammalian immunity engaged in immune reactions to phages, with studies dating back to 1920s, for example, Muckenfuss (1928). Serological cross-reaction represented the earliest criteria for bacteriophage classification, beside host range, particularly of phages into serologically related groups (Stent 1963). Since the intensity of immunological reactions decreases in correlation with morphological and biological differences between phages, serological classification greatly improved upon the host range-based classification of bacteriophages. Those early studies were based on a passive agglutination test for determination of neutralizing activity of bacteriophages by serum. The ability of bacteriophages to induce specific antibodies was also one of the earlier useful properties of these viruses. More recently, phage immunogenicity has been employed in medicine as a test for the immunocompetence of otherwise immunodeficient patients. In this test, HIV-infected patients can be monitored for the ability of T cells to provide help to B cells in antibody production, amplification, and isotype switching after phiX174 phage immunization (Fogelman et al. 2000). Antibody titer is anticipated to grow as a result of repeated medical application of phages, for example, as during phage therapy. The presence of antibodies within an individual’s serum against a specific bacteriophage, however, may result from previous, natural contact. These are so-called natural antibodies (Hajek 1967). Since phage-specific antibodies can cross-react with related phages, natural antibodies do not need to be induced by exactly the same phage strain but instead can result from previous contact with one or more related, especially serologically similar phage. In practical terms, natural antiphage antibodies may indicate that an individual has undergone a bacterial infection caused by a given host bacterium, for

324

K. Dąbrowska et al.

example, patients with staphylococcal infections have been found to demonstrate higher titers of antibodies specific to staphylococcal phages in comparison to healthy blood donors (Kucharewicz-Krukowska and Slopek 1987). The other inducers of natural antibodies can potentially be environmental phages that enter the body with food, water, etc., and/or which propagate on symbiotic bacteria such as those present in the gut or other parts of the body. In most cases, however, the true stimulator of the production of antiphage natural antibodies remains obscure. The frequency of natural antibodies in human populations is also difficult to estimate and most probably depends on many factors, including phage type and individual characteristics of investigated patients, including their individual microbiome. The frequency of antibodies specific to staphylococcal phages in a group of patients before phage therapy, that is, suffering from staphylococcal infections, was reported as approximately 23% (Kucharewicz-Krukowska and Slopek 1987). This high frequency suggests a role as well for gut-associated phages in stimulating antibody responses, and indeed the frequency of antibodies specific to T4 coliphages reached 81% in populations of healthy volunteers (Dąbrowska et al. 2014). Animal models have demonstrated that systemic administration of phages can effectively induce specific antibodies (Huff et al. 2010; Smith et al. 1987). Among the main classes of antibodies – IgM, IgG, IgA, IgE, and IgD – only IgM, IgG, and IgA were demonstrated to be induced by bacteriophages. The schema of antibody production in response to a challenge with a phage, however, seems very typical for antigen challenge generally: maximum IgM production can be observed within 5–10 days after the challenge and this is followed by the increase of IgG that tends to persist for a longer time (Hodyra-Stefaniak et al. 2015), see Fig. 8. Though not so far demonstrated to be associated with humoral responses to phage virions, one should not exclude the possibility that phage-specific IgE and IgD will be detected in future studies, particularly under conditions appropriate to induce these classes of antibodies. Not surprising, given their diversity, individual phage types differ in their antigenicity. It is unclear to what degree differences in phage antigenicity “map” onto measures of phage diversity, however, as not even moderately comprehensive analyses have been undertaken. Nevertheless, individual structural proteins forming phage particles can differ strongly in their ability to induce specific antibodies (Dąbrowska et al. 2014). This is consistent with the expectation that the molecular composition of individual phage capsid proteins should determine much of phage reactivity with the immune system. It may have further implications for the possibility to select, or even to construct phages with the molecular composition of a desired immune reactivity, such as construction of phages that are less “visible” to the immune system so as to be less efficient in induction of antibodies specific to the phage and thus neutralized less rapidly (note that this idea is distinct from modification of phage interaction with innate immunity, which is covered below). Decrease in the pace of neutralization can also be achieved by a chemical “cover,” for example, PEGylation (Goodridge 2010). Alternatively, otherwise similar phages with dissimilar serological properties in principle may be constructed to allow phage

Bacteriophage Pharmacology and Immunology

325

0,8

OD by ELISA

0,6

0,4

0,2

0,0 0

10

20

30

40

days anti-phage IgM

anti-phage IgG

Fig. 8 Induction of specific IgM and IgG antibodies in mice challenged with phage parenterally. Mice were injected subcutaneously with a Myoviridae phage (F8), 1010 plaque-forming units (chapter ▶ “Detection of Bacteriophages: Phage Plaques”) per mouse on day 0 and again on day 22. Antibodies specific to the phage were detected in blood by ELISA. Observed is a typical increase of IgM as a primary response which is followed by a typical increase of IgG (secondary response). Figure provided by K.D as based on unpublished data

switching during treatments should phage interaction within immune systems come to interfere with treatment success. There have been many antiphage antibody studies over the decades, but other types of specific immune responses to phages have been much less explored or appreciated. Specific cellular response to phages would engage T lymphocytes, since they play a central role in cell-mediated immunity. These lymphocytes are distinguished from other immunological cells by the presence of a T-cell receptor (TCR) on their surfaces, which is responsible for recognizing antigens bound to major histocompatibility complex molecules as found on vertebrate animal cell surfaces. Interestingly, the interaction between TCR and an antigen is of relatively low affinity and low specificity compared to antibodies produced by B cells; one TCR can recognize many antigens and conversely, T cells capable of recognizing foreign antigens develop via clonal selection before they can act as specific immunological cells. Langbeheim et al. (1978) provides some indication of the ability of a phage (MS-2) to induce a cellular response. It was evaluated in presensitized mice, by intradermal injection of the test antigens, resulting in local erythema and induration. Strong in vivo reactions to the injected phages were reported. Induction of a cellular response was also determined in vitro by measuring proliferative responses of lymph node cells to the test antigens. Srivastava et al. (2004), by contrast, demonstrated that the clearance of phages in the blood was similar in normal (control) and T-celldeficient mouse strain C57BL/6 J-Hfh11nu, implying that T cells in fact did not play a significant role in the inactivation of these phages in vivo.

326

K. Dąbrowska et al.

Innate Immunity Contrasting the predominant role of humoral immunity via antibodies in adaptive immune responses to phages, phagocytes are the major players in innate immunity to phages. Very early studies showed that the reticuloendothelial system of the liver and, predominantly, the spleen – a.k.a., the mononuclear phagocyte system – filters bacteriophages from circulation. Bacteriophages are phagocytized by Kupffer cells which are stellate macrophages present in the liver, splenocytes (mainly macrophages), or by peritoneal macrophages. Neutrophils have also been shown to be capable of engulfing phages (Kantoch 1958; Inchley 1969; Geier et al. 1973; Górski et al. 2012). It has been demonstrated that individual characteristics of phage capsids may determine phage reactivity not just with adaptive immunity but also with nonspecific, that is, innate parts of the mammalian immunity. Merril et al. (1996) used a serial passage scheme to isolate phage mutants possessing reduced sensitivity to filtration and elimination from the blood. These mutants were able to remain in the circulatory system for longer periods of time, and it was shown that substitution of a single amino acid in the major capsid protein was enough to achieve this “long circulating” phenotype. Sokoloff et al. (2000) engineered phage particles to achieve equivalent long-circulating phenotypes by introducing peptides with C-terminal lysine (rat model), arginine (rat model), or tyrosine (humans); these modifications resulted in lower sensitivity to antibody-dependent complement-mediated neutralization. These examples suggest again that phage reactivity with the immune system can be modified by changes in the molecular composition of phage particles. Phages have been shown to interact with nonadaptive humoral elements of innate immunity as well. The complement system, for example, consists of a number of small proteins that form a cascade leading to the activation of cell-killing membrane attack complexes as well as other antipathogen functions. This complex is targeted to invasive bacteria as well as to viruses (Perreau et al. 2007). Although bacteriophage particles are substantially unlike most eukaryotic viruses, particularly in that most phages lack a phospholipid envelope derived from host cells, these “naked” phage virions nevertheless seem to be sensitive to actions stemming from the complement system. The most probable antiphage action of complement therefore likely requires antigen-antibody complexes (immune complexes) for activation, implying a link between nonspecific and specific immune response to phages (Sokoloff et al. 2000; Dąbrowska et al. 2014; Hodyra-Stefaniak et al. 2015). As noted, the typical mechanisms of nonspecific immunity center on inflammation, which is a complex response that involves immune cells, blood vessels, and molecular mediators (cytokines). Cytokines are small signaling proteins and some of them are typical markers of inflammation, for example, IFN-gamma, TNF-alpha, Il-1, and Il-6. These markers rapidly increase in prevalence in response to the presence of PAMPs as indicators of pathogen invasion. In contrast to pathogens, including pathogenic viruses, phages do not induce inflammation markers in vivo. This corresponds to the fact that phages do not induce production of reactive oxygen species (ROS), which is a characteristic marker of activation of phagocytes

Bacteriophage Pharmacology and Immunology

327

(Miernikiewicz et al. 2013; Park et al. 2014). In that respect, bacteriophages should be discriminated from crude phage lysates that contain multiple PAMPs released during bacterial lysis. These PAMPs include peptidoglycan, lipopolysaccharide, and other bacteria-associated molecules and have strong pro-inflammatory activity. Since these PAMPs are typical bacterial lysis products, highly purified phage particles would lack such inflammatory characteristics (Miernikiewicz et al. 2013). It should be noted, however, that inflammatory markers (CRP, sedimentation rate, leukocytosis) in patients treated with phage lysates may decrease significantly during the treatment (Międzybrodzki et al. 2009).

Effects of Immune Response on Phage Pharmacokinetics In the case of organisms and materials that are foreign to the body, such as phages, the major role of the mammalian immune response is neutralization and/or removal. Thus, immune responses against phage particles generally will have a limiting effect on phage activity in vivo. Considering phage therapy pharmacokinetics in mammals, the relationship of variables representing concentrations of active phages tends to be inversely associated with variables representing the intensity of antiphage immune responses (Dąbrowska et al. 2014). These are specific antibodies, phagocyte activity, and production or activation of other immunological elements that are able to neutralize phage particles. These elements may differ in their individual contribution to the intensity of the neutralization, and this contribution may also significantly differ for different phages or circumstances. Confirmation of the ability of specific antibodies to negatively impact phage functionality comes from their impact on phage antibacterial potential in vivo (Dąbrowska et al. 2014, Hodyra-Stefaniak et al. 2015). This is in line with the role of antibody-producing B cells in phage-T7 clearance from the blood of mice as was highlighted by Srivastava et al. (2004). The presence of antiphage antibodies in serum has been postulated to be an important factor potentially limiting phage therapy effectiveness (Carlton 1999; Sulakvelidze et al. 2001). Often the fact that antibodies are demonstrably capable of blocking phage antibacterial activity in vitro is interpreted as a strong argument that they can neutralize phages in vivo. Accordingly, animal models usually show that preimmunization with a phage can block phage therapeutic activity (Smith et al. 1987; Huff et al. 2010). Carlton (1999) suggested that negative impacts of immune-system inactivation of phage virions might be overcome by applying more phages. A recent report of Łusiak-Szelachowska et al. (2014), however, shows that simple conclusions on the effect of immunization on the efficacy of phage therapy may be overly simplistic. Antiphage activity in patients’ sera depended on the route of phage administration and phage type, and high antiphage activity of sera during phage therapy was observed in only 12.3% patients receiving phages. Moreover, the induction of antiphage activity in sera during or after phage therapy did not interfere with the antibacterial efficacy of treatments. Also, the first small safety test of T4 phages applied orally to humans as reported by Bruttin and Brüssow (2005) revealed

328

K. Dąbrowska et al.

no T4-specific antibodies in the serum of subjects. Comparison of phage therapy studies in animals – which often conclude with significant increases in antiphage antibodies – to human clinical studies reveals that in most cases phage doses (total plaque-forming units per kg body weight) as applied to animals were much higher than those tested in humans (Letarov et al. 2010; Majewska et al. 2015). Notably, the routes of phage administration were also usually different in humans versus animals: oral or topical in humans, but parenteral (intraperitoneal, intravenous, subcutaneous, and intramuscular injections) in animals. These differences may result in discouraging conclusions from animal studies that, in fact, do not fully correspond to reports from human studies (Międzybrodzki et al. 2012; Łusiak-Szelachowska et al. 2014). The second powerful way that phages can be inactivated within bodies, as noted above, is via phagocytosis, which has been demonstrated as playing an important role in affecting circulating bacteriophages in vivo. Phagocytes engulf phage particles as foreign objects and the process seems to be rapid, that is, within minutes of phage delivery to the blood. Phage disruption inside phagocytic cells usually takes place within 15 min to 2 h (Aronow et al. 1964; Kantoch 1958). Two studies reported that neither macrophages nor neutrophils are substantially involved in the clearance of phages from blood after systemic administration (mouse models) (Srivastava et al. 2004; Uchiyama et al. 2009; Górski et al. 2012), but other studies suggest that phagocytes may have greater capabilities to neutralize phages when these cells are activated, for example, by typical PAMPs during bacterial infection (HodyraStefaniak et al. 2015). Phage therapy also has been found to not impair the killing ability of patients’ granulocytes and monocytes and may even upregulate this ability (Jończyk-Matysiak et al. 2015). In conclusion, with regard to immune system impact on phage pharmacokinetics, that is, the ability of phages to reach and sustain sufficient densities within the vicinity of target bacteria, phage sensitivity to immune system action probably also depends on specific conditions, for example, particularly phage dose. For T2 phages an experimental “cut-off value” of 107 plaque-forming units/ml was observed, that is, under this value the effect of phagocytosis on phage concentration was not detected (Kantoch 1961). Other important factors can be the immune status of patients (or model animals), which can be of key importance particularly in terms of understanding immune system impact on phage prevalence and persistence within animals. Differences in properties are likely to be seen between dissimilar phage types, however, greatly complicating the study of phage-immune system interaction.

Relation of Immune Response to Phage Pharmacodynamics Nonbactericidal pharmacodynamic effects of bacteriophages in mammals can be related to immunomodulatory effects of phage preparations. There are a few distinct aspects that need to be considered. First, “phage effects” should be separated from “phage preparation effects” since phage lysates can exert important effects that are mediated by bacterial components only, with no direct role of associated phage virions. Second, some effects can be truly attributed to phages, but they are not

Bacteriophage Pharmacology and Immunology

329

caused by the direct impact of virions on immunity but instead are associated with phage antibacterial activity. Phages, by killing bacteria, can normalize immune system stimulation, which has otherwise been boosted by bacterial infection (Górski et al. 2012). Phages also can interact with components of bacteria that have been released into the larger organism. Coliphages and their proteins, for instance, can be capable of binding to LPS found in solution, not only that incorporated into bacterial outer membrane (Miernikiewicz et al. 2016). Only very few effects can be conclusively attributed to phages and their direct interaction with body tissues, however, with no infectivity or toxic effects on mammalian cells reported (Merril 2008). This fact is crucial for tolerance by the body to phages, that is, with secondary pharmacodynamics effects (toxicities, side effects) that are attributable to phage virions themselves at worse only negligible in nature (Bruttin and Brüssow 2005; Kutter et al. 2010; Abedon et al. 2011; Abedon 2015d); see also (Curtright and Abedon 2011). As noted, mammalian phagocytes may engulf phage virions. During a bacterial infection, phagocytes also are expected to engulf and effectively kill infecting bacteria, which is an important part of innate immune defenses. We therefore may ask whether phages might interfere with the intracellular killing of bacteria by phagocytes, hypothetically, by modifying phagocyte activity, particularly cytokine release and endocytic capabilities. Intracellular killing of bacteria is one of the fundamental immune responses against invading pathogens, and thus, possible effects of phages on this mechanism would be of importance to therapeutic approaches (Jończyk-Matysiak et al. 2015). According to recent observations, however, phage therapy did not reduce the ability of patient phagocytes to kill bacteria and phagocytosis of phages does not affect the activity of phagocytes in patients who initially displayed a reduced ability to kill bacteria intracellularly. These observations support observations that experimental phage therapy seems to have no significant adverse effects on patient immune system functions (Jończyk-Matysiak et al. 2015). Nonbactericidal phage pharmacodynamic effects that can be observed are usually indirect, that is, they result from the tripartite interactions of mammalian immunity, phages, and especially target bacteria. A good example of this complex effect is inhibition of respiratory burst (ROS production) in peripheral blood polymorphonuclear leukocytes (PMNs) as stimulated by LPS serving as a standard activator. Significant reductions in such leukocyte activation were observed that probably were due not only to phage-PMN interactions but also to phage-LPS interactions (Międzybrodzki et al. 2008). Interestingly, some antiviral effects mediated by phages have also been described, as T4 phage was shown to inhibit attachment and replication of adenovirus (Przybylski et al. 2015). In general, positive results of phage treatment should be related to the normalization of immunological parameters through decreases in inflammation. Such antiinflammatory effects of phages have been reported earlier in animals as well as patients receiving phage therapy (Górski et al. 2006a; Międzybrodzki et al. 2009). Phage lysates that have not been purified prior to application to animals, that is, animals without experimental infection, in principle may cause significant increases

330

K. Dąbrowska et al.

in inflammatory markers indicating activation of the immune system, where such activation as mediated directly by medicinals generally is considered to be an undesired effect (Pincus et al. 2015). Phage preparations, including lysates used as antibacterials for the treatment of experimental infections in animals, nevertheless can reduce levels of pro-inflammatory cytokines (Międzybrodzki et al. 2008; Zimecki et al. 2009; Kumari et al. 2010; Górski et al. 2012; Międzybrodzki et al. 2012). Importantly, extensive analysis of immunological parameters in human patients presented by Górski et al. (2012) have shown that no significant overstimulation of the immune system by phage lysates is observed during phage therapy. Some authors suggest that phage therapy success may depend on phages altering bacterial sensitivity to the immune system, for example, as may be seen with phages capable of removing protective surfaces from bacteria (Mushtaq et al. 2004; Schmerer et al. 2014). All these observations illustrate the importance of the dynamic interactions between mammalian organism, infecting bacteria, and phages as three elements that together can contribute to the pharmacodynamic effects displayed by phages during phage therapy.

Conclusion To summarize pharmacology, from the perspectives considered here, one can differentiate the body’s impact on drugs from a drug’s impact on the body, which are described as pharmacokinetics and pharmacodynamics, respectively. The body’s impact on a drug – especially the pharmacokinetic aspects described as absorption, distribution, metabolism, and excretion, all in combination with drug dosing strategies – controls the potential to raise drug concentrations within the immediate vicinity of target tissues or, in the case of antimicrobial agents, within the vicinity of target microorganisms. The higher a drug’s density within a body then generally the greater a drug’s impact on that body, for better or for worse. To the extent that drug densities can be higher especially in association with target tissues, that is, as due to specificity in targeting, then typically the greater the potential for positive efficacy effects (primary pharmacodynamic effects), and the lower the potential for drug side effects (i.e., as a component of secondary pharmacodynamic effects). With bacteriophages, pharmacokinetics often can be considered to involve some form of phage penetration to target bacteria in combination with phage evasion of immune-system-mediated inactivation, though in certain circumstances excretion is relevant as well, for example, as in the treatment of urinary tract infections by systemic phage delivery (Abedon 2014a). Dosing can be topical, parenteral (along with other means of directly supplying phages systemically), or, at least potentially, per os delivery (Ryan et al. 2011), with generally a goal of achieving at least relatively high phage virion densities within the immediate vicinity of target bacteria, that is, within micrometers such that adsorption is highly likely. Over time, phages must adsorb to and then bactericidally infect a large fraction of targeted bacterial cells in order to successfully clear bacterial infections, or other instances of biocontrol of nuisance bacteria. Over macroscopic dimensions, such clearance likely

Bacteriophage Pharmacology and Immunology

331

requires phage densities of at least 107/ml sustained over long periods (many hours, days, or even weeks) or instead higher densities (e.g., 108/ml) sustained if treatments are expected to take place over shorter timescales (Abedon 2014a). With passive treatment, achievement of relatively high virion densities depends entirely on traditional methods of dosing. With active treatment, such dosing is assisted by what can be described instead as auto dosing, where phage replication during infection of target bacteria results especially in highly localized increases in virion densities. This potential can allow application of lower phage numbers topically or, instead, allow for less direct, usually systemic phage application, that is, given circumstances where direct delivery of high phage densities to target bacteria is not easily achieved. With mixed passive-active treatment, a concept otherwise not considered to any great extent in this chapter, an assumption is made that phage in situ replication can to some degree enhance phage therapy efficacy even when high phage densities nevertheless may be directly applied (Abedon 2014b). For instance, auto dosing in this case might achieve not so much relevant enhancements in overall phage densities but instead enhancements in phage penetration to otherwise difficult-to-reach bacteria. By increasing phage virion densities but in an extremely localized manner, especially in the vicinity of immediately adjacent bacteria, such highly localized enhancement of virion densities thus might contribute to phage active penetration into bacterial biofilms as well. There exists a possibility that applying extremely high phage densities may not always be ideal for achieving bacterial eradication, and one scenario in which this could be the case also is when active penetration is required for successful bacterial eradication. Here the idea is that with bacteria existing within clumps, such as microcolonies or biofilms, the outermost bacteria may be able to shield underlying bacteria from phage attack. In this case, should infection with higher multiplicities result in abortive infections that do not result in lysis (Abedon 2011d), then something other than extremely high phage densities (e.g., 109/ml) delivered to the immediate vicinity of target bacteria may be warranted. Such delivery of lower phage doses, for example, 107/ml, may be accomplished via repeated dosing, auto dosing, or some combination of the two. Indeed, repeated dosing can be warranted in circumstances where phage penetration ability either changes over time or is positively impacted by previous phage activity, or in which phage in situ amplification is insufficiently robust to sustain phage densities within the vicinity of target bacteria over relevant time frames. Theory aside, there is no substitute for empiricism, so if phage treatments are less efficacious than desired, use of higher or perhaps even lower phage densities might be explored as possible solutions. See Table 4 for summary of passive treatment, active treatment, and passive-active treatment as well as active penetration. The body’s immune system can negatively impact protein-based drug densities, for example, such as phage virion densities. Immune systems therefore can have limiting effects on phage concentrations in target tissues and organs and thereby, at least in principle, can interfere with efficacy (that is, with primary pharmacodynamic effects). Activation of immune responses nevertheless tends to be more

332

K. Dąbrowska et al.

Table 4 Comparing passive and active variations on phage therapy treatment

Passive treatment

Active penetration

Active treatment

Passiveactive treatment

Size of phage dose Much greater than numbers of to-be adsorbed bacteria Potentially approx. same as numbers of to-be adsorbed bacteria Not substantially greater than or even lower numbers of to-be adsorbed bacteria Much greater than numbers of to-be adsorbed bacteria

Requires direct, e.g., topical application or injection directly into an abscess Yes

Requires phageinduced bacterial lysis No

Requires in situ production of new phage virions No

Ultimately requires phage densities to locally exceed densities of target bacteriaa Yes

Potentially

Yes

Potentially

Yes

No

Yes

Yes

Yes

Yes

Yes

Yes

Yes

a

Phage densities result either from traditional dosing or instead from in situ phage replication and “local” means in the immediate vicinity of target bacteria, e.g., within micrometers

bacterial infection-associated rather than strictly phage-virion related. No direct toxic effect of phage particles applied to mammals, that is, secondary pharmacodynamic effects associated with purified phage particles, have been reported so far. Phages as one element of the phage-bacteria-mammalian immunity balance can give rise to positive immune system modulatory (immunomodulatory) effects, which mainly relates to normalization of pro-inflammatory mediators after boosting by bacterial infection. This means that phages are capable of impacting immune systems, thus affecting the reactivity of immunological factors. On the other hand, intestinal phages have been shown to mediate immunosuppressive effects that can be also relevant at other body sites due to the phenomenon of phage translocation (Górski and Weber-Dąbrowska 2005; Górski et al. 2006b; Barr 2017; Nguyen et al. 2017). These factors shape phage pharmacokinetics and phage fate in vivo. By extension, immunomodulatory capabilities of phage are a weighty determining factor of phage pharmacokinetics.

Bacteriophage Pharmacology and Immunology

333

Cross-References ▶ Bacteriophage as Biocontrol Agents ▶ Bacteriophage Utilization in Animal Hygiene ▶ Clinical Trials of Bacteriophage Therapeutics ▶ Early Therapeutic and Prophylactic Uses of Bacteriophages ▶ The Use of Bacteriophages in Veterinary Therapy Acknowledgments KD’s work was supported by National Science Centre in Poland, grant UMO-2012/05/E/NZ6/03314. AG’s work was supported by National Science Centre in Poland, grant DEC-2013/11/B/NZ1/ 02107.

References Abedon ST (2008) Phage, bacteria, and food. Appendix: rate of adsorption is function of phage density. In: Abedon ST (ed) Bacteriophage ecology. Cambridge University Press, Cambridge, UK, pp 321–324 Abedon ST (2009) Kinetics of phage-mediated biocontrol of bacteria. Foodborne Pathog Dis 6:807–815 Abedon ST (2010) The ‘nuts and bolts’ of phage therapy. Curr Pharm Biotechnol 11:1 Abedon S (2011a) Phage therapy pharmacology: calculating phage dosing. Adv Appl Microbiol 77:1–40 Abedon ST (2011b) Bacteriophages and biofilms: ecology, phage therapy, plaques. Nova Science Publishers, Hauppauge Abedon ST (2011c) Envisaging bacteria as phage targets. Bacteriophage 1:228–230 Abedon ST (2011d) Lysis from without. Bacteriophage 1:46–49 Abedon ST (2012a) Bacterial ‘immunity’ against bacteriophages. Bacteriophage 2:50–54 Abedon ST (2012b) Phage therapy best practices. In: Hyman P, Abedon ST (eds) Bacteriophages in health and disease. CABI Press, Wallingford, pp 256–272 Abedon ST (2012c) Spatial vulnerability: bacterial arrangements, microcolonies, and biofilms as responses to low rather than high phage densities. Viruses 4:663–687 Abedon ST (2014a) Bacteriophages as drugs: the pharmacology of phage therapy. In: Borysowski J, Międzybrodzki R, Górski A (eds) Phage therapy: current research and applications. Caister Academic Press, Norfolk, pp 69–100 Abedon ST (2014b) Phage therapy: eco-physiological pharmacology. Scientifica 2014:581639 Abedon ST (2015a) Bacteriophage secondary infection. Virol Sin 30:3–10 Abedon ST (2015b) Ecology of anti-biofilm agents I. Antibiotics versus bacteriophages. Pharmaceuticals 8:525–558 Abedon ST (2015c) Ecology of anti-biofilm agents II. Bacteriophage exploitation and biocontrol of biofilm bacteria. Pharmaceuticals 8:559–589 Abedon ST (2015d) Phage therapy of pulmonary infections. Bacteriophage 5:e1020260 Abedon ST (2016a) Bacteriophage exploitation of bacterial biofilms: phage preference for less mature targets? FEMS Microbiol Lett 363:fnv246 Abedon ST (2016b) Phage therapy dosing: the problem(s) with multiplicity of infection (MOI). Bacteriophage 6:e1220348 Abedon ST (2017a) Active bacteriophage biocontrol and therapy on sub-millimeter scales towards removal of unwanted bacteria from foods and microbiomes. AIMS Microbiol 3:649–688 Abedon ST (2017b) Bacteriophage clinical use as antibactertial “drugs”: utility precident. Microbiol Spectr 5: BAD-0003-2016

334

K. Dąbrowska et al.

Abedon ST (2017c) Information phage therapy research should report. Pharmaceuticals (Basel) 10:43 Abedon ST (2017d) Phage “delay” towards enhancing bacterial escape from biofilms: a more comprehensive way of viewing resistance to bacteriophages. AIMS Microbiol 3:186–226 Abedon ST (2018a) Bacteriophage-mediated biocontrol of wound infections, and ecological exploitation of biofilms by phages. In: Shiffman M (ed) Recent clinical techniques, results, and research in wounds. Springer Abedon ST (2018b) Phage therapy: various perspectives on how to improve the art. In: Medina C, López-Baena F (eds) Host-pathogen interactions. Humana Press, New York, pp 113–127 Abedon ST, Thomas-Abedon C (2010) Phage therapy pharmacology. Curr Pharm Biotechnol 11:28–47 Abedon ST, Kuhl SJ, Blasdel BG, Kutter EM (2011) Phage treatment of human infections. Bacteriophage 1:66–85 Ackermann H-W (2005) Bacteriophage classification. In: Kutter E, Sulakvelidze A (eds) Bacteriophages: biology and application. CRC Press, Boca Raton, pp 67–90 Aronow R, Danon D, Shahar A, Aronson M (1964) Electron microscopy of in vitro endocytosis of T2 phage by cells from rabbit peritoneal exudate. J Exp Med 120:943–954 Barr JJ (2017) A bacteriophages journey through the human body. Immunol Rev 279:106–122 Boratyński J, Syper D, Weber-Dąbrowska B, Łusiak-Szelachowska M, Poźniak G, Górski A (2004) Preparation of endotoxin-free bacteriophages. Cell Mol Biol Lett 9:253–259 Borysowski J, Międzybrodzki R, Górski A (2014) Phage therapy: current research and applications. Caister Academic Press, Norfolk Brussow H (2013) Bacteriophage-host interaction: from splendid isolation into a messy reality. Curr Opin Microbiol 16:500–506 Bruttin A, Brüssow H (2005) Human volunteers receiving Escherichia coli phage T4 orally: a safety test of phage therapy. Antimicrob Agents Chemother 49:2874–2878 Bryan D, El-Shibiny A, Hobbs Z, Porter J, Kutter EM (2016) Bacteriophage T4 infection of stationary phase E. coli: life after log from a phage perspective. Front Microbiol 7:1391 Bull JJ, Gill JJ (2014) The habits of highly effective phages: population dynamics as a framework for identifying therapeutic phages. Front Microbiol 5:618 Bull JJ, Regoes RR (2006) Pharmacodynamics of non-replicating viruses, bacteriocins and lysins. Proc R Soc Lond B Biol Sci 273:2703–2712 Campbell AM (2006) General aspects of lysogeny. In: Calendar R, Abedon ST (eds) The bacteriophages. Oxford University Press, Oxford, pp 66–73 Carlton RM (1999) Phage therapy: past history and future prospects. Arch Immunol Ther Exp 47:267–274 Casjens SR, Hendrix RW (2015) Bacteriophage lambda: early pioneer and still relevant. Virology 479–480:310–330 Chan BK, Abedon ST (2012a) Bacteriophage adaptation, with particular attention to issues of phage host range. In: Quiberoni A, Reinheimer J (eds) Bacteriophages in dairy processing. Nova Science Publishers, Hauppauge, pp 25–52 Chan BK, Abedon ST (2012b) Phage therapy pharmacology: phage cocktails. Adv Appl Microbiol 78:1–23 Chan BK, Abedon ST (2015) Bacteriophages and their enzymes in biofilm control. Curr Pharm Des 21:85–99 Chan BK, Abedon ST, Loc-Carrillo C (2013) Phage cocktails and the future of phage therapy. Future Microbiol 8:769–783 Chan BK, Turner PE, Kim S, Mojibian HR, Elefteriades JA, Narayan D (2018) Phage treatment of an aortic graft infected with Pseudomonas aeruginosa. Evol Med Public Health 1:60–66 Chanishvili N (2012a) A literature review of the practical application of bacteriophage research. Nova Science Publishers, Hauppauge Chanishvili N (2012b) Phage therapy – history from Twort and d’Herelle through Soviet experience to current approaches. Adv Virus Res 83:3–40

Bacteriophage Pharmacology and Immunology

335

Christie GE, Allison HA, Kuzio J, McShan M, Waldor MK, Kropinski AM (2012) Prophageinduced changes in cellular cytochemistry and virulence. In: Hyman P, Abedon ST (eds) Bacteriophages in health and disease. CABI Press, Wallingford, pp 33–60 Curtright AJ, Abedon ST (2011) Phage therapy: emergent property pharmacology. J Bioanal Biomed S3:010 Dąbrowska K, Miernikiewicz P, Piotrowicz A, Hodyra K, Owczarek B, Lecion D, Kaźmierczak Z, Letarov A, Górski A (2014) Immunogenicity studies of proteins forming the T4 phage head surface. J Virol 88:12551–12557 Dy RL, Richter C, Salmond GP, Fineran PC (2014) Remarkable mechanisms in microbes to resist phage infections. Annu Rev Virol 1:307–331 Fan X, Li W, Zheng F, Xie J (2013) Bacteriophage inspired antibiotics discovery against infection involved biofilm. Crit Rev Eukaryot Gene Expr 23:317–326 Fish R, Kutter E, Wheat G, Blasdel B, Kutateladze M, Kuhl S (2016) Bacteriophage treatment of intransigent diabetic toe ulcers: a case series. J Wound Care 25(Suppl 7):S27–S33 Fogelman I, Davey V, Ochs HD, Elashoff M, Feinberg MB, Mican J, Siegel JP, Sneller M, Lane HC (2000) Evaluation of CD4+ T cell function in vivo in HIV-infected patients as measured by bacteriophage phiX174 immunization. J Infect Dis 182:435–441 Geier MR, Trigg ME, Merril CR (1973) The fate of bacteriophage lambda in non-immune germfree mice. Nature (London) 246:221–223 Gill JJ, Hyman P (2010) Phage choice, isolation and preparation for phage therapy. Curr Pharm Biotechnol 11:2–14 Goodridge LD (2010) Designing phage therapeutics. Curr Pharm Biotechnol 11:15–27 Górski A, Weber-Dąbrowska B (2005) The potential role of endogenous bacteriophages in controlling invading pathogens. Cell Mol Life Sci 62:511–519 Górski A, Kniotek M, Perkowska-Ptasinska A, Mroz A, Przerwa A, Gorczyca W, Dąbrowska K, Weber-Dąbrowska B, Nowaczyk M (2006a) Bacteriophages and transplantation tolerance. Transplant Proc 38:331–333 Górski A, Wazna E, Dąbrowska B-W, Switala-Jelén K, Międzybrodzki R (2006b) Bacteriophage translocation. FEMS Immunol Med Microbiol 46:313–319 Górski A, Międzybrodzki R, Borysowski J, Dąbrowska K, Wierzbicki P, Ohams M, KorczakKowalska G, Olszowska-Zaremba N, Łusiak-Szelachowska M, Kłak M, Jończyk E, Kaniuga E, Gołas A, Purchla S, Weber-Dąbrowska B, Letkiewicz S, Fortuna W, Szufnarowski K, Pawełczyk Z, Rogóz P, Kłosowska D (2012) Phage as a modulator of immune responses: practical implications for phage therapy. Adv Virus Res 83:41–71 Górski A, Dąbrowska K, Hodyra-Stefaniak K, Borysowski J, Międzybrodzki R, Weber-Dąbrowska B (2015) Phages targeting infected tissues: novel approach to phage therapy. Future Microbiol 10:199–204 Gutíerrez D, Rodríguez-Rubio L, Martínez B, Rodríguez A, García P (2016) Bacteriophages as weapons against bacterial biofilms in the food industry. Front Microbiol 7:825 Hagens S, Loessner MJ (2010) Bacteriophage for biocontrol of foodborne pathogens: calculations and considerations. Curr Pharm Biotechnol 11:58–68 Hajek P (1967) Properties of natural 19S antibodies in normal pig serum against the FX174 and T2 phages. Folia Microbiol 12:551–556 Harper DR, Morales S (2012) Bacteriophage therapy: practicability and clinical need meet in the multidrug-resistance era. Future Microbiol 7:797–799 Harper DR, Parracho HMR, Walker J, Sharp R, Hughes G, Werthrén M, Lehman S, Morales S (2014) Bacteriophages and biofilms. Antibiotics 3:270–284 Henry M, Lavigne R, Debarbieux L (2013) Predicting in vivo efficacy of therapeutic bacteriophages used to treat pulmonary infections. Antimicrob Agents Chemother 57:5961–5968 Hobbs Z, Abedon ST (2016) Diversity of phage infection types and associated terminology: the problem with ‘Lytic or lysogenic’. FEMS Microbiol Lett 363:fnw047 Hodyra-Stefaniak K, Miernikiewicz P, Drapala J, Drab M, Jończyk-Matysiak E, Lecion D, Kaźmierczak Z, Beta W, Majewska J, Harhala M, Bubak B, Kłopot A, Górski A, Dąbrowska K

336

K. Dąbrowska et al.

(2015) Mammalian host-versus-phage immune response determines phage fate in vivo. Sci Rep 5:14802 Huff WE, Huff GR, Rath NC, Donoghue AM (2010) Immune interference of bacteriophage efficacy when treating colibacillosis in poultry. Poult Sci 89:895–900 Hyman P, Abedon ST (2010) Bacteriophage host range and bacterial resistance. Adv Appl Microbiol 70:217–248 Hyman P, Abedon ST (2012) Bacteriophages in health and disease. CABI Press, Wallingford Inchley CJ (1969) The activity of mouse Kuppfer cells following intravenous injection of T4 bacteriophage. Clin Exp Immunol 5:173–187 Janeway CA, Travers P, Walport M, Shlomchik MJ (2005) Immunology. Garland Science, New York Jończyk-Matysiak E, Łusiak-Szelachowska M, Kłak M, Bubak B, Międzybrodzki R, WeberDąbrowska B, Żaczek M, Fortuna W, Rogóz P, Letkiewicz S, Szufnarowski K, Gorski A (2015) The effect of bacteriophage preparations on intracellular killing of bacteria by phagocytes. J Immunol Res 2015:482863 Kantoch M (1958) Studies on phagocytosis of bacterial viruses. Arch Immunol Ther Exp 6:63–84 Kantoch M (1961) The role of phagocytes in virus infections. Arch Immunol Ther Exp 9:261–340 Kawai T, Akira S (2006) Innate immune recognition of viral infection. Nat Immunol 7:131–137 Khalifa L, Shlezinger M, Beyth S, Houri-Haddad Y, Coppenhagen-Glazer S, Beyth N, Hazan R (2016) Phage therapy against Enterococcus faecalis in dental root canals. J Oral Microbiol 8:32157 Kucharewicz-Krukowska A, Slopek S (1987) Immunogenic effect of bacteriophage in patients subjected to phage therapy. Arch Immunol Ther Exp 35:553–561 Kuhl S, Hyman P, Abedon ST (2012) Diseases caused by phages. In: Hyman P, Abedon ST (eds) Bacteriophages in health and disease. CABI Press, Wallingford, pp 21–32 Kumari S, Harjai K, Chhibber S (2010) Evidence to support the therapeutic potential of bacteriophage Kpn5 in burn wound infection caused by Klebsiella pneumoniae in BALB/c mice. J Microbiol Biotechnol 20:935–941 Kutateladze M, Adamia R (2008) Phage therapy experience at the Eliava institute. Med Mal Infect 38:426–430 Kutter E, Sulakvelidze A (2005) Bacteriophages: biology and application. CRC Press, Boca Raton Kutter E, De Vos D, Gvasalia G, Alavidze Z, Gogokhia L, Kuhl S, Abedon ST (2010) Phage therapy in clinical practice: treatment of human infections. Curr Pharm Biotechnol 11:69–86 Kutter E, Borysowski J, Międzybrodzki R, Górski A, Weber-Dąbrowska B, Kutateladze M, Alavidze Z, Goderdzishvili M, Adamia R (2014) Clinical phage therapy. In: Borysowski J, Międzybrodzki R, Górski A (eds) Phage therapy: current research and applications. Caister Academic Press, Norfolk, pp 257–288 Kutter EM, Kuhl SJ, Abedon ST (2015) Re-establishing a place for phage therapy in western medicine. Future Microbiol 10:685–688 Labrie SJ, Samson JE, Moineau S (2010) Bacteriophage resistance mechanisms. Nat Rev Microbiol 8:317–327 Langbeheim H, Teitelbaum D, Arnon R (1978) Cellular immune response toward MS-2 phage and a synthetic fragment of its coat protein. Cell Immunol 38:193–197 Letarov AV, Golomidova AK, Tarasyan KK (2010) Ecological basis of rational phage therapy. Acta Nat 2:60–71 Lindberg HM, McKean KA, Wang I-N (2014) Phage fitness may help predict phage therapy efficacy. Bacteriophage 4:e964081 Łobocka M, Hejnowicz MS, Gagala U, Weber-Dąbrowska B, Wegrzyn G, Dadlez M (2014) The first step to bacteriophage therapy: how to choose the correct phage. In: Borysowski J, Międzybrodzki R, Górski A (eds) Phage therapy: current research and applications. Caister Academic Press, Norfolk, pp 23–67 Łusiak-Szelachowska M, Żaczek M, Weber-Dąbrowska B, Międzybrodzki R, Kłak M, Fortuna W, Letkiewicz S, Rogóz P, Szufnarowski K, Jończyk-Matysiak E, Owczarek B,

Bacteriophage Pharmacology and Immunology

337

Górski A (2014) Phage neutralization by sera of patients receiving phage therapy. Viral Immunol 27:295–304 Majewska J, Beta W, Lecion D, Hodyra-Stefaniak K, Klopot A, Kazmierczak Z, Miernikiewicz P, Piotrowicz A, Ciekot J, Owczarek B, Kopciuch A, Wojtyna K, Harhala M, Mąkosa M, Dąbrowska K (2015) Oral application of T4 phage induces weak antibody production in the gut and in the blood. Viruses 7:4783–4799 McCallin S, Alam SS, Barretto C, Sultana S, Berger B, Huq S, Krause L, Bibiloni R, Schmitt B, Reuteler G, Brüssow H (2013) Safety analysis of a Russian phage cocktail: from metaGenomic analysis to oral application in healthy human subjects. Virology 443:187–196 Medzhitov R (2007) Recognition of microorganisms and activation of the immune response. Nature (London) 449:819–826 Merril CR (2008) Interaction of bacteriophages with animals. In: Abedon ST (ed) Bacteriophage ecology. Cambridge University Press, Cambridge, UK, pp 332–352 Merril CR, Biswas B, Carlton R, Jensen NC, Creed GJ, Zullo S, Adhya S (1996) Long-circulating bacteriophage as antibacterial agents. Proc Natl Acad Sci U S A 93:3188–3192 Miedzybrodzki R, Switala-Jelen K, Fortuna W, Weber-Dabrowska B, Przerwa A, LusiakSzelachowska M, Dabrowska K, Kurzepa A, Boratynski J, Syper D, Pozniak G, Lugowski C, Górski A (2008) Bacteriophage preparation inhibition of reactive oxygen species generation by endotoxin-stimulated polymorphonuclear leukocytes. Virus Res 131:233–242 Międzybrodzki R, Fortuna W, Weber-Dąbrowska B, Górski A (2009) A retrospective analysis of changes in inflammatory markers in patients treated with bacterial viruses. Clin Exp Med 9:303–312 Międzybrodzki R, Borysowski J, Weber-Dąbrowska B, Fortuna W, Letkiewicz S, Szufnarowski K, Pawełczyk Z, Rogóz P, Kłak M, Wojtasik E, Górski A (2012) Clinical aspects of phage therapy. Adv Virus Res 83:73–121 Miernikiewicz P, Dąbrowska K, Piotrowicz A, Owczarek B, Wojas-Turek J, Kicielińska J, Rossowska J, Pajtasz-Piasecka E, Hodyra K, Macegoniuk K, Rzewucka K, Kopciuch A, Majka T, Letarov A, Kulikov E, Maciejewski H, Górski A (2013) T4 phage and its head surface proteins do not stimulate inflammatory mediator production. PLoS One 8:e71036 Miernikiewicz P, Klopot A, Soluch R, Szkuta P, Kęska W, Hodyra-Stefaniak K, Konopka A, Nowak M, Lecion D, Kaźmierczak Z, Majewska J, Harhala M, Gorski A, Dąbrowska K (2016) T4 phage tail adhesin gp12 counteracts LPS-induced inflammation in vivo. Front Microbiol 7:1112 Motlagh AM, Bhattacharjee AS, Goel R (2016) Biofilm control with natural and geneticallymodified phages. World J Microbiol Biotechnol 32:67 Muckenfuss RS (1928) Studies on the bacteriophage of d’Hérelle. XI. An inquiry into the mode of action of antibacteriophage serum. J Exp Med 48:709–722 Mushtaq N, Redpath MB, Luzio JP, Taylor PW (2004) Prevention and cure of systemic Escherichia coli K1 infection by modification of the bacterial phenotype. Antimicrob Agents Chemother 48:1503–1508 Nguyen S, Baker K, Padman BS, Patwa R, Dunstan RA, Weston TA, Schlosser K, Bailey B, Lithgow T, Lazarou M, Luque A, Rohwer F, Blumberg RS, Barr JJ (2017) Bacteriophage transcytosis provides a mechanism to cross epithelial cell layers. MBio 8:e01874 Olszowska-Zaremba N, Borysowski J, Dąbrowska K, Górski A (2012) Phage translocation, safety, and immunomodulation. In: Hyman P, Abedon ST (eds) Bacteriophages in health and disease. CABI Press, Wallingford, pp 168–184 Pancer Z, Cooper MD (2006) The evolution of adaptive immunity. Annu Rev Immunol 24:497–518 Parasion S, Kwiatek M, Gryko R, Mizak L, Malm A (2014) Bacteriophages as an alternative strategy for fighting biofilm development. Pol J Microbiol 63:137–145 Park K, Cha KE, Myung H (2014) Observation of inflammatory responses in mice orally fed with bacteriophage T7. J Appl Microbiol 117:627–633 Payne RJH, Jansen VAA (2001) Understanding bacteriophage therapy as a density-dependent kinetic process. J Theor Biol 208:37–48

338

K. Dąbrowska et al.

Payne RJH, Phil D, Jansen VAA (2000) Phage therapy: the peculiar kinetics of self-replicating pharmaceuticals. Clin Pharmacol Ther 68:225–230 Payne RJH, Jansen VAA (2003) Pharmacokinetic principles of bacteriophage therapy. Clin Pharmacokinet 42:315–325 Perreau M, Guerin MC, Drouet C, Kremer EJ (2007) Interactions between human plasma components and a xenogenic adenovirus vector: reduced immunogenicity during gene transfer. Mol Ther 15:1998–2007 Pincus NB, Reckhow JD, Saleem D, Jammeh ML, Datta SK, Myles IA (2015) Strain specific phage treatment for Staphylococcus aureus infection is influenced by host immunity and site of infection. PLoS One 10:e0124280 Pirnay JP, De VD, Verbeken G, Merabishvili M, Chanishvili N, Vaneechoutte M, Zizi M, Laire G, Lavigne R, Huys I, Van den Mooter G, Buckling A, Debarbieux L, Pouillot F, Azeredo J, Kutter E, Dublanchet A, Górski A, Adamia R (2011) The phage therapy paradigm: prêt-à-porter or sur-mesure? Pharm Res 28:934–937 Pirnay JP, Blasdel BG, Bretaudeau L, Buckling A, Chanishvili N, Clark JR, Corte-Real S, Debarbieux L, Dublanchet A, De VD, Gabard J, Garcia M, Goderdzishvili M, Górski A, Hardcastle J, Huys I, Kutter E, Lavigne R, Merabishvili M, Olchawa E, Parikka KJ, Patey O, Pouilot F, Resch G, Rohde C, Scheres J, Skurnik M, Vaneechoutte M, Van PL, Verbeken G, Zizi M, Van den Eede G (2015) Quality and safety requirements for sustainable phage therapy products. Pharm Res 32:2173–2179 Przybylski M, Borysowski J, Jakubowska-Zahorska R, Weber-Dąbrowska B, Górski A (2015) T4 bacteriophage-mediated inhibition of adsorption and replication of human adenovirus in vitro. Future Microbiol 10:453–460 Rus H, Cudrici C, Niculescu F (2005) The role of the complement system in innate immunity. Immunol Res 33:103–112 Ryan EM, Gorman SP, Donnelly RF, Gilmore BF (2011) Recent advances in bacteriophage therapy: how delivery routes, formulation, concentration and timing influence the success of phage therapy. J Pharm Pharmacol 63:1253–1264 Sabour PM, Griffiths MW (2010) Bacteriophages in the control of food and, waterborne pathogens. ASM Press, Washington, DC Schmerer M, Molineux IJ, Bull JJ (2014) Synergy as a rationale for phage therapy using phage cocktails. PeerJ 2:e590 Schooley RT, Biswas B, Gill JJ, Hernandez-Morales A, Lancaster J, Lessor L, Barr JJ, Reed SL, Rohwer F, Benler S, Segall AM, Taplitz R, Smith DM, Kerr K, Kumaraswamy M, Nizet V, Lin L, McCauley MD, Strathdee SA, Benson CA, Pope RK, Leroux BM, Picel AC, Mateczun AJ, Cilwa KE, Regeimbal JM, Estrella LA, Wolfe DM, Henry MS, Quinones J, Salka S, Bishop-Lilly KA, Young R, Hamilton T (2017) Development and use of personalized bacteriophage-based therapeutic cocktails to treat a patient with a disseminated resistant Acinetobacter baumannii infection. Antimicrob Agents Chemother 61:e00954-17 Seed KD (2015) Battling phages: how bacteria defend against viral attack. PLoS Pathog 11: e1004847 Sillankorva S, Azeredo J (2014) The use of bacteriophages and bacteriophage-derived enzymes for clinically relevant biofilm control. In: Borysowski J, Międzybrodzki R, Górski A (eds) Phage therapy: current research and applications. Caister Academic Press, Norfolk Smith HW, Huggins MB, Shaw KM (1987) Factors influencing the survival and multiplication of bacteriophages in calves and in their environment. J Gen Microbiol 133:1127–1135 Sokoloff AV, Bock I, Zhang G, Sebestyen MG, Wolff JA (2000) The interactions of peptides with the innate immune system studied with use of T7 phage peptide display. Mol Ther 2:131–139 Speck P, Smithyman A (2016) Safety and efficacy of phage therapy via the intravenous route. FEMS Microbiol Lett 363:fnv242 Srivastava AS, Kaido T, Carrier E (2004) Immunological factors that affect the in vivo fate of T7 phage in the mouse. J Virol Methods 115:99–104 Stent GS (1963) Molecular Biology of Bacterial Viruses. WH Freeman and Co., San Francisco, CA

Bacteriophage Pharmacology and Immunology

339

Sulakvelidze A, Kutter E (2005) Bacteriophage therapy in humans. In: Kutter E, Sulakvelidze A (eds) Bacteriophages: biology and application. CRC Press, Boca Raton, pp 381–436 Sulakvelidze A, Alavidze Z, Morris JG Jr (2001) Bacteriophage therapy. Antimicrob Agents Chemother 45:649–659 Summers WC (2005) History of phage research and phage therapy. In: Waldor M, Friedman D, Adhya S (eds) Phages: their role in bacterial pathogenesis and biotechnology. ASM Press, Washington, DC Summers WC (2012) The strange history of phage therapy. Bacteriophage 2:130–133 Szermer-Olearnik B, Boratyński J (2015) Removal of endotoxins from bacteriophage preparations by extraction with organic solvents. PLoS One 10:e0122672 Uchiyama J, Maeda Y, Takemura I, Chess-Williams R, Wakiguchi H, Matsuzaki S (2009) Blood kinetics of four intraperitoneally administered therapeutic candidate bacteriophages in healthy and neutropenic mice. Microbiol Immunol 53:301–304 Wei W, Krone SM (2005) Spatial invasion by a mutant pathogen. J Theor Biol 236:335–348 Zimecki M, Artym J, Kocięba M, Weber-Dąbrowska B, Borysowski J, Górski A (2009) Effects of prophylactic administration of bacteriophages to immunosuppressed mice infected with Staphylococcus aureus. BMC Microbiol 9:169

Phage Infection and Lysis John J. Dennehy and Stephen T. Abedon

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Productive, Reductive, and Destructive Phage Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage-Productive Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage-Reductive Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage-Destructive Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Growth Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Strictly Lytic Phages for Phage Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Infection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Eclipse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Gene Expression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Genomes and Replication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Latent Period Continues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .And Continues? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Rise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Host Physiology Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Virion Release . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Holin-Mediated Lysis from Within . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inhibition of Peptidoglycan Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chronic Release . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Determination of Phage Growth Parameter Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Virion Durability Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eclipse, Latent Period, Burst Size, and Rise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Population Growth Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Continuous Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plaques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

342 345 345 357 357 359 360 361 363 364 364 366 367 368 369 369 370 374 374 375 376 376 376 377 377

J. J. Dennehy Biology Department, Queens College and The Graduate Center of the City University of New York, New York, NY, USA e-mail: [email protected] S. T. Abedon (*) Department of Microbiology, The Ohio State University, Mansfield, OH, USA e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_53

341

342

J. J. Dennehy and S. T. Abedon

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378 Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379

Abstract

Viruses are differentiated from other mobile genetic elements by the encapsidation of their genomes during some stage of their life cycles. It is during their infection of bacteria that bacteriophage genomes are both generated and encapsidated. Overall, the process of virus infection involves virion acquisition of host cells, the infection itself, and then subsequent release of progeny virions from these cells: adsorption, infection, and release respectively. Successful phage infections also can be either productive or latent. Latent infections for phages generally are described as lysogenic and entail either insertion of a phage’s genome into its host’s chromosome, as a prophage, or instead prophage existence as a plasmid. Productive infections can be differentiated by their associated mechanisms of virion release. Depending on the phage, this can be lytic, but alternatively can involve chronic virion extrusion or budding. The lytic mechanism, which is an intracellularly effected, phage-induced lysis of phage-infected bacteria, appears to be far more common among bacteriophages than chronic release. Elsewhere in this volume, lysogenic infections and lysis-mediating phage endolysins are considered in depth. Here we focus on phage-productive infections and the various mechanisms of phage-induced bacterial lysis. Our emphasis also is on so-called phage growth parameters, including infection durations (latent period) and the number of virions produced per phage-infected bacterium (burst size). Growth parameters can affect the phage ability to negatively impact bacteria, i.e., as ideally is seen in the course of the phage therapy of bacterial infections.

Introduction At any given moment, all parasites are engaged either in active parasitism of a host or in an extra-organismal transmission, i.e., a “search” that ends either with acquisition of a new host or instead in parasite death. For phages, all of the parasitizing stage is intracellular. In general terms, the process of phage acquisition of a host bacterium is described as adsorption as resulting in virion attachment to the host cell (chapter ▶ “Adsorption: Phage Acquisition of Bacteria”); the parasitizing stage is infection; and the initiation of the extracellular search stage is release. The life cycle of a phage thus consists of release ➔ movement ➔ attachment ➔ infection ➔ release. That is, virion release from a host cell, virion search for a new host cell (as largely though not entirely driven by virion diffusion), virion attachment to a host cell, phage infection of a host cell, and then the next round of release of virion progeny from phage-infected host cells. Phage infections that result in the release of phage progeny are said to be productive.

Phage Infection and Lysis

343

At an organismal level, these various phenomena typically can be assigned values that collectively are useful descriptors of phage life cycle properties, values we refer to here as growth parameters. Growth parameters include such things are phage virion adsorption rates (especially, the adsorption rate constant), phage infection durations (latent period), and the number of virions produced per infection (phage burst size). See Fig. 1 for summary. For the sake of phage use as antibacterial agents, i.e., for phage therapy, at a minimum some degree of phage attachment to and subsequent destruction of bacteria must occur. In addition, it often can be helpful in phage therapy for phage infections to produce new phage progeny, as described quantitatively by the phage burst size. Infections can be successful or unsuccessful for the phage. Here we focus specifically on those infections that are successful, and especially on those that

Fig. 1 Adsorption, infection, and release are phage organismal-level properties that also can be described as life-history characteristics or organismal traits, and which are quantified for phages as growth parameters. These organismal aspects – including virion durability, adsorption rates (chapter ▶ “Adsorption: Phage Acquisition of Bacteria”), infection durations (latent period), and infection fecundities (burst sizes) – are all built upon numerous molecular aspects of phages, as well as molecular aspects of their bacterial hosts (“Biochemistry,” etc., in the figure). Those both molecular and organismal phenotypic aspects in turn are underlain by a phage’s genotype, as well as that of its host (“Genetics” in the figure). Lastly, to the right, phages interact with their environments as organisms, resulting in a phage’s ecology, which as an applied ecology we can also describe as phage therapy pharmacology (chapter ▶ “Bacteriophage Pharmacology and Immunology”). This figure is inspired by that of Abedon (2009)

344

J. J. Dennehy and S. T. Abedon

generate new, that is, progeny phage virions. Lysogenic cycles are covered elsewhere in this volume (chapter ▶ “Temperate Phages, Prophages, and Lysogeny”) as too are unsuccessful infections (chapter ▶ “Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy”). With lytic productive infections, the infection stage ends with phage-induced host-cell lysis, which results in virion release. With chronic productive infections, by contrast, virion release does not coincide with infection termination. The various steps of successful phage infection cycles are indicated in Fig. 2. The mechanisms employed by different phages to effect these life cycle steps tend to vary considerably. Nonetheless, these steps also tend to possess certain commonalities. In this chapter, we consider these various steps of phage infection including

Fig. 2 The phage life cycle, with emphasis on productive lytic cycles. This cycle can be differentiated into (1) periods of extracellular virion (or free phage) movement, encounter, and attachment (collectively, “Search” or extracellular search) and (2) a period of “Infection” that begins after virion attachment and which ends, at least for phage lytic cycles, with virion release from cells. “Infection” has been described also as a “Virocell” stage (chapter ▶ “Adsorption: Phage Acquisition of Bacteria”) and takes place entirely intracellularly, as contrasts especially with “Free virions,” as occur extracellularly. Note that “Post-eclipse” may be described also as period of virion accumulation but otherwise does not appear to have been formally named during the early development of phage biology nomenclature, versus, e.g., the phage latent period or the phage eclipse. Note also that “Free virion” and “Free phages” are used here synonymously. A phage’s generation time spans both the search and infection phase whereas a phage’s fecundity (burst size) is a product solely of the infection phase

Phage Infection and Lysis

345

in terms of mechanisms and variation. Mechanisms of phage-induced bacterial lysis from within are highlighted here (i.e., normal phage-induced bacterial lysis), while phage-induced bacterial lysis from without, particularly via the action of purified phage lysins used as antibacterial medicaments, is considered elsewhere in this volume (chapter ▶ “Enzybiotics: Endolysins and Bacteriocins”). Additionally, more ecologically oriented aspects of the material covered in this chapter as well as more evolutionary perspectives can be found in a subsequent chapter ▶ “Bacteriophage Ecology.” Our aim in this chapter is to provide a broad overview of organismal properties associated with phage infections and this is rather than in-depth molecular descriptions. This emphasis is both for reasons of available space and because typically molecular details are less relevant to phage therapy success than simply appreciating a phage’s ability to display bactericidal, bacteriolytic, or virion production activities (chapter ▶ “Bacteriophage Pharmacology and Immunology”). Thus, numerous molecular mechanisms give rise to, for example, differences in phage adsorption rates, infection durations, and burst sizes, as well as virion durabilities – which also can vary under different conditions or while infecting different host strains – but distinctions between these mechanisms are not often relevant to phage therapy success. Definitions of terms associated with phage infections can be found in Table 1.

Productive, Reductive, and Destructive Phage Infections Phage infections, broadly speaking, can be differentiated into ones that are phageproductive versus phage-destructive versus phage-reductive (Abedon 2008, 2020; Abedon et al. 2009). See Fig. 3 for an overview (Fig. 2, by contrast, concentrates especially on lytic productive infections). In this section, we provide an introduction to these phenomena including how they fit into the concepts of phage growth parameters. In addition, we consider an important subset of productive infections, which include especially successful infections caused by strictly lytic phages.

Phage-Productive Infections Productive phage infections are ones that result in both the production and release of phage progeny virions. Depending on the phages, this release can occur either with or without host-cell lysis. Those phages that release progeny virions without lysing their host cell can be described as displaying chronic infections or chronic release, and such infections generally are neither bacteriolytic nor bactericidal. For example are the filamentous phages of family Inoviridae (Russel and Model 2006) which release their virions chronically via a process of extrusion. Members of the seemingly much rarer phage family, Plasmaviridae (Maniloff et al. 1982; Maniloff and Dybvig 2006), instead release virions chronically via budding, which they are able to accomplish presumably because the bacteria they infect lack cell walls (Maniloff

346

J. J. Dennehy and S. T. Abedon

Table 1 Definitions of key terms Abortive infection

Antiholin Assembly

Attachment Bactericidal

Bacteriolytic

Batch

Burst Burst size Capsid

Chaperonin Chemostat

Phage infection of a bacterium that ends with bacterial death but either with no released virions (narrower definition) or with insufficient numbers of released virions to cumulatively support phage plaque formation (a broader as well as operational definition). Abortive infections are always bactericidal and often bacteriolytic Phage protein that interferes with holin functioning, thereby lengthening a phage’s latent period Process of intracellular formation of virion progeny, involving phage proteins making up capsids and other virion components coming together (self assembly) and also packaging of the phage genome into the maturing capsid Specific virion interactions with the surface of a host cell, thus allowing initiation of the infection process Referring here to phage infections that end with the death of the infected bacterium whether associated with a lytic phage infection or instead with a phage abortive infection. Bactericidal infections are not necessarily also bacteriolytic, though most bactericidal phage infections presumably are also bacteriolytic and all bacteriolytic infections are also bactericidal Referring here to phage infections that end with lysis of the targeted bacterium. All bacteriolytic infections are also bactericidal but not all bacteriolytic phage processes are virion productive. For example, abortive infections can be bacteriolytic but not result in release of phage virions and lysis from without is bacteriolytic well before the start of virion production during a phage infection Growth of a microbial population within a container that contains a fixed volume of media, e.g., such as a test tube or flask. Batch culture growth typically ends either when microorganisms run out of resources or instead if waste products become excessive. For phages, it is generally declines in availability of bacteria that determines the end of batch growth, though alternatively transitions of bacteria from log phase to stationary phase instead can block further phage population growth Informal description of a phage infection’s release as resulting in a burst size of free virions Number of virion progeny produced per phage-infected bacterium Especially that aspect of a virion particle that surrounds (encapsidates) the virus genome, e.g., the head in tailed phages (the tail can be viewed instead as an accessory or appendage to the capsid) Proteins that aid, that is, “chaparone,” the folding of other proteins Device that allows for the continuous culture of microorganisms by feeding growth media to a culture vessel at a flow rate that is matched by the flow of media out of the vessel. More complex continuous cultures for phages (known as two-stage chemostats) (continued)

Phage Infection and Lysis

347

Table 1 (continued)

Chronic release

Circular chromosome

Circular permutation

Cleavage Closed-circular Complementary

Concatemer

Continuous culture

Contractile tail

Daughter strand

can be devised by directing the outflow from a first-stage that contains exponentially growing bacteria to a second-stage containing both bacteria and phage Progeny phage transition from an intracellular to extracellular state that does not involve bacterial lysis. Instead, a process of virion extrusion across the intact bacterial cell envelope occurs, or less commonly, for certain phages chronic release instead can occur via virion budding Nucleic acid, single-stranded or double-stranded, that possesses no ends but instead consists of a continuous structure that therefore can be conceptualized as circular (though its actual shape geometrically is more that of a tangle than that of a circle). With circular chromosomes rolling-circle replication is possible. Note that a circular chromosome is not equivalent to a circularly permuted chromosome A linear phage chromosome that is longer than its number of genes because some genes, found at its ends, are redundantly present. Circular permutation occurs as a consequence of headful packaging Cutting of chromosomes from concatemers toward packaging, e.g., such as at Pac sites Equivalent to simple “circular” for chromosomes As referring to the nucleic acid found in a double helix, each strand is complementary rather than identical to the other, e.g., with the nitrogenous base adenine found in one strand and thymine in the other. Complementary nucleic acid strands also have the property of being antiparallel, meaning that nucleic acid strand backbones have a polarity and, with double helices as consisting of complementary nucleic acid, the polarity of each strand runs in the opposite direction DNA consisting of multiple genomes that are present in tandem, that is, in line. In the course of phage-genome replication, often concatemers are generated which are then cut by various means so that individual genomes may be packaged into individual phage capsids Growth of a microbial population within a container containing replenishable media, such as a chemostat, with that media typically replenished in small amounts over time, such as continuously. Volumes of media do not increase over time, however, as media already present will be lost at the same rate that new medium is gained The tails of phage family Myoviridae are essentially molecular machines that can undergo conformational changes following irreversible attachment. These conformational changes include a contraction, or shortening, of the phage tail, resulting in delivery of the phage’s DNA into the host cytoplasm Newly synthesized nucleic acid strand, i.e., single-stranded polymer, that is complementary as well as antiparallel to the sequence of DNA found in a parental strand template. In semiconservative DNA replication, the resulting new double (continued)

348

J. J. Dennehy and S. T. Abedon

Table 1 (continued)

Destructive infection

Displacement

Doubling time

Early gene

Eclipse

Encapsidation Endolysin Extracellular search

Fitness

helices consist of a base-paired combination of an unbroken parental strand and an unbroken daughter strand Loss of phage viability following phage attachment to a bacterium, e.g., as involving restrictive or abortive infections. “Destructive” here is not a reference to the phage impact on the bacterium but instead is a reference especially to a possible bacterium impact on the phage, particularly given bacterium display of mechanisms of phage resistance that act after virion attachment Replacement of one complementary strand of a double helix with another complementary strand, whether occurring in the course of replication (where a parental strand is displaced by a daughter strand) or instead in the course of homologous recombination (where, in the course of invasion of a double helix, e.g., see intragenomic recombination, one preexisting strand is replaced by another preexisting strand). During rolling-circle replication, displacement by a daughter strand can occur without also replication of the displaced strand, i.e., replication of only one strand, whereas at a replication fork displacement of both parental strands by daughter strands occurs effectively simultaneously The interval over which a population doubles in size during exponential growth, i.e., a measure of population growth rate. See also “Malthusian parameter.” Phage genes that are expressed soon after the start of productive infections and as are often involved in host takeover. Contrast with phage late genes as well as with phage middle genes Period of productive infections during which completion of the maturation of at least one virion progeny has not yet occurred. The eclipse consists in part of a “gearing up” period toward virion progeny assembly See “packaging” Phage enzyme that digests bacterial cell walls, particularly as effected from within Virion movement through the extracellular environment, which ideally for the phage is toward virion attachment to a host bacterium. The search most notably involves virion diffusion, but also can involve virion movement by other means, e.g., such as hitchhiking on animals, and also involves recognition of hostbacteria cell-surface molecules following encounter. The extracellular search thus begins with virion release, leads to encounter, and ends at the point of irreversible virion attachment. Though diffusion is a passive process, other aspects can be more active including reversible phage binding to animal mucus as well as the specific interactions with bacteria that are associated with attachment Measure of a combination of generation time, number of offspring produced, and ability of those offspring to remain viable until they also reproduce. Especially, shorter generation times, larger numbers of offspring produced, and greater (continued)

Phage Infection and Lysis

349

Table 1 (continued)

Free phage

Free virion

Gene expression

Generation time

Growth parameter

Headful packaging

Holin

Host takeover Induction

Infection

Infective center Intragenomic recombination

organism durability will tend to result in a greater fitness, though there also can be trade-offs between improvement in one parameter versus another, such as latent period versus burst size for phages Mature phage virion that is found in the extracellular environment and has not yet irreversibly attached to a host bacterium. The concept of free phage is equivalent to that of free virion As equivalent to free phage for bacterial viruses, this is the phage entity involved in the virion extracellular search and also the entity that contrasts with virocells Templated generation of RNA transcripts as catalyzed by RNA polymerase and involving a process known as transcription. Often the resulting RNA transcripts are then, in a continuation of this process of gene expression, used to template the polymerization of amino acids into polypeptides, the latter a process known as translation Measure of the interval between “birth” and production of the next generation of progeny. In general, the shorter an organism’s generation time, all else held constant (including no trade-offs), then the greater an organism’s population growth rate Quantifiable aspects of virus life cycles, including the adsorption rate constant, latent period length, and burst size, but also virion inactivation rates Means of cutting of concatemers for inclusion in phage capsids that does not rely on Pac sites but instead is dependent on the amount of DNA that fits into the capsid. With headful packaging, circular permutation tends to also be present Phage protein involved in defining the timing of phage-induced bacterial lysis from within as well as allowing phage endolysin to digest the bacterial cell wall Concept of cell metabolic activity being modified especially by productively infecting viruses toward virion production Conversion of a phage lysogenic cycle to a productive cycle, e.g., such as due to a host experiencing DNA damage from UV irradiation or exposure to the chemical agent, mitomycin C Process that a bacteriophage carries out within a bacterium following phage genome uptake and, for lytic cycles, prior to virion release. “Infection” should not be used synonymously with either “Adsorption” or “Attachment” since, for the sake of avoiding ambiguity, direct phage interactions with host bacteria should be separated into the three distinct steps: prior to infection (including encounter and attachment), during infection, and then after infection (as seen upon release). More or less equivalent to plaque-forming unit but explicitly referring to either a free phage or a phage-infected bacterium Invasion of nucleic acid by nucleic acid associated with the same genome (intragenomic), e.g., as associated with circular permutations. (continued)

350

J. J. Dennehy and S. T. Abedon

Table 1 (continued) Late gene

Latent infection

Latent period

Linear chromosome

Lysis

Lysis from within (LI)

Lysis from without (LO)

Lysis inhibition (LIN)

Phage genes that are expressed later during productive infections, e.g., genes involved in virion production that will not be needed until after nucleic acid replication has occurred. Contrast especially with early genes but also with middle genes Virus infection that does not directly result in release nor maturation of virion progeny, but nevertheless involves intracellular virus genome replication. For bacteriophages these are described as lysogenic infections, or lysogenic cycles; contrast especially with productive infection Duration during a productive infection that encompasses the entirety of the infection period. This definition assumes that the transition from reversible attachment to irreversible attachment is sufficiently fast as to be ignored and that infection begins with the process of phage genome uptake into the bacterial cytoplasm Main nucleic acid associated with an organism that possesses ends rather than being continuous. Contrasts with circular chromosomes. Linear chromosomes may or may not be circularly permuted (which is not the same as being circular), where linear chromosomes possessing Pac sites will tend to not be circularly permuted Process of cell-envelope destruction that terminates lytic productive infections, results in virion release, and partially solubilizes the hosting bacterium. Lysis is caused by the actions of specific phage proteins rather than something that just “happens” spontaneously, i.e., not as a consequence of bacteria simply becoming filled up with intracellular virions. Lysis also can be differentiated into lysis from within processes (more common) versus lysis from without processes (less common). Lysis is always associated with successful lytic-productive phage infections and also can be associated with abortive phage infections. Lysis of course underlies all phage bacteriolytic activities Phage-induced bacterial lysis that is initiated from within phageinfected bacteria and which serves to define the end of phage latent periods. Lysis from within is the normal phage lysis observed at the end of phage lytic cycles and usually is what is being referred to when phage-induced bacterial lysis is described without qualification Phage-induced bacterial lysis caused by digestion of bacterial cell walls as caused by irreversibly attaching phage virions and which is not associated with production of new phages. It is possible that most phage types are not able to induce LO, besides large myoviruses such as phage T2 and phage T4. Lysis from without also describes the action of purified endolysins applied to bacterial cultures as antibacterial agents, though this latter is not a process that is otherwise considered in this chapter Mechanism seen in certain phages in which secondary adsorption induces an extension of the period of virion accumulation portion of latent periods. Specifically, it is lysis (continued)

Phage Infection and Lysis

351

Table 1 (continued)

Lysogen Lysogenic cycle

Malthusian parameter

Maturation

Middle gene

MOI

from within that is delayed and this is in response to another phage adsorbing to the already lytically infected bacterium. The process of lysis inhibition requires multiple phage genes to be successfully expressed, as are provided by the initially infecting phage (primary phage). LIN also has been observed primarily with large, coliphage, myoviruses such as phage T4 A bacterial cell harboring a lysogenic cycle Type of phage latent infection during which the host cell continues to grow and divide, and the phage genome is replicated as a prophage along with the host genome, but neither virion production nor virion release occurs. Lysogenic cycles can resolve into productive infections via a process known as induction. Thus, lytic cycles can begin either upon phage genome uptake or following the induction of a phage lysogenic cycle. Strictly speaking, only temperate prophages that are capable of going through lytic cycles following induction can be lysis (“lyso”) generating (“genic”). Temperate, chronically releasing phages also exist, however, e.g., CTXPhi, the vibriophage encoding cholera toxin. Technically, the latent infections of chronically releasing phages cannot be lysis generating, though generally they are described as lysogenic as well Measure of population growth rate, essentially the slope of an exponential increase as graphed as the natural logarithm (ln) of organism numbers versus time. As population growth rates increase, the Malthusian parameter also increases whereas the equivalent measure known as doubling time instead decreases as growth rates increase. The Malthusian parameter is named after Thomas Malthus of population exponential growth fame Assembly of a functional phage virion, i.e., one that is capable in its extracellular state of adsorbing to a bacterium in order to initiate a new infection. Mature phage virions, however, are not also free phages until their release from a cell. In addition, for chronically released phages, the maturation process (and assembly) is completed in the course of virion release Phage genes that are expressed later during productive infections than early genes and sooner than late genes. Not all phages display middle genes Abbreviation of multiplicity of infection, used here synonymously with “multiplicity of adsorption.” The concept of multiplicity of infection was developed prior to the understanding that phage infection does not necessarily always follow phage adsorption, e.g., a MOI of 5 does not always mean that the genomes of 5 phages have all entered the same bacterium, i.e., see “Superinfection exclusion.” See also “Multiplicity.” Note that a less rigorous and less useful though still commonly seen definition of MOI is the ratio of phages added to adsorbable bacteria, and not all authors distinguish between these two meanings of multiplicity, as can be described as MOIinput versus MOIactual (continued)

352

J. J. Dennehy and S. T. Abedon

Table 1 (continued) Morphogenesis Multiplicity Obligately productive

One-step growth Organismal property

Origin of replication

Pac site

Packaging

Parental strand Period of virion accumulation (PVA)

Plaque-forming unit (PFU)

Plasmid

Post-eclipse

Process of generation of a virion particle as involving a combination of assembly and maturation Typically used as short for multiplicity of infection or MOI Description of a phage or a successful infection by a phage that is genetically unable to display lysogenic cycles. See similarly “Strictly lytic” See “Single-step growth” Referring to more holistic properties of organisms versus molecular properties, e.g., phage organismal properties include latent period length, burst size, and virion durability. Organismal properties, that is, often involve the action of multiple interacting gene products versus, e.g., the kinetics of individual enzymes Location on genomes at which DNA replication is initiated (or RNA replication for RNA viruses), e.g., the point of initiation of the replication forks associated with theta replication Phage genome sequences indicating the ends of to-be-packaged linear chromosomes and which are both recognized for cleavage and recognized for the initiation of packaging Process by which an intracellular virion genome becomes enclosed within the capsid of what will be a virion particle. Packaging occurs during infections in the course of virion assembly/maturation That nucleic acid strand from which new nucleic acid undergoes templated polymerization during replication The post-eclipse portion of a phage latent period. The phage latent period thus consists of a period that is prior to the intracellular maturation of the first progeny virion (a period known as the eclipse) and a period that is post the point of maturation of that first virion (post-eclipse or period of virion accumulation). Note that some authors call this period instead the “Rise” but this should be discouraged as the rise traditionally has been used to describe a different aspect of the phage life cycle, i.e., as coinciding instead with virion release Entity capable of forming a plaque. Phage titers often are designated as PFUs rather than in actual phage numbers since experiments operationally will typically directly measure the former (numbers of plaques formed) rather than the latter (number of phage virions actually present). Discrepancies can arise if virions become clumped (thereby lowering numbers of PFUs relative to numbers of viable virions) or if relative efficiencies of plating on a given bacterial indicator host simply are smaller than one (thus also lowering numbers of PFUs relative to numbers of viable virions) Accessory DNA carrying accessory genes, as associated with cells. Plasmids by definition are both not chromosomes and not integrated into chromosomes. Plasmids often are closed-circular. Certain phages during lysogeny can exist as plasmid prophages rather than as bacterial chromosome-integrated prophages See “Period of virion accumulation” (continued)

Phage Infection and Lysis

353

Table 1 (continued) Priming (of replication)

Procapsid

Productive infection Professionally lytic Prophage

Population growth rate

Primary infection

Promotor Pseudolysogeny

Reductive infection

DNA polymerase enzymes, as required for DNA replication, typically are able to extend nucleic acid polymers, but not initiate the production of such polymers. Different enzymes therefore are required to initiate polymer production and the generation of these initial polymers is described as priming. Primers for DNA replication generally consist of short RNA polymers that are generated by a DNA template-dependent RNA polymerase enzyme known as a primase Prior to completion of the virion maturation process, and therefore serving as an intermediate in virion assembly, capsids can exist instead as procapsids. Conversion of procapsids into capsids is typically associated with DNA packaging Virus infection that directly results in release of virion progeny; contrasts with latent infection A strictly lytic phage that also is not descended from a temperate phage The state of phage genomes during lysogenic cycles. Can exist, depending on either the phage or circumstances, either integrated into the host DNA or instead as an unintegrated plasmid Measure of the rate of increase in a population’s size over time, particularly as seen under conditions where resources are not substantially limiting. For phages, population growth rates as a function of productive infections are measured especially under conditions of low phage multiplicity. Note that phage population growth rates are a function not just of extracellular environmental conditions, as well as phage properties, but also densities of target bacteria. More target bacteria being present, all else held constant, generally results in faster phage population growth rates Description of the first phage infecting a bacterium or instead the first phage to initiate a productive infection following adsorption. Used in particular to distinguish from the concept of phage secondary adsorption Region found prior to a gene’s open reading sequence to which RNA polymerase binds to initiate gene transcription A stalled infection as may be observed in response to nutrient limitations. During pseudolysogeny, the phage genome is found within a bacterium’s cytoplasm but does not replicate nor take on a prophage state. Pseudolysogenic states can be resolved, such as upon nutrient addition, into productive infections, reductive infections, or even destructive infections, depending on the phage or circumstances. Note, however, that different authors have defined “pseudolysogeny” to mean a variety of states that are different from the usage employed here An infection where the phage remains viable but nevertheless is not virion productive. Most commonly these would be latent infections, i.e., lysogenic cycles, with the term “reductive” derived from that of a phage being “reduced” to a prophage state. Reductive infection is used here, however, to describe not just (continued)

354

J. J. Dennehy and S. T. Abedon

Table 1 (continued)

Relative fitness Release

Replication Replication fork

Restrictive infection

Rise

RNA polymerase

Rolling-circle replication

Saltatory

lysogenic cycles but also various temporarily stalled productive infections that can be described as pseudolysogeny Measure of organism fitness as compared with that of another organism possessing a different genotype Process of transition of a virion from an intracellular state to an extracellular state. For lytic phages, this by definition involves lysis, whereas for chronically released phages release by definition does not involve lysis Nucleic acid-templated generation of new nucleic acid copies, often resulting for phages in the generation of concatemers Structure existing during replication of a double-stranded nucleic acid, i.e., as the explicit site of that replication. Replication forks have been studied particularly for DNA replication in which one double helix is catalytically converted into two double helixes, each of which possesses one DNA strand from the original double helix and a newly synthesized DNA strand Phage infection in which the phage is inactivated but the phageinfected bacterium remains viable. This is exemplified by the successful impact of restriction endonucleases on infecting phages, though is not necessarily limited to that mechanism (re: CRISPR-Cas). Contrast with abortive infection, i.e., where the host instead dies despite phage inactivation. While restrictive infections typically represent a phage dies-bacterium lives scenario, not all authors describe all such phenomena explicitly as restrictive, especially if not involving the action of restrictionmodification systems Period of extracellular virion accumulation during single-step growth experiments. This term should not also be used to describe the period of virion accumulation, that is, as occurs intracellularly, since “Rise” was described as a phenomenon somewhat prior to the first description of intracellular virion accumulation also as an increase in numbers of assembled virions, but one that does not occur extracellularly In general, this is a DNA-dependent enzyme (but also potentially RNA-dependent for RNA viruses) that polymerizes RNA based upon a template, i.e., as resulting in transcription. Gene expression begins with the “reading” of genes using RNA polymerase and what genes are expressed at a given time typically can be dependent especially on what genes are being read by RNA polymerase Means of replication of closed-circular nucleic acid that involves continuous polymerization, “round and round” the circular genetic material, resulting in multiple, linear, connected copies of genomes known as concatemers (though other mechanisms exist by which concatemers of phage genetic material can be formed). Contrast with theta replication which generates double-stranded rather than the single-stranded DNA (or RNA for RNA phages) Happening in a single “leap” rather than as a more continuous or gradual process, as has been used to describe phage-induced (continued)

Phage Infection and Lysis

355

Table 1 (continued)

Secondary adsorption

Semi-conservative DNA replication

Shine-Dalgarno sequence

Sigma factor

Sigma replication Single-step growth

Spanin Strictly lytic

Structural genes

Superinfection exclusion Superinfection immunity

bacterial that once initiated occurs with great rapidity within individual phage-infected bacteria Refers to the adsorption of a phage to a bacterium that is already phage infected. Secondary adsorptions do not necessarily result in coinfections as not all phages that successfully adsorb to an already phage-infected bacterium also experience successful genome uptake Standard approach to the replication of double-stranded nucleic acid, i.e., as achieved at replication forks, where two strands of “old” nucleic acid each becomes paired with “new” nucleic acid, thus resulting in two double helices each of which consists of one “old” (or “parent”) strand and one “new” (or “daughter”) strand The portion of messenger RNA that is involved in binding that RNA to ribosomes to initiate translation and generally is found upstream of an open reading frame’s start codon Protein that facilitates RNA polymerase binding to specific promoters, i.e., as a way of controlling which genes within a cell are expressed over time A form of rolling-circle replication as seen with phage λ following theta replication Experimental approach where phages are allowed to go through only a single round of adsorption, infection, and release, particularly under conditions where it is predominantly only single rather than multiple phages that are infecting each phageinfected bacterium (low MOI, e.g., 0.1). This experimental approach allows measures of phage latent periods and burst sizes explicitly under circumstances where induction of lysis inhibition is unlikely Phage protein involved in breaching the outer membrane of Gram-negative bacteria during lysis from within Referring to a type of phage that both does not chronically infect and is not temperate, that is, which displays as its productive infections only lytic cycles and is not able to display lysogenic cycles. Also known as obligately lytic or virulent. “Professionally lytic” phages are also strictly lytic, though not only not temperate but not closely related to temperate phages either Especially genes giving rise to nonregulatory proteins but used here to describe those genes that give rise to proteins involved in virion structure and morphogenesis Prevention of secondarily adsorbing phage genome uptake due to genes expressed by a primary phage. Prevention of secondarily adsorbing phages that experience successful genome uptake from otherwise successfully infecting. “Immunity” explicitly is a consequence of primary phages producing prophage repressor proteins that they otherwise use to prevent lysogenic phage infections from becoming productive phage infections, as these proteins also interfere with the ability of phages of the same immunity type (homoimmune) to (continued)

356

J. J. Dennehy and S. T. Abedon

Table 1 (continued)

Temperate phage

Template Terminase

Theta replication

Three transmembranedomain configuration

Trade-off

Transcription

Translation Uptake

Virion

Virocell

Virulent

successfully infect despite gaining access to a bacterium’s cytoplasm Phage that is capable of displaying lysogenic cycles but which also, at different times or under different circumstances, is capable of instead displaying productive infections Equivalent to parental strand, i.e., that nucleic acid sequence that is copied, as a complementary strand, in the course of replication DNA-cutting enzymes used by some phages for converting genome concatemers into individual phage genomes for packaging Means of replication of circular genomes that involves two replication forks, i.e., as equivalent to DNA replication starting at a single point in a linear chromosome, except with theta replication the two replication forks move toward each other because the genome instead is closed-circular. This creates an intermediate that looks like the Greek letter, theta (θ), with the two replication forks metaphorically found at the two points where the horizontal line meets the circle. Contrast with rolling circle replication which generates single-stranded rather than double-stranded DNA Membrane protein structure involving a polypeptide crossing a lipid bilayer a total of three times, i.e., as seen with the phage λ holin protein, but not with its antiholin protein, this despite the two proteins possessing nearly identical primary structures, i.e., amino acid sequences Improvements in one aspect of an organism that are associated with reductions in functionality in another aspect. For example, for phages, an increase in capsid stability could result in a reduction in rates of virion maturation Templated RNA polymerization as catalyzed by the enzyme, RNA polymerase. The regions of genomes that are transcribed generally will correspond to genes Ribosome-catalyzed, RNA-templated polymerization of amino acid subunits into polypeptides and proteins Movement of a phage’s genome, following attachment, from outside of a cell into a cell’s cytoplasm. The start of a phage infection arguably occurs in the course of such genome translocation Not metabolizing encapsidated virus state. A mature bacteriophage virion found unattached and outside of a host cell is called a free phage Description of a cell that is undergoing virus infection, indicating that this state represents a metabolic “partnership” between an infecting virus and the infected cell. Contrast virocell especially with free phage, i.e., with free virions, as virocells can contain intracellular (nonfree) virions. Contrast the virocell stage of virus life cycles also with the extracellular search As used here, a description of a phage that is strictly lytic

Phage Infection and Lysis

357

et al. 1982, Maniloff and Dybvig 2006). As this budding mechanism is associated with nonbactericidal phage infections, it is of little interest to phage therapy, at least thus far in its development, so are not covered in this chapter. The chapter instead emphasizes lytic cycles and especially those associated with the tailed phages, though all tailless phages other than members of phage families Inoviridae and Plasmaviridae also display lytic productive infections. Different phages types as considered from a perspective of virion structure are discussed in chapter ▶ “Structure and Function of Bacteriophages.”

Phage-Reductive Infections Reductive infections are characterized by a combination of phage genome survival and a temporary lack of phage virion production (Abedon 2008, 2020; Abedon et al. 2009). The term “reductive” comes from Lwoff (1953), p. 272 (emphasis his), “When the infection is going to be reductive, when the bacterium is going to be lysogenized, the genetic material of the infecting phage or germ is “reduced” into prophage.” With respect to phage latent states, the term lysogeny refers specifically to latent states where the bacteriophage genome exists as a replicating prophage. Lysogeny is considered in greater depth in a separate chapter ▶ “Temperate Phages, Prophages, and Lysogeny.” Alternatively, phage genomes may persist intracellularly without integration into the host genome. These latent states also include pseudolysogeny (Miller and Day 2008; Abedon 2009c; Los and Wegrzyn 2012). Reductive infections can transition into productive infections, which, for lysogenic cycles, is described as prophage induction. Reductive infections generally are considered to be problematic in effecting phage therapy since, with reductive infections, phagemediated bacterial death (bactericidal activity) typically is avoided, or at least delayed.

Phage-Destructive Infections Infections where neither the phage genome survives nor are phage progeny produced may be described as destructive. That is, during such infections, phages are subjected to processes that result in destruction of the infecting phage. This destruction, though, should not be confused with destruction of the bacterial host as seen upon phage-induced bacterial lysis. Such losses of phage viability in the course of infection can be the result of bacterial restriction-modification systems, abortive infection systems, CRISPR-Cas systems, lysogen-expressed superinfection immunity, superinfection exclusion, or simply can occur as a consequence of biochemical incompatibilities between infecting phages and infected hosts (Hyman and Abedon 2010; Labrie et al. 2010; Stern and Sorek 2011; Samson et al. 2013; Chaturongakul and Ounjai 2014) (chapter ▶ “Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy”).

358

J. J. Dennehy and S. T. Abedon

Fig. 3 Five possible outcomes of phage infection as falling under the categories of destructive infection, reductive infection, and productive infection. These differ in terms of whether the phage or the bacterium lives or dies, and whether new virions are either produced and released (productive infection) or neither produced nor released (reductive infection or destructive infection). Note that for phages, latent infections are caused by temperate phages and that such infections typically are described as lysogenic rather than using the term “latent.” Not listed in the figure under the heading of reductive infection is the concept of pseudolysogeny, which can involve phage-genome survival along with bacterial survival, but without phage replication (Abedon 2009c). Note further that the terms “Abortive infection” and “Restrictive infection” as used here are not employed necessarily as strictly by all authors. With meanings as used here, thus only abortive infections and lytic infections are bactericidal (“Bacterium dies”). For further discussion of phage infection types, see (Hobbs and Abedon 2016; Abedon 2020)

As noted, phage destruction does not necessarily coincide with bacterial destruction, i.e., bactericidal or bacteriolytic phage infection activities do not necessarily coincide with failures of phages to successfully infect. Indeed, of the various phenomena listed in Fig. 3, only those resulting in either lytic infections or abortive infections will, by definition, result in bacterial death. Thus, phage destructive infections generally are not always useful for phage therapy. It is important, however, to distinguish among destructive infection types in terms of impacts on bacteria, particularly abortive infections versus restrictive infections, as restrictive infections by definition (as used here) are not also destructive of bacteria (Table 1; Fig. 3). Indeed, even abortive infections will be more limited in their usefulness in phage therapy than lytic productive infections.

Phage Infection and Lysis

359

Phage Growth Parameters The above descriptors, productive, reductive, and destructive, are qualitative. Alternatively, it is possible to consider phage life cycles from quantitative perspectives, i.e., in terms of growth parameter values. Four such measures are commonly considered, and these are the length of time required for virion adsorption to occur (adsorption rate), the length of the phage infection period (latent period), the number virion progeny produced per phage-infected bacterium (burst size), and the durability of the virion particles (as measured quantitatively in terms of virion inactivation rates). The middle two of these are inextricably linked to the topics covered in this chapter, and thus are of primary interest here. Adsorption instead is covered in chapter ▶ “Adsorption: Phage Acquisition of Bacteria.” Also, we can define productive infections, reductive infections, and destructive infections in terms of associated growth parameter values, which is the emphasis of this subsection.

Latent Periods Phage-productive infections and phage-reductive infections can be distinguished in terms of lengths of latent period. A productive infection has a latent period that often is well-defined in duration and which otherwise is finite in length. A reductive infection, by contrast, can last indefinitely, coming to a close only with loss of the infecting phage (e.g., curing) or transition to productive infection (induction). Thus, productive infections are associated with relatively short latent periods (e.g., much less than many days long) whereas reductive infections are associated with relatively long latent periods (e.g., often much more than many days long). Burst Sizes Productive infections, reductive infections, and destructive infections in turn can be distinguished in terms of burst sizes. A productive infection has a burst size of greater than zero whereas both reductive and destructive infections have burst sizes of zero. Reductive infections, unlike destructiveinfections, however, retain a potential to display infections in which burst sizes are greater than zero, i.e., as following prophage induction. Indeed, with lysogeny, an infection’s lineage has the potential to display multiple bursts, i.e., multiple new productive infections. This potential is a consequence of lysogen binary fission in combination with multiple prophage induction events, while an obligately productive infection typically will display only one burst per initial phage adsorption (at least for lytic, productive infections), and a phage-destructive infection by definition will display no burst at all. Utility for Phage Therapy From a perspective of phage impact on target bacteria, e.g., as during phage therapy, generally higher adsorption rates are preferable to lower adsorption rates, and thus greater adsorption rate constants are preferable, particularly constants associated with that adsorption as achieved in situ (chapter ▶ “Adsorption: Phage Acquisition of Bacteria”). There can be limits to the utility of possessing ever higher adsorption

360

J. J. Dennehy and S. T. Abedon

rates, however, due to diminishing returns. These issues are considered in greater detail in the ecology chapter ▶ “Bacteriophage Ecology.” For lytically infecting phages, a burst size of greater than zero is always associated with bacterial death and always implies that a fully phage-destructive infection has not occurred. Such bactericidal phage infections are crucial for successful phage therapy (chapter ▶ “Bacteriophage Pharmacology and Immunology”). In addition, for phage therapy larger burst sizes often are more desirable than smaller burst sizes since release of new virions has the effect of increasing the in situ dosing of phage particles, as typically is considered to be a good thing (chapter ▶ “Bacteriophage Pharmacology and Immunology”). Lastly, generally reductive infections, or destructive infections in which the phage dies but the bacterium does not (i.e., restrictive infections), are considered to lack utility in phage therapy, and can even be antagonistic to phage therapy success as they have the explicit effect of reducing the bactericidal as well as bacteriolytic impacts of phages.

Strictly Lytic Phages for Phage Therapy For phage therapy, at minimum and as noted, phages must be bactericidal. In addition, it is helpful for phage infections to be productive, since this can lead to increases in phage numbers (and associated titers) within treated individuals, especially in direct, spatial association with phage impact on target bacteria (Fig. 4). Reductive infections, and to a certain degree, destructive infections as well, interfere with phage bactericidal activity, and both reductive infections and destructive infections interfere with the production of new phage virions in situ. Indeed, phage-reductive infections can be worse than phage-destructive infections because not only are bacteria by definition not killed, but the resulting bacterial lysogens are now phage-killing entities, i.e., due to the expression by bacterial lysogens of what is known as superinfection immunity (Blasdel and Abedon 2012), as gives rise to restrictive infections. Phage-destructive infections can be avoided through informed phage choice, as mechanisms of resistance to phages is a determinant of phage host range (Hyman and Abedon 2010; Ross et al. 2016; Hyman 2019). Thus, if a phage is able to kill a bacterium, then that bacterium by definition will be within that phage’s bactericidal host range. Similarly, if a phage is able to produce new virion particles upon infection (virion production), such as leading in the laboratory to plaque formation, then by definition that bacterium is part of a phage’s productive host range. Lysogenic cycles may be fully avoided only by not employing temperate phages for phage therapy. We describe phages that cannot display lysogenic cycles variously as obligately productive or strictly productive, or instead as virulent (the latter, i.e., strictly lytic/obligately lytic). As chronically infecting phages generally are not bactericidal, then they should be avoided for phage therapy as well. Thus, ideally for phage therapy one employs strictly lytic, aka., obligately lytic phages. Furthermore, for reasons such as avoidance of phages that can transduce bacterial DNA (chapter ▶ “Bacteriophage-Mediated Horizontal Gene Transfer: Transduction”), it is best

Phage Infection and Lysis

361

Fig. 4 Depiction of phage infection effectiveness (here, “performance”) as toward phage therapy utility. Passive treatment requires only that phage virions adsorb (attach) and then kill bacteria (bactericidal activity). Active treatment, by contrast, requires that phages adsorb, attach, and kill bacteria as well as produce new phage virions (productive infection). Bacteriolytic activity, as considered distinctly, may have utility especially toward bacterial biofilm eradication. These aspects of phage therapy pharmacology are considered further in the chapter ▶ “Bacteriophage Pharmacology and Immunology.” Particularly, it is strictly lytic phages which display the greater phage therapy infection performance. At least arguably, greater phage performance and therefore antibacterial activity is seen with greater virion durability, faster virion adsorption, shorter phage latent periods, and larger phage burst sizes

also to avoid phages for phage therapy that have recently descended from temperate phages. Such strictly lytic phages that also are not closely related to temperate phages may be described as professionally lytic (Hobbs and Abedon 2016). Here emphasis is primarily on productive infections that end with phage-induced host lysis, though examples especially from phage λ, a temperate phage that has been subject to substantial characterization in this area, nevertheless are provided.

Infection Bacteriophage infections by definition occur within host bacteria, turning bacteria into what can be described instead as virocells (Forterre 2012). The virocell stage of phage life cycles spans the period from virion acquisition of a bacterial host, that is,

362

J. J. Dennehy and S. T. Abedon

given phage genome uptake into the bacterial cytoplasm, at least to the point of release of new virion particles, and indeed beyond this point given chronic rather than lytic release (Fig. 2). As noted above, once infections have been initiated, they can be either phage-reductive or phage-destructive, but also phage-productive (Fig. 3), with our emphasis here especially on the latter. Particularly for lytic productive infections, the duration of infections is traditionally described as a latent period (Fig. 5). In-between steps include an eclipse and then a post-eclipse period of virion accumulation (Fig. 2). A third step, as seen with single-step growth experiments (Hyman and Abedon 2009), is known as the rise. This technically is a period of termination of the latent period that coincides with phage-induced bacterial lysis.

Fig. 5 Depiction of a single-step growth experiment, including with determination of intracellular period of virion accumulation (PVA). Adsorption is synchronized to initiate infections, and this is followed by dilution of resulting infective centers into warm broth to allow for unimpeded phage metabolism throughout the experiment. Traditionally, initial phage multiplicities are somewhat less than 1, e.g., 0.1, and resulting “infective centers” are diluted prior to the end of latent period (and therefore prior to lysis) to limit virion adsorption to new bacteria once those virions have been released from phage-infected bacteria. The end of the “eclipse” is defined as the point at which titers of accumulated intracellular virions, obtained via bacterial artificial lysis (dashed blue curve), equal the number of phage-infected bacteria (the dashed horizontal line started at 100). As noted, PVA stands for period of virion accumulation, i.e., as occurring after the eclipse. The latent period is followed by a rise in extracellular phage titers, as distinct from the rise in extracellular titers instead seen with artificial lysis. The number of phages indicated below the label, “Lysis from Within,” is designated as “mostly” extracellular because some phage-infected bacteria are assumed to persist until the end of the rise. A phage’s burst size is equal to the number of phages present at the end of the rise divided by the number of phage-infected bacteria that are present prior to the start of the rise. This figure was adapted from the experiment presented by Doermann (1952). For more on single- or one-step growth experiments, see Hyman and Abedon (2009) as well as Kropinski (2018)

Phage Infection and Lysis

363

We discuss the rise in this section rather than as a step that occurs after infection because of its association with the mentioned single-step growth experiments (Fig. 5), but do so also to make sure that readers are unambiguously aware that the rise is not equivalent to the intracellular, post-eclipse, period of virion accumulation. Rather, the rise is a period of phage-progeny extracellular virion accumulation.

The Eclipse Productive phage infections generally can be reduced to two consecutive steps, that which occurs prior to the intracellular production of the first progeny virion and that which happens after. This dividing line reflects observations from the 1950s (Doermann 1952) that there is a period following phage attachment when no phage virions can be recovered as plaque-forming units even given artificial lysis of otherwise still-intact phage-infected bacteria. This initial period is then followed by a period during which phage virions instead can be recovered upon this artificial lysis of phage infections (Fig. 5). The first period was termed the “eclipse,” as perhaps can be viewed as an allusion to the occasional disappearance of celestial objects due to intervening bodies, in this case it being the phage virion stage that is “eclipsed” by the early phage-infected bacterium stage (Fig. 2). It is important to realize, however, that nearly all steps associated with phage infections, save for the period of virion accumulation and subsequent virion release, occur during the eclipse. That is, nearly all aspects of gene expression, host takeover, and complete intracellular virion morphogenesis are initiated prior to the later, post-eclipse stage of phage infection. This makes sense as all aspects of phage infection necessary to produce phage virions must occur prior to the end of the eclipse. The length of the eclipse, i.e., the eclipse period, represents another phage growth parameter, or at least a sub-parameter, one that affects the number of virion progeny produced by a phage infection. Though to a degree the eclipse as a concept is physiologically almost arbitrary, its duration nonetheless is both real and has real consequences regarding the kinetics of phage progeny production during infections. We can ask, therefore, why does a given phage, under a given set of circumstances, display the eclipse duration that it does? The answer to this question presumably has to do with a balancing of the length of the eclipse period – where all else held constant shorter presumably is better (thereby giving rise to overall shorter phage latent periods and therefore shorter phage generation times), – with the complexity of the phage infections (where more complex infections presumably would require more time, but also presumably supply some sort of as yet not well understood advantage to phages). Given that all phage infections must display some maturation steps along with gene expression, eclipse periods cannot be of length zero (i.e., with the period of virion accumulation hypothetically beginning immediately after phage genome uptake). So too, however, there presumably are costs, particularly in terms of overall phage generations times, of having excessively long eclipse periods. In addition, a phage may be able to more effectively resist destruction upon infection (destructive infection) at the expense of eclipse period lengthening, such as one sees

364

J. J. Dennehy and S. T. Abedon

with phage T7, which delays the full initiation of infections so as to first block, via the expression of an antirestriction protein, the action of host restriction endonucleases (Molineux 2006).

Phage Gene Expression During the eclipse, phage genes are expressed, the phage genome is replicated, and the processes of virion morphogenesis are initiated. Often phage genes are differentiated into those whose expression begins early in the phage infection and therefore early during the eclipse (early genes) versus those genes whose expression begins late in the phage infection and therefore late in the eclipse (late genes) (Yang et al. 2014). With some phages, e.g., T4, there are also genes that are expressed at intermediate times, or middle genes (Hinton 2010). Early genes are generally involved in host takeover or are simply expressed over long periods. Late genes, by contrast, tend to be relevant to the generation of progeny virions and lysis of the host. Early genes and late genes often occupy separate regions on phage genomes and are associated with characteristic gene expression promoter sequences. In the transition from early to late gene expression, there tend to be modifications in the specificity of RNA polymerases. This often occurs, for example, through the phage expression of RNA polymerase specificity-changing, gene promoter-binding proteins known as sigma factors, e.g., Mosig and Eiserling (2006).

Phage Genomes and Replication Phage genome replication provides not only more templates for transcription but also new phage DNA (or RNA for RNA-genomed phages) for packaging into newly generated phage procapsids (chapter ▶ “Structure and Function of Bacteriophages”). Different phage types display a number of variations on the standard semiconservative DNA replication scheme. In terms of tailed phages, which possess dsDNA genomes, phage λ provides an illustrative example. Early after infection, the circularized λ genome, i.e., now closed-circular rather than a linear chromosome, is replicated bidirectionally from its origin of replication, in a process called θ replication (i.e., theta replication; or circle-to-circle rather than circle-to-concatemer replication). This is equivalent to how bacteria with their circular chromosomes replicate. In the first 15 min following genome uptake into the bacterial cytoplasm, about 5–7 consecutive rounds of θ replication take place, resulting in the production of 50–100 λ circular chromosomes, and all of these copies can serve as templates for RNA transcription (Wegrzyn and Wegrzyn 2005; Narajczyk et al. 2007). At some point, via an unknown mechanism, θ replication gives way to unidirectional σ replication (i.e., sigma replication, which is a form of rolling-circle replication; Fig. 6). The latter produces long, linear DNA molecules containing many copies (concatemers) of the λ genome. These concatemeric molecules are cut by the enzyme, terminase,

Phage Infection and Lysis

365

Fig. 6 Four variations on phage DNA and its replication. These are separate examples rather than a sequence of steps, and most of the DNA replication involved originates at specific DNA sequences known as origins of replication. 1. In rolling circle replication (upper, left), generation of multiple copies of a genome can be initiated from only a single origin of replication, generating a linear concatemer of DNA. This DNA can either remain single-stranded or instead can be converted through further DNA replication to double-stranded DNA, depending on the phage. 2. Alternatively (lower, right), replication can originate by nonstandard means, such as due to strand invasion by other DNAs (recombination initiation), which is particularly possible given phage genome circular permutation. 3. Circular permutation (lower, left) is not equivalent to circular chromosomal DNA but genomes instead are slightly longer than necessary to contain all of a phage’s genes, resulting in each phage genome containing two copies of a small proportion of genes, as found at the ends of these linear chromosomes. How many genes are found in excess varies between individual mature virions. The amount of DNA packaged into virions with circularly permuted genomes is determined by capsid size, hence headful packaging. 4. Alternatively, rather than being circularly permutated (middle, right), DNA packaging into capsids can be facilitated by DNA sequences known as Pac sites, which are locations on DNA where cutting is targeted to result in conversion of DNA concatemers into single rather than single-plus genome lengths

into appropriately sized λ genomes, which are then packaged into newly produced procapsids (Wegrzyn and Wegrzyn 2005, Narajczyk et al. 2007). For phage T4, also a tailed phage, DNA replication too involves origins of replication. In addition, however, it involves multiple rounds of DNA replicationpriming intragenomic recombination steps, producing what are described as large, multibranched concatemers of T4 DNA. These concatemers are then cut into slightly longer than single-genome length units (as resulting in circular permutation) as they

366

J. J. Dennehy and S. T. Abedon

are inserted into phage procapsids in a process known as headful packaging. To a degree, phage T4 DNA replication and subsequent packaging are temporally coordinated since in phage T4 transcription from late-gene promotors, and therefore of virion structural genes, does not begin until DNA replication begins (Mosig and Eiserling 2006). Rolling-circle replication is used by a number of phages as well as other viruses and even plasmids. The process is efficient as it requires only a single priming step, though is limited to replicating closed-circular DNAs. Key is the use of one parental strand as a template and then ongoing displacement of the complementary strand, i.e., as it is replaced by newly synthesized complementary strand. Indeed, the newly synthesized DNA strand itself can be displaced by even more recently synthesized DNA. In this manner, a long, linear concatemer of DNA can be produced off of a single template strand, which will then require further DNA synthesis to generate a complementary strand for dsDNA genomes. Packaging then involves cutting either genome-sized (at Pac sites) or headful-sized DNA lengths (headful packaging). Alternatively, the displaced DNA can be cut and ligated to generate closed-circular DNA. See Fig. 6 for illustration of various DNA-replication-associated concepts.

The Latent Period Continues. . . The stage of phage infection that follows the eclipse does not, to the best of our knowledge, possess a consistently employed formal name. It has been referred to (Hyman and Abedon 2009) as a post-eclipse, intracellular phage growth period, reproductive period, adult period, period of phage-progeny accumulation, or a period of virion maturation. The latter is inaccurate, however, as the process of intracellular virion maturation begins prior to the end of the eclipse. For the sake of convenience, we designate this post-eclipse stage here also as a “period of virion accumulation,” which we will abbreviate as PVA (and which more accurately for lytic phages is a period of virion intracellular accumulation). Note that the phage PVA should not be described in terms of a “rise” in phage numbers within infected bacteria since a different aspect of phage infection is already described as the phage rise (see below). The PVA, by definition, begins following the phage eclipse, and continues until phage-induced bacterial lysis for lytic phages, or the loss of infection viability or otherwise cessation of phage production for chronically infecting phages. Indeed, for phages that chronically release, the PVA is not an intracellular period of virion accumulation, but instead one of ongoing extracellular virion accumulation. The overall PVA duration appears to vary between infections by the same phage type even given otherwise identical infection conditions (Singh and Dennehy 2014), and certainly varies among phage types, as well as with certain types of phage mutations, the latter as considered below in terms of mechanisms of phage-induced bacterial lysis. Over the course of the PVA, phage virions appear to accumulate intracellularly in a linear, that is, constant rate (Wang et al. 1996), or at least linear accumulation is a common assumption. Such kinetics of intracellular virion accumulation makes logical sense, though, assuming that some aspect of virion

Phage Infection and Lysis

367

morphogenesis must serve as a virion-morphogenesis rate-limiting step and so long as the rate of this step does not change over the course of an infection. For example, this might include the number of ribosomes available or the infected bacterium’s rate of ATP production. The overall length of a phage infection, particularly as associated with lytic infections, is called the phage latent period. If one considers only a single phageinfected bacterium, then the end of the latent period coincides with phage-induced bacterial lysis. Alternatively, for an adsorption-synchronized phage population, the latent period can be considered to end with the release of phage virions from the first lysing bacterium. The ensuing period, during which the other infected bacteria in the population are lysed, is termed the phage rise, as considered below.

. . .And Continues? To the extent that new virions are produced at a more or less constant rate by a single lytic-phage infected bacterium, then the length of the PVA multiplied by the rate of intracellular virion accumulation should approximate the resulting infection burst size. Burst size, that is, is the total number of phage progeny virions produced per host–cell infection, and which serves as another key phage growth parameter. Longer latent periods, so long as the increase in length is associated with the PVA, as well as faster virion accumulation rates, thus will give rise to larger phage burst sizes. Postponing host lysis nonetheless directly delays the initiation of subsequent infections by newly produced phages since lytic phages retain their virion progeny intracellularly prior to bacterial lysis. This trade-off between producing more phage progeny intracellularly and allowing those progeny to initiate their own progenyproducing infections has given rise to a series of models devoted to calculating the optimal lysis timing for a lytic phage (Abedon 1989; Wang et al. 1996; Abedon et al. 2001; Bull et al. 2004; Bull 2006; Wang 2006; Bonachela and Levin 2014). The general findings of these models relate latent period length optima – that is, as allowing maximal rates of phage population growth – to numbers of permissive hosts available for infection if the current cell is lysed. Specifically, the more host bacteria that are present, then the faster progeny numbers can be increased by their infecting new hosts, versus increasing numbers solely by continuing intracellular maturation and virion accumulation within a still unlysed host. Thus, more opportunities for creation of new phage infections should provide a benefit to shorter phage latent periods. If the eclipse period is fixed in length, however, then this means that such latent period shortening can only come at the expense of the length of the PVA, and therefore directly at the expense of the phage burst size. The length of a given phage’s PVA thus presumably represents some compromise between a phage displaying shorter generation times (via a shorter PVA) and displaying larger burst sizes (via a longer PVA). It is important to recognize, however, that the PVA can be shortened and burst size increased simultaneously due to improved host physiological conditions (Hadas et al. 1997). Thus, one should not overly generalize the idea that shorter phage latent periods automatically result in smaller phage burst sizes.

368

J. J. Dennehy and S. T. Abedon

Rather, the key perspective is that PVAs that have been shortened due to phage mutation will tend to result in smaller burst sizes. Burst size dependence on the PVA helps to explain advantages associated with the lysis inhibition phenotype that is seen in some phages. Here phage adsorption to already phage-infected bacteria results in an induced PVA extension and associated increased infection burst size (Abedon 1990, 2009a, 2019). The attachment of phages to a phage-infected bacterium, i.e., secondary adsorption, signals to the infecting phage, that is, the virocell, that there may be strong competition for hosts in the surrounding environment. In this situation, the phage is better-off staying put within the bacterium it is already infecting, and this is rather than exit the host but then fail to find an uninfected bacterium to infect. With temperate phages such as phage λ, rather than multiple adsorptions inducing an extension of PVA, instead multiple adsorptions can result in an increased tendency for infections to display lysogenic rather than lytic cycles (Abedon 2017a). That is, an extension of a period preceding the PVA (though for lysogenic cycles that period is not typically described as an eclipse, nor the resulting lysogenic cycle described as a latent period). With chronically released phages there is no equivalent inherent conflict between phage production and infection duration, though faster rates of phage production nonetheless may result in reduced infection durations (Breitbart et al. 2005). For phage therapy, lysis inhibition would be the most important of these phenomena as, unlike the others, it is seen with obligately lytic phages, but it nevertheless is uncertain the extent of prevalence of this phenotype among therapeutic phages.

The Rise The transition from infection to free virions occurs over the course of what is traditionally described as the phage rise. The phage rise is defined as the period during single-step growth experiments over which phage titers increase, that is, rise in number (Fig. 5). Specifically, these are phage populations in which the start of infections has in some manner been synchronized, such as in terms of adsorption. In these experiments, variance in the timing of lysis results in a more gradual rather than instantaneous increase in phage numbers, up to some maximum number of new virions as determined by a phage’s burst size. The rise begins when the first phageinfected cell lyses, which during single-step growth experiments is when the number of infective centers increases above the number of phage-infected bacteria. This also, for many authors, serves to define the end of the phage latent period. The rise in phage numbers then continues until all phage-infected bacteria in the culture have lysed. Note that it is essential to make sure that released phages during the phage rise, during single-step growth experiments (Hyman and Abedon 2009), are unable to adsorb bacteria. This serves three purposes. First, any phages which happen to adsorb bacteria that are already phage infected (secondary adsorptions) will no longer be available as plaque-forming units, i.e., they will instead experience destructive infections (Fig. 3). This point is true even if the phage in question is

Phage Infection and Lysis

369

unable to display superinfection exclusion (Abedon 1994, 2017b). Thus, one will expect an underestimation of burst sizes if such adsorptions to already phageinfected bacteria area allowed to occur. Second, for phages that are able to display lysis inhibition, these secondary adsorptions of already phage-infected will extend the adsorbed phage infection’s latent periods and also substantially multiply the resulting burst size. Lastly, if released phages adsorb to phage-uninfected bacteria, then these new infections can give rise to new phage bursts, also artificially inflating the phage burst size. Nonetheless, it is not entirely uncommon to see in the literature so-called single-step growth experiments which appear, in reality, to be “multiple”step growth experiments, which at best are illustrations of phage exponential growth under a given set of laboratory conditions rather than measures of a phage’s rise or burst size. We return to this concern in a subsequent section, since arguably there is no more valuable single phage-characterization assay concerning phage infection and determination of associated growth parameters than single-step growth experiments, properly performed.

Host Physiology Considerations For most phages, the robustness of productive infections will vary as a function of not only host genetics but also of host physiology including as can vary with environmental conditions (Hadas et al. 1997). At an extreme, phage-productive infections will tend to stall when bacterial hosts enter stationary phase (Miller and Day 2008; Bryan et al. 2016). This outcome can be seen during batch culture phage population growth when bacteria grow to high densities, thereby resulting in failures of phages to lyse these cultures, but also during phage plaque growth (chapter ▶ “Detection of Bacteriophages: Phage Plaques”). Phage plaques for many phages, that is, are limited in size, and these limitations largely are thought to be consequence of bacterial lawn entrance into stationary phase. For certain phages, however, bacterial lawn entrance into stationary phase does not block a continuation of plaque growth, as can result in extremely large plaques such as seen with coliphage T7 (Yin 1991).

Virion Release In this section, we consider phage virion release from infected bacteria. Emphasis is on phage-induced bacterial lysis along with major variations in lysis mechanisms. More formally, this has been described as a lysis from within, or LI (Kao and McClain 1980; Young 1992), in order to distinguish it from a pair of somewhat unrelated phenomena that are described instead as lysis from without, or LO (Abedon 2011). One form of LO is described elsewhere in this volume as induced by purified phage lysins (chapter ▶ “Enzybiotics: Endolysins and Bacteriocins”), but historically LO denotes lysis induced by high multiplicities of phage adsorption to bacteria (Delbrück 1940). Generally, phage-induced bacterial lysis from within

370

J. J. Dennehy and S. T. Abedon

possesses two defining components, which are the timing of initiation of lysis, on one hand, and the mechanism of bacteria cell-wall destruction on the other. Associated with one or both of these phenomena is a combination of the cessation of infection metabolism and release of phage progeny virions that up to this point are trapped intracellularly. The physiological purpose of LI is this phage progeny release. Ecologically, however, phage-induced bacterial lysis is important as well to nutrient cycling within ecosystems, that is, as lysis represents the first step of bacterial decomposition as stems from the action within environments of phages (chapter ▶ “Bacteriophage Ecology”). In simpler systems, i.e., as associated with single-stranded lytic phages, only a single protein is required to effect LI. For double-stranded phages, however, two or more proteins can be involved. Typically these will include both a holin and an endolysin, the latter, a.k.a. a lysin (chapter ▶ “Enzybiotics: Endolysins and Bacteriocins”), though various lysis timing regulating proteins including antiholins may be involved as well. The holins themselves can vary fundamentally in their actions, though in all cases they perform two basic functions: (1) metabolically poisoning and thereby terminating phage infections and (2) allowing endolysins to enzymatically degrade cells walls. More recently described are proteins known as spanins which are involved in breaching the outer membrane of Gram-negative bacteria during the lysis process. Numerous general reviews of phage lysis exist, published particularly by Young and colleagues (Young 1992, 2005, 2013, 2014; Young et al. 2000; Bernhardt et al. 2002; Young and Wang 2006; Chamakura and Young 2018; Cahill and Young 2019).

Holin-Mediated Lysis from Within The majority of phage lysis systems involve, minimally, a combination of two molecules: a holin and an endolysin. The holins are responsible for controlling the timing of LI while the endolysins are responsible for enzymatically degrading the cell wall of the infected bacterium. The enzymatic nature of endolysins, or lysins as they are described for short, is covered elsewhere in this volume (chapter ▶ “Enzybiotics: Endolysins and Bacteriocins”). Here we focus instead on phage holins as well as additional protein factors involved in controlling or effecting the lysis process.

Lysis-Mediating Phage Proteins Holins, as their name partially implies, are responsible for producing holes in the bacterial plasma membrane. So far as is understood, all double-stranded phages employ holins to effect LI. The holes can be large enough to allow endolysins to pass through the plasma membrane to gain access to the bacterial peptidoglycan layer (Wang et al. 2003), or alternatively the holes, known as pinholes because of their roughly nanometer rather than roughly micrometer size, can release endolysins that are tethered to the plasma membrane in the periplasm (Young 2014). In either case, the formation of these holes results in cessation of host metabolism as the proton

Phage Infection and Lysis

371

motive force that fuels ATP synthesis is disrupted. The disruption of the inner membrane by holins and the subsequent enzymatic digestion of cell walls by endolysin are followed, in Gram-negative phages, by the disruption of the outer membrane by spanin proteins (Berry et al. 2012; Kongari et al. 2018). Once the last cellular barrier is subverted, the cell’s turgor pressure ensures the explosive expulsion of the cytoplasmic contents, including progeny phages, into the surrounding medium. Holin-mediated timing of cell lysis is controlled in part by holin-gene transcription and then translation rates in combination with the threshold concentrations of holin protein required to form holes. The holin concentrations required to initiate hole formation depend on the holin protein sequence, modifications of which can produce early or late lysing phage phenotypes (Dennehy and Wang 2011; Kannoly et al. 2020). As general transcription and translation rates are functions in part of environmental conditions, growth in low nutrient media can extend the duration of the latent period (Hadas et al. 1997; Dennehy and Wang 2011). Furthermore, modification of the holin-gene promoter and Shine–Dalgarno sequences, as these also can impact gene transcription and translation rates, each can speed up or delay holin accumulation and eventual lysis. Notwithstanding these various mechanisms that can control the timing of hole formation, once a hole small hole forms in the membrane, the cell rapidly lyses in seconds (Wang et al. 2000; Gründling et al. 2001). The mechanism for the rapid, or saltatory progression of lysis is believed to be the accumulation of previously dispersed membrane-associated holin molecules at the site of the lesion, and also the transformation of antiholin, the holin inhibitor, into functional holin because of the removal of the proton motive force. Thus, after the collapse of the proton motive force, holin, and antiholin jointly will contribute, for many phages, to an approximately micron-sized hole in the membrane sufficient to allow large numbers of endolysin molecules to access their substrate, the peptidoglycan cell wall. See Fig. 7 for a model of hole formation. Antiholins As noted, some phages, such as bacteriophage λ, produce a lysis inhibitor known as an antiholin. The phage λ antiholin is expressed from the same open reading frame as its holin, and only differs from the holin by virtue of an extra two amino acids at the n-terminus (Young 2014). The net positive charge exhibited by the extra amino acids prevents the antiholin from accessing the holin’s three transmembrane-domain configuration (Gründling et al. 2000). Consequently, the antiholin cannot participate in initiation of hole formation, but nonetheless can dimerize with the holin prior to hole formation. This dimerization with antiholin inhibits holin from participating in hole formation and thus represents the molecular basis of the antiholin activity. The inhibition lasts until sufficient sufficient numbers of holin dimers have accumulated, resulting in hole formation despite the antiholin’s presence and thereby abolishing the bacterium’s proton motive force. The resulting membrane depolarization enables antiholin to access the three transmembrane-domain configuration characteristics of the holin, thereby allowing the antiholin to become functionally equivalent to a

372

J. J. Dennehy and S. T. Abedon

Fig. 7 Model of holin action. The holin as encoded by phage λ is a plasma membrane protein with three transmembrane domains. The inactive holin is depicted with two hydrophilic faces on two of the transmembrane domains oriented toward each other (inactive monomer, upper left; note the green faces oriented toward the middle as found at the top of the holin depiction). These monomers can oligomerize while retaining this inactive orientation. Once holin protein has accumulated within the plasma membrane in sufficient quantities, forming into so-called death rafts of holin proteins, the inactive form partially or fully switches to an active form (see model, lower left), which then spontaneously forms holes with the hydrophilic faces now oriented toward the lumen of the holes. This hole formation poisons the plasma membrane, which prompts further holin switching from inactive to active form. Endolysin (not shown) is now able to digest the cell wall of the phageinfected bacterium and lysis ensues. Also not shown here is the role of antiholin proteins. This figure is derived from that presented by Cahill and Young (2019)

holin. See Box 1 for further discussion of the possible utility of antiholin expression by phage λ. Box 1 Why Have an Antiholin?

In wild type phage λ, holin and antiholin are expressed in a 2:1 ratio (Chang et al. 1995). Since each antiholin inhibits one holin (antiholin’s dimerization with itself is negligible), two-thirds of the output of the phage λ S gene, which encodes both proteins, is initially functionally inactive. Given that proteins are energetically expensive to produce, the existence of this system thus (continued)

Phage Infection and Lysis

373

Box 1 Why Have an Antiholin? (continued)

constitutes an evolutionary puzzle. We must assume that nature is not profligate; nonetheless, no fitness-enhancing function for antiholin has been experimentally demonstrated. It is conceivable, however, that antiholin serves as many as three roles. First, antiholin may prevent premature lysis by dampening stochastic gene expression from the λ late gene promoter. Our reasoning for this hypothesis stems from the fact that gene expression systems are well-known to generate unpredictable protein numbers in identical cells in the same environment (Raj and van Oudenaarden 2008). Consider that phages will be encountering cells with variable numbers of RNA polymerases and ribosomes. Moreover, due to the inherent probabilistic nature of biochemical reactions, these complexes are engaged in a stochastic manner. Given the holin’s relatively low threshold for nucleation, ~1000 molecules (Chang et al. 1995), it is plausible that this threshold could be reached during the eclipse period. This outcome would be tantamount to phage suicide – or more specifically, to an abortive infection – because progeny virions would not have yet been produced. Second, if the cell should die before holin-induced lysis occurs, then any replicated progeny would be trapped within the dead cell. Antiholin may prevent this circumstance because it is converted to holin following the termination of the proton motive force. While insufficient holin may be present to lyse the cell, the combined amount of holin and antiholin may be enough to form holes and lyse the cell even though the cell is otherwise physiologically dead. In other words, the phage infection in effect produces emergency but normally nonfunctional holin in reserve, thereby making it easier to balance the lysis-timing and lysis-effecting roles of holins since no longer must sufficient quantities of holin necessary to start lysis also be sufficient quantities of holin to finish lysis. Third, antiholin may reduce the impact of noise in gene expression on the timing of lysis via an incoherent feed-forward regulatory system (Ghusinga et al. 2017). Incoherent feed-forward regulatory systems are control system variants (others include, for example, positive and negative feedback) where an effector and its inhibitor are expressed from the same promoter (Singh and Dennehy 2014, Ghusinga et al. 2017). For stable proteins, such as holin, it has been shown that incoherent feed-forward control reduces gene expression noise more effectively than feedback control (Chang et al. 1995). Other features of the λ holin expression system, including the high transcription rate, the low translational “burst size” (i.e., the average number of proteins produced per mRNA molecule), and the optimal holin lysis-triggering threshold concentration, also ensure that lysis occurs at precisely scheduled times (Kannoly et al. 2020).

374

J. J. Dennehy and S. T. Abedon

Less well studied but no less interesting are the antiholins of phage T4. T4 antiholins are the products of the phage rI and rIII genes. The resulting RIII antiholin is unusual in that it acts on the cytoplasmic rather than periplasmic-side the phage T4 holin, which is encoded by the phage T4 gene t (Chen and Young 2016). The RI antiholin is even more unusual than RIII in that it serves as an extrinsically activatable antiholin, that is, activated by something that comes from outside of the phage-infected bacterium rather than inside, i.e., rather than by cell metabolic poisoning. Specifically, the secondary adsorption signal that stimulates the expression of T4 lysis inhibition is in some manner received and acted upon by the RI antiholin, which then serves to interfere with T-holin lysis-timing activity (Tran et al. 2005).

Inhibition of Peptidoglycan Production Toward phage-induced lysis from within, there exists an alternative strategy to that of destroying intact cell-wall peptidoglycan via phage endolysin-mediated digestion following holin-based hole formation. That other approach is instead to prevent cell wall formation and remodeling during the process of bacterial growth. The primary advantage to this latter approach toward lysis from within is that interfering with the completion of cell wall formation is simpler in terms of the gene products involved, and it may also more directly link lysis timing with resulting burst size. For example with phage MS2, an RNA phage, translation of its lysis protein and capsid protein are linked, suggesting that sooner lysis would be associated with larger phage burst sizes and vice versa (Bernhardt et al. 2002). Notwithstanding the exact utility for their doing so, it appears to be particularly the simpler of phages which employ such single-gene, nonholin-endolysin-based lysis systems. These are the single-stranded, nontailed, lytic phages, members of phage families Leviviridae (ssRNA phages) and Microviridae (ssDNA phages, both with icosahedral capsids (chapter ▶ “Structure and Function of Bacteriophages”)). These nonholin and nonendolysin phage lytic proteins have been described as “protein antibiotics” (Bernhardt et al. 2002), that is, phage-encoded proteins that interfere with peptidoglycan production (Chamakura and Young 2018) as analogous to the interference with peptidoglycan production seen with the nonprotein antibiotic, penicillin. More prominent as potential cell-wall inhibiting protein antibiotics, however, are the endolysins of tailed phages, as have been described instead as “Enzybiotics” (chapter ▶ “Enzybiotics: Endolysins and Bacteriocins”).

Chronic Release The primary, though certainly not only challenge for phages, in terms of their release to the extracellular environment, is breaching the bacterial cell wall. Most phages accomplish this by degrading the cell wall to a point that it fails, physiologically terminating both the phage infection and the phage-infected bacterium. The resulting

Phage Infection and Lysis

375

bacterial lysis has the effect of creating relatively large holes in the cell envelope of host bacteria through which phage virions can pass, as well as to a degree solubilizing the bacterium. Alternatively, certain bacteria exist which lack cell walls and at least one phage is known that passes through a cell-wall-less cell envelope via budding, i.e., as is commonly seen also with numerous enveloped eukaryotic viruses. This phages is Acholeplasma phage L2 and it is the sole isolated member of the phage family, Plasmaviridae (Maniloff and Dybvig 2006). A third approach is seen with members of phage family Inoviridae, the filamentous phages. These phages extrude their virions across intact cell-wall-containing cell envelopes and do so in a likely much more common form of phage virion chronic release than virion budding. From a phage therapy perspective, extrusion as a means of virion release is not highly relevant since while such phages may be engineered in various ways to kill bacteria (Hagens et al. 2004; Yacoby and Benhar 2008; Moradpour et al. 2009; Ngo-Duc et al. 2020; Peng et al. 2020), actual phage release, should it even occur, is not involved in the antibacterial process. Thus, unlike lysis, chronic release is not mechanistically considered in much detail here. Nevertheless, as follows is a short overview of the process. The DNA of the filamentous inoviruses is single-stranded and does not accumulate in the infected bacterium cytoplasm. Instead that DNA complexes with phage protein (pV) and this complex is then recognized by a second phage protein (pI) that is host plasma-membrane associated. Protein pI is found at regions in the cell envelope where the inner and outer membranes interact, due to associations of pI and pXI proteins in the inner membrane and pIV protein in the outer membrane (all of which are phage proteins). In the process of extrusion, the pV protein is stripped from the phage DNA while various other phage proteins are added (pIII, pVI, pVII, pVIII, pIX). The now mature phage is released through a channel formed by the pIV protein and thereby from the bacterium. Unlike lytic phages which exit phageinfected bacteria through large holes, filamentous phages, presumably due in large part to the narrowness of their virions (about 6.5 nm), are able to pass through small holes in the bacterial cell envelope, holes which are not highly disruptive to either ongoing cell or infection metabolism (Russel and Model 2006).

Determination of Phage Growth Parameter Values Phage organismal characteristics have long been subject to experimental determination. Experimental approaches include though are not limited to virion survival experiments under various conditions, virion adsorption rate determinations, variations on single-step growth curves, and also determinations of phage population growth rates, chemostat dynamics, and the study of plaque formation. In this section, we provide brief overviews of these approaches. Many of these methods involve determinations of phage titers, or numbers of phage-infected bacteria, which usually involves plaque-based enumeration (chapter ▶ “Detection of Bacteriophages: Statistical Aspects of Plaque Assay”).

376

J. J. Dennehy and S. T. Abedon

Virion Durability Determination Virion survival curves are commonly generated to observe the impact of environmental extremes, such as in pH or temperature. Survival curves also are used to determine the effects of virion processing, or instead the impact of virion storage or exposure to disinfecting agents. Other aspects involve issues of virion interaction with animal immune systems (chapter ▶ “Bacteriophage Pharmacology and Immunology”) as well as phage survival following spraying on crops (chapter ▶ “Crop Use of Bacteriophages”). The standard protocol for determination of phage survival characteristics is straightforward, involving repeated enumeration of a virion population that has been subjected to some constant, often challenging condition over some span of time. Inactivation is typically exponential (exponential decline). In some cases, such as following exposure to ultraviolet light, the inactivation also can be conditional depending on the circumstances associated with or taking place prior to enumeration (Weinbauer et al. 1997).

Eclipse, Latent Period, Burst Size, and Rise Single- or one-step growth curves involve, ideally, a synchronized start to phage infections. This can be accomplished via phage adsorption to a population of metabolically inhibited target bacteria, usually at relatively low phage multiplicities, or instead in the course of lysogen induction. These now productively infected bacteria are then enumerated at regular intervals until phage numbers increase, corresponding to phage release, and then stabilize in number at a level determined by a phage’s burst size. When it is that virion release first occurs can serve to define the end of the phage latent period as well as the start of the phage rise. It is possible, however, to skip many of the middle time points involved in latent period and rise determination so as to determine burst size only. In retaining these middle time points, enumeration following artificial lysis of cells allows characterization of the phage eclipse, which ends at the point of the intracellular maturation of the first virion progeny (Fig. 5). Further discussion can be found in Delbrück (1946), Benzer et al. (1950), Hyman and Abedon (2009), and Kropinski (2018). Doermann provides explorations of means of artificially lysing phage-infected bacteria to determine eclipse periods in association with single-step growth determinations (Doermann 1951, 1952; Anderson and Doermann 1952).

Phage Population Growth Rates Phage population growth rates can be determined either for a single phage type or instead for mixtures of different phage types, the latter, e.g., such as a phage mutant versus its wild-type parent. With single phage types, for a given set of conditions, host type, and host densities, a phage exponential growth rate, called a Malthusian parameter, can be determined or instead, equivalently, a phage population doubling

Phage Infection and Lysis

377

time may be assessed. The rate of this growth will be a function of phage generation times and burst sizes. Phage durability need not play a substantial role, though antiphage antagonists, such as relatively low titers of antiphage serum, can be incorporated into growth-rate determinations if that is desired. Population growth rates with mixtures of phages typically will be determined for two competing phages with the intention of determining what can be described as a relative fitness. That is, the rate at which one phage population increases in number relative to the other. This approach is employed to determine the degree to which distinct populations of the same phage isolate differ in terms of their evolutionary fitness, such as in competing a viable phage mutant with a wild type phage. Both separate and competing phage population growth curves are presented in Abedon et al. (2001, 2003), and see Dennehy et al. (2007) and Shao and Wang (2008) for relative fitness calculations.

Continuous Culture Continuous culture growth, as occurs in devices known as chemostats, can be somewhat more complex than simple batch phage population growth. This occurs because at least four distinct phage population states can exist with continuous culture: (1) phage population growth, (2) phage population decline, (3) phage population extinction, and (4) phage- as well as bacterium-population-size steady states. Extinction, furthermore, can occur either because phages cannot replicate faster than they are lost from growth vessels, such as due to too-low densities of susceptible host bacteria, or instead, at an extreme, because phages have driven the population of host bacteria to extinction. By including the possibility of phageresistant hosts as well as phage host-range mutations, yet additional scenarios are possible (chapter ▶ “Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy”). Nevertheless, basic phage organismal characteristics still determine which of these scenarios will be present in a continuous culture, particularly as determined by phage burst size, but also virion adsorption rates and latent periods. An important article documenting phage chemostat growth is that of Bohannan and Lenski (1997), as has been deconstructed numerically in terms of growth parameters (Abedon 2009b).

Plaques Plaque formation involves not just phage population growth but also variations in both phage location and the physiological states of host bacteria, all at the same time (chapter ▶ “Detection of Bacteriophages: Phage Plaques”). As with phage population growth within more fluid environments, it is those phages with optimal combinations of shorter latent periods and larger bursts sizes that are expected to give rise to faster increases in plaque size. Perhaps particularly latent period length impacts these rates (Abedon and Culler 2007a), though burst size of course should be highly

378

J. J. Dennehy and S. T. Abedon

relevant as well in terms of increases in numbers of phages present per plaque (Abedon and Culler 2007b). The impact of virion adsorption rates (chapter ▶ “Adsorption: Phage Acquisition of Bacteria”), by contrast, is less straightforward, with particularly high affinity of virions for target bacteria potentially slowing down phage movement into the periphery of plaques and therefore slowing down rates of plaque-size increase. In any case, for most phages the entrance of their bacterial hosts into stationary phase blocks further phage population growth and limits phageinduced bacterial lysis. Stationary phase, therefore, will impose limits on the size of phage plaques. For study of the theoretical impact of different phage growth parameters of rates of plaque-size increase, see Abedon and Culler (2007a) and see Gallet et al. (2011) for experimental exploration.

Conclusion Since especially the 1960s, the study of phage biology has been substantially a molecular endeavor. Phages, however, exist as collections of different molecules rather than as the more isolated molecules that are required for precise molecular study. These phage-molecule collections we can describe as either free virions or virocells, both of which are representative of individual phages as organisms.

Fig. 8 Summary of phage organismal properties and underlying factors. For discussion of especially adsorption rates (see chapter ▶ “Adsorption: Phage Acquisition of Bacteria”)

Phage Infection and Lysis

379

Indeed, predating the intensive study of phage molecules was the study instead of phages primarily as whole organisms and in terms of whole-organism traits (Fig. 8), e.g., such as in terms of single-step growth experiments or abilities to form plaques (chapter ▶ “Detection of Bacteriophages: Phage Plaques”). Many of the phage organismal-level phenotypes can be described quantitatively as phage-growth parameters, e.g., latent period, burst size, or adsorption rate constant, and in this chapter we have concentrated on those phenotypes associated especially with virocells. Together with the organismal phenotypes associated with phage virions (chapter ▶ “Adsorption: Phage Acquisition of Bacteria”), these phage organismal phenotypes largely define a phage’s interaction with its environments, whether these are the natural environments outside of the body (chapter ▶ “Bacteriophage Ecology”) or instead as seen during phage treatment of bacterial infections within bodies (chapter ▶ “Bacteriophage Pharmacology and Immunology”).

Cross-References ▶ Adsorption: Phage Acquisition of Bacteria ▶ Bacteria-Phage Antagonistic Coevolution and the Implications for Phage Therapy ▶ Bacteriophage Ecology ▶ Bacteriophage Pharmacology and Immunology ▶ Bacteriophage Use in Molecular Biology and Biotechnology ▶ Bacteriophage-Mediated Horizontal Gene Transfer: Transduction ▶ Crop Use of Bacteriophages ▶ Detection of Bacteriophages: Phage Plaques ▶ Detection of Bacteriophages: Statistical Aspects of Plaque Assay ▶ Enzybiotics: Endolysins and Bacteriocins ▶ Food Safety ▶ Structure and Function of Bacteriophages ▶ Temperate Phages, Prophages, and Lysogeny Acknowledgments JJD acknowledges financial support from the National Institutes of Health NIGMS through grant number 1R01GM124446-01.

References Abedon ST (1989) Selection for bacteriophage latent period length by bacterial density: a theoretical examination. Microb Ecol 18:79–88 Abedon ST (1990) Selection for lysis inhibition in bacteriophage. J Theor Biol 146:501–511 Abedon ST (1994) Lysis and the interaction between free phages and infected cells. In: Karam JD, Kutter E, Carlson K, Guttman B (eds) The molecular biology of bacteriophage T4. ASM Press, Washington, DC, pp 397–405 Abedon ST (2008) Ecology of viruses infecting bacteria. In: Mahy BWJ, Van Regenmortel MHV (eds) Encyclopedia of virology, 3rd edn. Elsevier, Oxford, pp 71–77

380

J. J. Dennehy and S. T. Abedon

Abedon ST (2009a) Bacteriophage intraspecific cooperation and defection. In: Adams HT (ed) Contemporary trends in bacteriophage research. Nova Science Publishers, Hauppauge, pp 191–215 Abedon ST (2009b) Deconstructing chemostats towards greater phage-modeling precision. In: Adams HT (ed) Contemporary trends in bacteriophage research. Nova Science Publishers, Hauppauge, pp 249–283 Abedon ST (2009c) Disambiguating bacteriophage pseudolysogeny: an historical analysis of lysogeny, pseudolysogeny, and the phage carrier state. In: Adams HT (ed) Contemporary trends in bacteriophage research. Nova Science Publishers, Hauppauge, pp 285–307 Abedon ST (2011) Lysis from without. Bacteriophage 1:46–49 Abedon ST (2017a) Commentary: communication between viruses guides lysis-Lysogeny decisions. Front Microbiol 8:983 Abedon ST (2017b) Phage “delay” towards enhancing bacterial escape from biofilms: a more comprehensive way of viewing resistance to bacteriophages. AIMS Microbiol 3:186–226 Abedon ST (2019) Look who’s talking: T-even phage lysis inhibition, the granddaddy of virus-virus intercellular communication research. Viruses 11:951. https://pubmed.ncbi.nlm.nih.gov/ 31623057/ Abedon ST (2020) Phage-phage, phage-bacteria, and phage-environment communication. In: Witzany G (ed) Biocommunication of phages. Springer, Cham Abedon ST, Culler RR (2007a) Bacteriophage evolution given spatial constraint. J Theor Biol 248:111–119 Abedon ST, Culler RR (2007b) Optimizing bacteriophage plaque fecundity. J Theor Biol 249:582–592 Abedon ST, Duffy S, Turner PE (2009) Bacteriophage ecology. In: Schaecter M (ed) Encyclopedia of microbiology. Elsevier, Oxford, pp 42–57 Abedon ST, Herschler TD, Stopar D (2001) Bacteriophage latent-period evolution as a response to resource availability. Appl Environ Microbiol 67:4233–4241 Abedon ST, Hyman P, Thomas C (2003) Experimental examination of bacteriophage latent-period evolution as a response to bacterial availability. Appl Environ Microbiol 69:7499–7506 Anderson TF, Doermann AH (1952) The intracellular growth of bacteriophages. II. The growth of T3 studied by sonic disintegration and by T6-cyanide lysis of infected cell. J Gen Physiol 35:657–667 Benzer S, Hudson W, Weidel W, Delbrück M, Stent GS, Weigle JJ, Dulbecco R, Watson JD, Wollman EL (1950) A syllabus on procedures, facts, and interpretations in phage. In: Delbrück M (ed) Viruses 1950. California Institute of Technology, Pasadena, pp 100–147 Bernhardt TG, Wang I-N, Struck DK, Young R (2002) Breaking free: “protein antibiotics” and phage lysis. Res Microbiol 153:493–501 Berry J, Rajaure M, Pang T, Young R (2012) The spanin complex is essential for lambda lysis. J Bacteriol 194:5667–5674 Blasdel BG, Abedon ST (2012) Superinfection immunity. In: Mayloy S, Hughes K (eds) Brenner’s encyclopedia of genetics. Elsevier/Academic, Amsterdam Bohannan BJM, Lenski RE (1997) Effect of resource enrichment on a chemostat community of bacteria and bacteriophage. Ecology 78:2303–2315 Bonachela JA, Levin SA (2014) Evolutionary comparison between viral lysis rate and latent period. J Theor Biol 345:32–42 Breitbart M, Rohwer F, Abedon ST (2005) Phage ecology and bacterial pathogenesis. In: Waldor MK, Friedman DI, Adhya SL (eds) Phages: their role in bacterial pathogenesis and biotechnology. ASM Press, Washington, DC, pp 66–91 Bryan D, El-Shibiny A, Hobbs Z, Porter J, Kutter EM (2016) Bacteriophage T4 infection of stationary phase E. coli: life after log from a phage perspective. Front Microbiol 7:1391 Bull JJ (2006) Optimality models of phage life history and parallels in disease evolution. J Theor Biol 241:928–938

Phage Infection and Lysis

381

Bull JJ, Pfennig DW, Wang I-W (2004) Genetic details, optimization, and phage life histories. Trends Ecol Evol 19:76–82 Cahill J, Young R (2019) Phage lysis: multiple genes for multiple barriers. Adv Virus Res 103:33–70 Chamakura K, Young R (2018) Phage single-gene lysis: finding the weak spot in the bacterial cell wall. J Biol Chem 294:3350–3358 Chang C-Y, Nam K, Young R (1995) S gene expression and the timing of lysis by bacteriophage lambda. J Bacteriol 177:3283–3294 Chaturongakul S, Ounjai P (2014) Phage-host interplay: examples from tailed phages and gramnegative bacterial pathogens. Front Microbiol 5:442 Chen Y, Young R (2016) The last r locus unveiled: T4 RIII is a cytoplasmic antiholin. J Bacteriol 198:2448–2457 Delbrück M (1940) The growth of bacteriophage and lysis of the host. J Gen Physiol 23:643–660 Delbrück M (1946) Bacterial viruses or bacteriophages. Biol Rev 21:30–40 Dennehy JJ, Abedon ST, Turner PE (2007) Host density impacts relative fitness of bacteriophage Φ6 genotypes in structured habitats. Evolution 61:2516–2527 Dennehy JJ, Wang IN (2011) Factors influencing lysis time stochasticity in bacteriophage lambda. BMC Microbiol 11:174 Doermann AH (1951) Intracellular phage growth as studied by premature lysis. Fed Proc 10:591–594 Doermann AH (1952) The intracellular growth of bacteriophages I. liberation of intracellular bacteriophage T4 by premature lysis with another phage or with cyanide. J Gen Physiol 35:645–656 Forterre P (2012) The virocell concept and environmental microbiology. ISME J 7:233–236 Gallet R, Kannoly S, Wang IN (2011) Effects of bacteriophage traits on plaque formation. BMC Microbiol 11:181 Ghusinga KR, Dennehy JJ, Singh A (2017) First-passage time approach to controlling noise in the timing of intracellular events. Proc Natl Acad Sci U S A 114:693–698 Gründling A, Bläsi U, Young R (2000) Biochemical and genetic evidence for three transmembrane domains in the class I holin, λ S. J Biol Chem 275:769–776 Gründling A, Manson MD, Young R (2001) Holins kill without warning. Proc Natl Acad Sci U S A 98:9348–9352 Hadas H, Einav M, Fishov I, Zaritsky A (1997) Bacteriophage T4 development depends on the physiology of its host Escherichia coli. Microbiology 143:179–185 Hagens S, Habel A, von Ahsen U, von Gabain A, Bläsi U (2004) Therapy of experimental Pseudomonas infections with a nonreplicating genetically modified phage. Antimicrob Agents Chemother 48:3817–3822 Hinton DM (2010) Transcriptional control in the prereplicative phase of T4 development. Virol J 7:289 Hobbs Z, Abedon ST (2016) Diversity of phage infection types and associated terminology: the problem with 'Lytic or lysogenic'. FEMS Microbiol Lett 363:fnw047 Hyman P (2019) Phages for phage therapy: isolation, characterization, and host range breadth. Pharmaceuticals (Basel) 12:35. https://pubmed.ncbi.nlm.nih.gov/30862020/ Hyman P, Abedon ST (2009) Practical methods for determining phage growth parameters. Methods Mol Biol 501:175–202 Hyman P, Abedon ST (2010) Bacteriophage host range and bacterial resistance. Adv Appl Microbiol 70:217–248 Kannoly S, Gao T, Dey S, Wang IN, Singh A, Dennehy JJ (2020) Optimum threshold minimizes noise in timing of intracellular events. iScience 23:101186 Kao SH, McClain WH (1980) Roles of T4 gene 5 and gene s products in cell lysis. J Virol 34:104–107 Kongari R, Rajaure M, Cahill J, Rasche E, Mijalis E, Berry J, Young R (2018) Phage spanins: diversity, topological dynamics and gene convergence. BMC Bioinf 19:326

382

J. J. Dennehy and S. T. Abedon

Kropinski AM (2018) Practical advice on the one-step growth curve. Methods Mol Biol 1681:41–47 Labrie SJ, Samson JE, Moineau S (2010) Bacteriophage resistance mechanisms. Nat Rev Microbiol 8:317–327 Los M, Wegrzyn G (2012) Pseudolysogeny. Adv Virus Res 82:339–349 Lwoff A (1953) Lysogeny. Bacteriol Rev 17:269–337 Maniloff J, Dybvig K (2006) Mycoplasma phages. In: Calendar R, Abedon ST (eds) The bacteriophages. Oxford University Press, Oxford, pp 636–652 Maniloff J, Haberer K, Gourlay RN, Das J, Cole R (1982) Mycoplasma viruses. Intervirology 18:177–188 Miller RV, Day M (2008) Contribution of lysogeny, pseudolysogeny, and starvation to phage ecology. In: Abedon ST (ed) Bacteriophage ecology. Cambridge University Press, Cambridge, UK, pp 114–143 Molineux IJ (2006) The T7 group. In: Calendar R, Abedon ST (eds) The bacteriophages. Oxford University Press, Oxford Moradpour Z, Sepehrizadeh Z, Rahbarizadeh F, Ghasemian A, Yazdi MT, Shahverdi AR (2009) Genetically engineered phage harbouring the lethal catabolite gene activator protein gene with an inducer-independent promoter for biocontrol of Escherichia coli. FEMS Microbiol Lett 296:67–71 Mosig G, Eiserling F (2006) T4 and related phages: structure and development. In: Calendar R, Abedon ST (eds) The bacteriophages. Oxford University Press, Oxford Narajczyk M, Baranska S, Wegrzyn A, Wegrzyn G (2007) Switch from theta to sigma replication of bacteriophage lambda DNA: factors involved in the process and a model for its regulation. Mol Gen Genomics 278:65–74 Ngo-Duc TT, Alibay Z, Plank JM, Cheeney JE, Haberer ED (2020) Gold-decorated M13 I-forms and S-forms for targeted photothermal lysis of bacteria. ACS Appl Mater Interfaces 12:126–134 Peng H, Borg RE, Dow LP, Pruitt BL, Chen IA (2020) Controlled phage therapy by photothermal ablation of specific bacterial species using gold nanorods targeted by chimeric phages. Proc Natl Acad Sci U S A 117:1951–1961 Raj A, van Oudenaarden A (2008) Nature, nurture, or chance: stochastic gene expression and its consequences. Cell 135:216–226 Ross A, Ward S, Hyman P (2016) More is better: selecting for broad host range bacteriophages. Front Microbiol 7:1352 Russel M, Model P (2006) Filamentous bacteriophages. In: Calendar R, Abedon ST (eds) The bacteriophages. Oxford University Press, Oxford, pp 146–160 Samson JE, Magadan AH, Sabri M, Moineau S (2013) Revenge of the phages: defeating bacterial defences. Nat Rev Microbiol 11:675–687 Shao Y, Wang I-N (2008) Bacteriophage adsorption rate and optimal lysis time. Genetics 180:471–482 Singh A, Dennehy JJ (2014) Stochastic holin expression can account for lysis time variation in the bacteriophage lambda. J R Soc Interface 11:20140140 Stern A, Sorek R (2011) The phage-host arms race: shaping the evolution of microbes. Bioessays 33:43–51 Tran TA, Struck DK, Young R (2005) Periplasmic domains define holin-antiholin interactions in T4 lysis inhibition. J Bacteriol 187:6631–6640 Wang I-N (2006) Lysis timing and bacteriophage fitness. Genetics 172:17–26 Wang I-N, Deaton J, Young R (2003) Sizing the holin lesion with an endolysin-β-galactosidase fusion. J Bacteriol 185:779–787 Wang I-N, Dykhuizen DE, Slobodkin LB (1996) The evolution of phage lysis timing. Evol Ecol 10:545–558 Wang I-N, Smith DL, Young R (2000) Holins: the protein clocks of bacteriophage infections. Annu Rev Microbiol 54:799–825

Phage Infection and Lysis

383

Wegrzyn G, Wegrzyn A (2005) Genetic switches during bacteriophage lambda development. Prog Nucleic Acid Res Mol Biol 79:1–48 Weinbauer MG, Wilehelm SW, Suttle CA, Garza DR (1997) Photoreactivation compensates for UV damage and restores infectivity to natural marine virus communities. Appl Environ Microbiol 63:2200–2205 Yacoby I, Benhar I (2008) Targeted filamentous bacteriophages as therapeutic agents. Expert Opin Drug Deliv 5:321–329 Yang H, Ma Y, Wang Y, Yang H, Shen W, Chen X (2014) Transcription regulation mechanisms of bacteriophages: recent advances and future prospects. Bioengineered 5:300–304 Yin J (1991) A quantifiable phenotype of viral propagation. Biochem Biophys Res Com 174:1009–1014 Young R (1992) Bacteriophage lysis: mechanisms and regulation. Microbiol Rev 56:430–481 Young R (2005) Phage lysis. In: Waldor MK, Friedman DI, Adhya SL (eds) Phages: their role in pathogenesis and biotechnology. ASM Press, Washington, DC, pp 92–127 Young R (2013) Phage lysis: do we have the hole story yet? Curr Opin Microbiol 16:790–797 Young R (2014) Phage lysis: three steps, three choices, one outcome. J Microbiol 52:243–258 Young R, Wang I-N (2006) Phage lysis. In: Calendar R, Abedon ST (eds) The bacteriophages. Oxford University Press, Oxford, pp 104–125 Young R, Wang I-N, Roof WD (2000) Phages will out: strategies of host cell lysis. Trends Microbiol 8:120–128

Part III History of Bacteriophages

The Discovery of Bacteriophages and the Historical Context William C. Summers

Contents Discovery in Science: No Eureka Moment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microbes Before Twort and d’Herelle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Viruses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . “Fact Making”: Bacterial Lysis and “Lytic Principles” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cultures, Cells, Microbial Mutations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Twort and d’Herelle: The Canonical Account . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Canonical History of Bacteriophage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Becoming Phage: 1920–1940 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Discovery Completed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discovery and Priority Revisited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

388 390 390 390 391 392 393 393 394 397 398 399 399

Abstract

Scientific discovery and priority are entangled concepts, and no more so than in the case of bacteriophage. The initial recognition of certain bacteriolytic phenomena constituted a “fact” that begged for explanation. The recognition of such a fact, however, cannot be fully accepted as a discovery. The “fact” must be integrated into existing or novel paradigms and explanations and meanings must be accepted by the scientific community. At some point during that integration process, the fact seems to expand into a stable discovery by accretion of other facts, contexts, theoretical explanations, and general understanding. Thus, discovery seems to be a process rather than event. The first “facts” related to bacteriophage were observations made in the second decade of the twentieth century. The integration of these observations into scientific paradigms was a process that was not fully complete until about 1940. The discovery of phage, in this sense, did not W. C. Summers (*) Yale University, New Haven, CT, USA e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_11

387

388

W. C. Summers

occur with a eureka moment, but instead by the usual scientific process of evolution over time. One casualty of this way of thinking about discovery is the habit of assigning priority based on something with intrinsic uncertainty.

Discovery in Science: No Eureka Moment Science as an institution and scientists as individuals assign much importance to the concept of “discovery” and its associated concept of “priority,” yet there has been surprisingly little scholarly attention devoted either to the definition of these concepts or their analysis. While scientists, and less frequently these days, historians, make lists of “firsts,” of “founding fathers,” of “landmarks,” and of other tokens of “credit,” the meagre scholarly analyses of the notion of discovery and priority suggests that these notions are vague, underdetermined, and usually poorly supported by the historical facts. The “discovery of bacteriophages” is no exception to this tendency. This essay will expand this narrative beyond the simple assigning of credit and pinpointing of some arbitrary date or event as a eureka moment. In his famous presidential address to the American Sociological Society in 1957, Robert Merton examined the concept of priority in scientific discovery, and argued that the often bitter and acrimonious disputes about priority of discovery arose from the nature of scientific norms themselves. He noted that “ownership” of a scientific discovery was unlike other ownership in that the product of discovery (knowledge) was freely given away, was exploited by others, and usually the only thing the “owner” had left at the end of the day, was “credit,” that is, the esteem and approbation of one’s colleagues: “In short, property rights in science become whittled down to just this one: the recognition by others of the scientist’s distinctive part in having brought the result into being.” (Merton 1957) The philosopher David Hull has shown in his analysis of the field of evolutionary biology that a crucial component of credit in science is that one’s work is used by others, i.e., cited in their own publications (Hull 1988). Ignored work is work that is interpreted as without value, so in a way, the currency of science can be seen as the extent to which others find one’s work useful for further scientific progress. Indeed, questions of priority and credit are not new phenomena in science. History of science is full of accounts of disputes, name-calling, bruised egos, and revisionist attributions. In 1834, the famously peripatetic perpetual secretary of the French Academy of Science, François Arago, noted in comments on the priority dispute over the discovery of hydrogen in 1793 by James Watt and Henry Cavendish that the description “‘about the same time’ proves nothing; questions as to priority may depend on weeks, on days, on hours, on minutes.” (Arago 1839) The concept of priority is especially fraught, partly because it depends on another notion, that of “discovery” which is even more difficult to clarify. One of the few scholarly studies of “discovery” in terms of its structure has been the work of Thomas Kuhn. In an underappreciated essay in Science before publication of Kuhn’s famous Structure of Scientific Revolutions, he examined the question of “what” and “when” with respect to a scientific “discovery” (Kuhn 1962). Kuhn

The Discovery of Bacteriophages and the Historical Context

389

points out that there are two main types of discoveries in science: “those which are not predicted from accepted theory in advance and which therefore [catches] the assembled profession by surprise” (e.g., X-rays) and the existence of “objects [that] had been predicted from theory before they were discovered, and [those] who made the discoveries therefore knew from the start what to look for” (e.g., missing elements in the periodic table). He then examines several cases of the first sort of discovery and shows that a fine-grained historical analysis can give only a vague answer to the what and when questions. What constitutes the actual discovery? At what point can a thing be said to be discovered? As Kuhn notes, “it must be arbitrary just because discovering a new sort of phenomenon is necessarily a complex process which involves recognizing both that something is and what it is. Observation and conceptualization, fact and the assimilation of fact to theory, are inseparably linked to the discovery of novelty” (Kuhn 1962). Frederic Holmes has studied the detailed laboratory records for two key discoveries in modern molecular biology, Seymour Benzer’s discovery of the T4rII genetic system for fine structure genetic analysis (Holmes 2006) and the verification of semiconservative DNA replication by Mathew Meselson and Franklin Stahl (Holmes 2001), both of which seem to be “simple” discoveries, and both of which have been characterized by their discoverers as happening over a very short time frame, what some have called eureka moments. When, however, the day-to-day experimental records are examined, it is not possible to assign a critical day or even week as the time when the “discovery” occurred. In Benzer’s case, about 9 months elapsed between his first puzzling observation to his clear conceptualization of the nature of the T4rII phenomenon. At no point during this period of intense investigation can one unambiguously say the “discovery” took place. The same conclusion applies to the work of Meselson and Stahl: the discovery of semiconservative replication seemed to gradually emerge during an extended period of stepwise investigations. “Within the rather vaguely delimited interval of internal history, there is no single moment or day which the historian, however complete his data, can identify as the point at which the discovery was made.” (Kuhn 1962) In the analysis of the discovery of bacteriophage, the same difficulties arise. First, we must decide just what constitutes discovery of phage. At one end of the spectrum is the recognition of a novel phenomenon, for simplicity, say “unexpected bacterial lysis,” and at the other end of the spectrum is the general scientific acceptance of “phage as bacterial viruses” with definite chemical and biological properties. Certainly sometime between these two events, we can locate the discovery of bacteriophage. Unfortunately, this simple answer is obviously unsatisfactory, so we must try to dissect in more detail, the events in this interval. One criterion for discovery that Kuhn suggests seems useful: the work that fits anomalous observations into the “lawlike” structure of existing knowledge. Thus, the interpretation of observations, usually based on extended investigations, is crucial in the process of discovery. Discovery involves, then, a statement of what is being discovered, not just a report of an unexpected observation, and in addition, conceptualization of the phenomenon in the language and theories of the science of the time.

390

W. C. Summers

In the account that follows, the discovery of bacteriophage will be described not as an event, but as a saga, playing out in the first half of the twentieth century, one in which observations, hypotheses, experiments, and technologies evolved from puzzling results termed “the lytic principle” to the widely accepted conception of “bacterial viruses.”

Microbes Before Twort and d’Herelle Viruses Once methods for growing microbes on sterile media were developed in the latter half of the nineteenth century, it became apparent that microbes were ubiquitous, and that methods for sterilization were important technologies. Thus, heat and chemical sterilization methods were studied, antiseptics were introduced into medical practice, and sanitary science blossomed. One technology that would be crucial in virology was that of filtration. First as a source of “sterile” water, and later as a criterion for microbial classification, filtration was developed both as a practical and scientific tool. The term “virus” has had a long and varying history: initially applied to any “poison” it was adopted in the nineteenth century to designate the pathogenic agent in materials such as putrefying organic matter, wound infections, and contagious diseases. In the 1890s, the application of sterilization filters to the “virus” of certain diseases, initially foot and mouth disease and tobacco mosaic disease, suggested that some infectious agents could pass through these filters that usually retained known bacteria and yeasts. Thus, two sorts of virus were distinguished, “filterable” (i.e., filter-passing) and “nonfilterable” (i.e., retained by these types of filters). For nearly 40 years, the term “filterable viruses” was used in the microbiological literature. Gradually, however, as bacteria and yeasts became better understood as microbes, they were no longer designated by the older term, virus, and the use of that term narrowed to be applied only to the “filterable viruses” and by about 1930 such microbes were commonly called simply “viruses.”

Filters The earliest effective filters that retained bacteria were made of unglazed porcelain and were devised by Charles Chamberland (1884), a young protégé of Pasteur. Such filters were devised as a way to obtain a continuous source of sterile water for laboratory use. They were widely commercialized for home use as well, and were marketed as Pasteur-Chamberland filters. Another type of filter, also for bacteriological use, was devised in Germany and used diatomaceous earth (called Kieselguhr) from a mine near Hanover owned by Dr. Berkefeld whose name became eponymous for this type of filter (Berkefeld 1890). These filters could be crudely calibrated as to their porosity and became a standard technology to estimate the size of particles, including microbes, that were retained or, alternately, “filterable.”

The Discovery of Bacteriophages and the Historical Context

391

At least by the first decade of the twentieth century, dialysis membranes were being employed to “filter” viruses, and by the 1930s, they had been developed to the point that Elford and Andrewes (1932) could use their carefully calibrated “gradocol” membranes to estimate the sizes of bacteriophages. Even though many agents of plant and animal diseases were accepted as “filterable viruses” by the early twentieth century (e.g., vaccinia, polio, herpes simplex, hog cholera, tobacco mosaic), the status of bacteriophage as filterable viruses was strangely controversial.

“Fact Making”: Bacterial Lysis and “Lytic Principles” One can speculate that from the first work with pure cultures of bacteria, both in liquid media and on solid surfaces, researchers would have observed occasional instances where the culture failed to grow as expected, or showed spontaneous lysis, or patchy growth on solid media. One can also speculate that since these events were “failures” rather than “positive” features, they were usually attributed to technical errors or other unexplained but trivial causes. Sometimes these observations, however, seemed to merit comment, and some were reported in the scientific literature. In a detailed survey of the early literature, Abedon and his colleagues (2011) have identified 30 reports of bacterial lytic phenomena that they consider as possibly the result of phage action published between 1896 and 1917. They term these reports as the “prehistory” of phage because they occurred prior to the canonical publications of F.W. Twort and F. d’Herelle, and any recognition of a recognizable “fact” in the sense of a reproducible observation about nature that seemed to be a “thing.”1 One clear step in the process we are investigating is the “making of the fact” that bacterial lysis is a real phenomenon of nature, a “fact” that needs explanation. This fact making seems to have occurred in the period of 1911–1917 with observations described and published by both F.W. Twort and Félix d’Herelle. d’Herelle reported that he first observed clear spots in cultures of a bacterium that he was developing for biological pest control in Mexico (Coccobacillus sauterelles) in 1911. In contrast to earlier investigators, d’Herelle took this observation as a fact that needed explanation and he believed the spots to be the result of an infection of the culture with an “ultramicrobe.” This notion was not without precedence since the pathogenicity of the hog cholera bacillus was, at the time, attributed to the cooperation of an invisible filterable agent with the bacterium itself (De Schweinitz and Dorset 1903). However, d’Herelle’s 1911 observation, in itself, cannot be said to constitute the “discovery” but rather the recognition of a new fact of nature, perhaps the beginning of the “discovery.” Historians of science recognize a distinction between a “fact” and an “interpretation” of that fact. A fact is often an experimental or natural observation, e.g., the observation of plaques or the lysis of a culture, which by themselves are noncontroversial. Their interpretation, however, attempts to fit the facts into accepted paradigms or theories, or as Kuhn (1962) has argued, may force paradigm changes. New technologies often are thought of as “fact making” because they provide new types and extents of observations.

1

392

W. C. Summers

Frederic W. Twort published work in 1915 in which his principle aim was to find ways to grow filterable viruses in vitro with various culture media. He reasoned that viruses need nutritional cofactors to grow in vitro (based on his earlier work on finding essential growth factors for Johne’s bacillus). Since preparations of viruses (such as vaccinia) usually (always?) contained bacteria, he reasoned that perhaps these bacteria produced the essential cofactors for viral growth. His research program consisted of exploring various mixed cultures of bacteria and filterable viruses on various kinds of media in attempts to grow the filterable viruses outside the living host. Just what that appearance might be was, however, unspecified in Twort’s account (Twort 1915; Pirie 1990) To understand the context of fact making with respect to bacterial lysis, it is relevant that there were two other phenomena that involved bacterial lysis that were possible competitors for the Twort-d’Herelle phenomenon as a novel fact. The first was complement-mediated bacteriolysis, called at the time “Pfieffer’s phenomenon” after Richard Pfeiffer, the German microbiologist who first described the lysis of bacteria by serum from immunized animals (Pfeiffer 1894). In 1919, Jules Bordet received the Nobel Prize for his work on clarifying and explaining the mechanism of this immune phenomenon. A bit later, the second confounding fact was the discovery of lysozyme by Alexander Fleming (1922), an enzyme from tears and hen egg white that could lyse certain bacteria. The lytic phenomenon observed by Twort and d’Herelle might be simply another expression of one of these two known facts of nature. Indeed, one of the major objections to d’Herelle’s interpretation of bacteriolysis that he called bacteriophage came from Bordet and Ciuca (1920) as well as Bordet’s protégé, Andre Gratia (1921) all of whom preferred to interpret d’Herelle’s phenomenon as related to immune cytolysis or possibly enzymatic action by a “vitiated” cellular enzyme.

Cultures, Cells, Microbial Mutations Microbiology up until the 1930s was divided by two opposing conceptual views: one held that bacteria represented only one species of organism and that the diversity of forms that could be observed was the result of differentiation into differing life stages, or growth phases. This view was termed pleomorphism, and was supported by the well-known variation known as bacterial sporulation among other observations. The opposing view held that each observed bacterial form represented a fixed species. This position was termed monomorphism. While the “pure culture technique” introduced by Koch and Pasteur showed that many bacteria seemed to “breed true” and hence supported the monomorphism school, the observation of changes in cultural behavior over time and the appearance of new forms of bacteria with growth phase led to a widely accepted belief in the phenomenon termed “cyclogeny” during the nineteen teens, 20s and 30s (Amsterdamska 1991). Cyclogeny held that bacterial cultures went through cycles of growth and change and that new forms of the same organism appeared during these cycles. The cyclogenic view was dependent on the view that the relevant unit of study was the entire bacterial culture. One of the important problems that attracted

The Discovery of Bacteriophages and the Historical Context

393

such study was the change in pathogenic virulence of bacterial cultures with prolonged growth, in the case of the pneumococcus, the “rough-smooth” colony morphology transitions. In a series of insightful and clarifying studies of this phenomenon of “dissociation” (so-called because one pure line of bacteria could dissociate into two or more types), the young, soon to be famous as a science writer, Paul deKruif (1962), showed that if one took the bacterial cell, rather than the entire culture, as the unit of study, it was clear that the changes seen in cyclogeny were really bacterial mutations that rarely occurred, but that could eventually be selected by particular growth conditions. DeKruif’s pioneering work opened the way for more extensive study of bacterial genetics and led to the demise of the cyclogenic theories. The importance of this shift from the culture to the cell as the relevant object of study was important in the early work on bacteriophage (Summers 1991). The most obvious effect of phage that was often observed was the lysis of an entire culture, and early research frequently described phage as a “lytic principle” because of this. Another important early phage phenomenon related to overall cultural conditions was the phenomenon of “secondary cultures.” If a culture that was lysed by phage was incubated for some length of time, often days or weeks, it was observed that bacteria grew up and gave rise to a “secondary” culture. The nature of these secondary cultures was rather mysterious: often the bacteria, when grown on solid media, were of different colony morphology from the original culture and frequently these bacteria could not be lysed by phage again. Sometimes the cultures, when filtered and tested on other bacteria, promoted lysis again. Some investigators interpreted these changes as induction of cylcogenic changes, other investigators suggested that the phage acted as a mutagen and changed the bacteria to new forms, and yet others suggested that mutations to phage resistance occurred with out-growth of the resistant forms. Quite a bit later, it was realized that some of the secondary cultures were lysogens that were both immune to phage infection and producing low levels of spontaneously induced phage. Again, when the focus of investigation shifted away from gross cultural behavior to the cell-phage interactions, it became clearer what was happening in the phenomenon of secondary cultures.

Twort and d’Herelle: The Canonical Account2 The Canonical History of Bacteriophage In 1915, Frederick W. Twort, a British microbiologist working at the Brown Institution, a veterinary medicine establishment in London, was trying to grow the so-called filterable viruses on laboratory culture media without living cells, and 2

Full-length biographies of Twort (1877–1950) (In Focus, Out of Step, Alan Sutton 1993, by his son Anthony Twort) and of d’Herelle (1873–1949) (Félix d’Herelle and the Origin of Molecular Biology, Yale University Press 1999, by the present author) have been published. See also Fildes (1951).

394

W. C. Summers

from a sample of vaccinia material (typically obtained from pustules on calves) noted that some of the bacterial colonies that grew out exhibited a sporadic phenomenon of disintegration which he called “glassy transformation.” The material from the disintegrated colonies could cause further disintegration and glassy transformation of other colonies. Twort (1915) published this finding in The Lancet with certain speculations as to the interpretation of this result. Later, in 1917, by an entirely different route of investigation, Felix H. d’Herelle, a French-Canadian microbiologist working at the Pasteur Institute, observed lysis of cultures of dysentery bacilli, and on solid medium he observed the production of clear spots which were devoid of bacteria. This lytic phenomenon, and the production of clear spots, could be propagated indefinitely even at high dilution, and d’Herelle (1917), when he published this finding in Comptes Rendus de l’Academie des Sciences, proposed that this phenomenon was caused by an invisible microbe that parasitized the dysentery bacteria, and further, that this killing of dysentery bacteria by this invisible parasite could explain immunity and recovery from infectious diseases such as dysentery. d’Herelle’s formulation of bacteriophage (his term) as a microbial infection of bacteria, and as a biological agent of immunity, attracted the attention of microbiologists interested in infectious diseases, especially Jules Bordet, the recent Nobelist and director of the Pasteur Institute in Brussels. Because d’Herelle framed his work as a challenge to Bordet, the latter responded with both a scientific response, but also by championing Twort’s earlier, neglected, work as undercutting d’Herelle’s priority claims to having “discovered” bacteriophage. Thus began a long and acrimonious priority dispute, and for years, agnostic authors adopted the terminology “Twort-d’Herelle Phenomenon” to avoid both the priority issue and d’Herelle’s conception of bacteriophage as “ultraviruses” (Duckworth 1976; van Helvoort 1992)

Becoming Phage: 1920–1940 When considered as a process of evolving understanding and conceptual refinement, the “discovery” of bacteriophage spans approximately two decades, from the recognition of a “fact of nature” to the nearly universal acceptance of phage as viruses that infect bacteria. During these decades, both technical advances and changing understanding of viruses in general and bacteria in particular influenced the theorizing about bacteriophage. Several conceptual problems and controversial ideas needed resolution before phage could be fully accepted as viruses. These included: (1) Are phages particulate? (2) Is there one “species” of phage or many? (3) Are phages living? (4) What is the chemical composition of phage? (5) How big are phages? (6) Are their properties heritable? (7) How do they reproduce? (8) What is lysogeny? (9) Where do phages come from? In the early papers on phage, there was considerable debate and ambiguity as to the nature of phage. Twort’s original paper was not only agnostic on the nature of phage, but he seemed to invite confusion by his musings: the “dissolving substance” might be a “separate form of life” or, alternately, “an enzyme secreted by the

The Discovery of Bacteriophages and the Historical Context

395

micrococcus which leads to its own destruction and the production of more enzyme” but “the possibility of its being an ultra-microscopic virus has not been definitely [italics in the original] disproved.” (Twort 1915) On the other hand, he refers to “spots” or “points” of bacterial dissolution which certainly suggest he was observing phage plaque formation. d’Herelle (1917), in his first account of bacteriophage, based on his notion that pathogenicity of dysentery was controlled by the interaction of both bacteria and a filterable agent, noted the same spots, but interpreted them as colonies of the filterable agent that reproduced and lysed the bacteria locally. His insight that such colonies or taches vierges (Fr. virgin spots), later called plaques, might provide a way to enumerate or quantify the bacteriophage represent a clear advance in understanding the concept of phage. Further, as he suggested, this fact suggested that phage were particulate, like known microbes, rather than “fluid” or “in solution” like known enzymes. D’Herelle (1926) even cited the authority of Albert Einstein, whom he consulted on the interpretation of a dilution experiment in which the distribution of activity followed Poisson statistics expected for quantized particles. In spite of these apparently convincing experiments, the particulate nature of phage was a matter of contention for some years. It was the invention and application of the electron microscope to phage preparations in the late 1930s that won over the diehards. At that point, the discovery of the particulate nature of phage seemed complete. Even so, the status of phage as biological objects was unclear for a long time: Was there one “species” of phage with variable manifestations or did phage exist in many different species as did other organisms? If phage was inanimate like a chemical substance, it might be expected to be of only one form, such as glucose is of only one form. d’Herelle believed in the “unicity” of phage, at least originally, a belief perhaps a remnant of the earlier monomorphism/pleomorphism debate in bacteriology. Later work, including that by d’Herelle, however, showed that phage isolates were distinct in physical properties as well as antigenic properties, undermining the concept of unicity and leading to the accepted belief by the end of the 1930s that phage exist in biologically distinct and stable forms similar to other organisms. The question of the status of phage as being “alive” was an intriguing and fraught question at the beginning. This is partly because “life” was described in terms of macroscopic properties of familiar organisms: movement, irritability, assimilation of nutrients, reproduction, variation, and the like. With nearly invisible organisms, such as phage as well as other viruses, these properties were not easily assessed. Reproduction, and to some extent, variation were observable by extant methods, but metabolism and sensibility or irritability were not. Attempts to measure respiration in concentrated phage preparations were unsuccessful (Bronfenbrenner 1926). Likewise, physical properties such as relative stability to heat and other stresses suggested that phage were different from other kinds of living things. The crystallization of some animal and plant viruses in the 1930s raised the general question as to the living status of all virus-like entities. Crystallization was taken to be the sine qua non of chemical purity, and seemed to put viruses, and by analogy, phages, into the field of the chemist rather than the biologist. While this ambiguity has persisted in the minds of some, microbiologists soon came to accept the fact that viruses,

396

W. C. Summers

including phages, are sui generis and came to be considered, as d’Herelle had argued, “obligate intracellular parasites.” Gradually by the end of the period of consolidation of beliefs about phage, the sizes and chemical nature of phage were clarified by the application of new laboratory methods: sedimentation and ultrafiltration. High speed centrifuges, developed by Svedberg and others, and calibrated membranes for filtration experiments allowed experimental determination of sizes of the phage particles. Better purification of phages by sedimentation and dialysis allowed early chemical analyses especially by Martin Schlesinger (1936) to convince phage workers that the particles were of discrete sizes (Fig. 1) and that they contained both protein and nucleic acids. A central concern in the development of consensus as to the nature of phage that needed resolution to complete the process of “discovery” was the problem of reproduction. To be recognized as living, phages had to reproduce and in the third and fourth decades of the twentieth century, the methods for studying reproduction of microbes were lacking. Those powerful tools came along a bit later: isotopic tracers, chromatographic methods of chemical analysis, and both optical and physical methods of microchemistry. Likewise, the status of microbes, including bacteria, as genetic organisms was still debated. Without clear sexual reproduction and a cytologically visible nucleus with chromosomes, many researchers viewed microbes as having some fundamentally different type of heredity from “higher” organisms. Further, the puzzling phenomenon of lysogeny in which phage could apparently

Fig. 1 Properties of a set of phages isolated in d’Herelle’s laboratory in Paris in the 1930s. Note the small phage φ X 174, which later would be widely used in phage research (Reproduced from Sertic and Boulgakov 1935)

The Discovery of Bacteriophages and the Historical Context

397

establish symbiosis with the bacteria suggested a novel and problematic biology. The recent discovery of autocatalytic proenzymes such as pepsinogen and trypsinogen, which in some ways mimicked phage in being able to promote the appearance of new proteolytic activity, provided competing explanations for phage reproduction (Northrop 1939). Even as late as the 1950s, some hold-out researchers supported this model. During the 1930s, however, refinements in simple experiments such as plaque isolation, which showed that phage properties such as host range, burst size, plaque morphology, and particle size were stable, convinced most workers that phage behaved as a simple genetic organism. Occasional variation of plaque forms, for example, suggested that mutations in phage could be obtained. More and more, phages were being compared to “raw genes.” Indeed, lysogeny was becoming seen as a genetic property of the bacteria which harbored phage in some latent form, comparable to the latent genetic plans in developing higher organisms, termed Anlagen by the German embryologists (Wollman 1935; Burnet 1968). The use of two key experimental designs did much to complete the acceptance of phage as bacterial viruses. These were the “one-step growth” experiment and the “single-burst” experiment (Ellis and Delbrück 1939). Adding phage to a bacterial culture, which was then sampled at various times and centrifuged to separate bacteria from phage, showed the phage seemed to attach to the bacteria, and mostly disappeared as free phage. After a reproducible period of time, characteristic of that particular strain of phage, there was an abrupt increase in free phage (the step) accompanied by the disappearance by lysis of the bacteria. This increase in phage, often a 100-fold or more, was taken to represent the intracellular reproduction of the input phage. The step-wise nature of this process suggested an infectious cycle driven by the phage, rather than a more or less continuous process of bacterial metabolism (Fig. 2). The other experiment used immediate dilution of the infected culture to the point where the Poisson distribution predicted one or fewer bacteria per culture tube. After the time expected for the “burst,” the entire culture tube was assayed for phage. The results showed that some tubes had no phage and others had many phages, the number being rather constant and characteristic of the particular phage strain. This type of experiment suggested, again, that phage reproduction seemed to be an all-ornone event based on a particular bacterium being infected with a phage. These crucial experiments convinced most phage workers that d’Herelle was fundamentally correct in his conception of phage as a bacterial virus.

Phage Discovery Completed By 1940, the phenomenon called “bacteriophage” was widely recognized among the small coterie of interested microbiologists to be understood and accepted as being viruses of bacteria. This recognition, however, did not mean that phages were well understood, just that their existence and basic properties were finally agreed upon, the basic requirement to say that they had been “discovered.”

398

W. C. Summers

x

Log phage concentration

9

x

x x

8 x

x

7 x

6

x

x

2-15 x 3-3 + 3-13

x

5 +

4 Min. 0

++

+

20

40

60

80

100 Time

120

140

160

180

120

Fig. 2 One-step phage growth curve showing waves or “bursts” of phage (Reproduced from Ellis and Delbrück 1939)

For example, in their landmark 1939 paper on “The Growth of Bacteriophage,” Emory Ellis and Max Delbrück start off by stating “Certain large protein molecules (viruses) possess the property of multiplying within living organisms.” They then go on to provide support for the bacterial virus model of phage proposed by d’Herelle in his initial publication. Major textbooks written at the end of this period likewise have abandoned the rather vague “vitiation” ideas of Bordet and adopted the conception of phage as virus (Bronfenbrenner 1928; Zinsser and BayneJones 1938). Clearly there was no doubt that there was acceptance and consensus to the concept of bacteriophage as a virus of bacteria even though there was still substantial uncertainty about just what a virus is. As late as 1957, Andre Lwoff was exasperated to the point of asserting that “viruses are viruses” . . . that is all there is to it (Lwoff 1957). In Kuhn’s discovery process, by 1940 (and arguably earlier), it was clear that the phage discovery had occurred. The question, however, remains, can we pinpoint when this “discovery” took place? By the somewhat fine-grained historical account presented here, it seems that this question is misguided. There is no one moment, the eureka moment, when we can say that the concept of phage went from being tentative to being certain. This evolution happened gradually over a rather long period of controversy, uncertainty, and accumulation of facts and knowledge. Phage discovery was a process, not an event.

Discovery and Priority Revisited As David Hull and others have argued, scientists are usually intensely concerned with “credit” for their work, and this takes the form of approbation and value of one’s work by esteemed colleagues. New knowledge is the prize in science as scientists

The Discovery of Bacteriophages and the Historical Context

399

struggle to make sense of the natural world. New knowledge is the currency of the scientific realm, and priority of discovery creates wealth. Priority, then, as Arago recognized nearly two centuries ago, is crucial. There are, however, two difficulties: how to determine priority and how to determine if the event assigned to that priority is of the assumed significance. In recent times of nearly instantaneous mass communication, the scientific community has generally agreed that priority can be established by the date of wide dissemination of the new knowledge, usually by some means of publication. The detailed rules for this consensus vary by discipline and custom, and often simultaneous publication occurs by accident or design. History of science gives us plenty of examples of priority disputes and their resolution. In the unusual case of phage, the priority dispute between Twort and d’Herelle was precipitated by a third party, Bordet, for his own reasons. Again, this priority dispute was interesting because the main controversy was not about timing but about the nature of the actual thing that was discovered. Were Twort and d’Herelle even talking about the same natural phenomenon? If not, of course, there would be no priority dispute. As this controversy wore on for over a decade, with threats of judicial proceedings and public third-party adjudication of experimental results, the nature of the thing that was claimed to be discovered was not at all agreed upon (Gratia 1930). What was discovered? A bacterial virus? A lytic enzyme? A bacterial gene? An autocatalytic protein? From the case of bacteriophage, it is clear that priority and discovery are entangled problems. While priority presumes, and may require a eureka moment, discovery, as Kuhn first showed, is rarely so easily characterized.

Cross-References ▶ Detection of Bacteriophages: Electron Microscopy and Visualization ▶ Detection of Bacteriophages: Statistical Aspects of Plaque Assay ▶ Early Therapeutic and Prophylactic Uses of Bacteriophages ▶ Phage Infection and Lysis ▶ Temperate Phages, Prophages, and Lysogeny

References Abedon ST et al (2011) Bacteriophage prehistory. Is or is not Hankin, 1896, a phage reference? Bacteriophage 1:174–178 Amsterdamska O (1991) Stabilizing instability: the controversy over cyclogenic theories of bacterial variation during the interwar period. J Hist Biol 24:101–222 Arago F (1839) Historical eloge of James Watt by M. Arago . . . Trans. from the French, with additional notes and an appendix by James Patrick Muirhead. J. Murray, London Berkefeld W (1890) Simple means for scouring filter. US Patent 422,112, 25 Feb 1890 Bordet J, Ciuca M (1920) Exsudats leucocytaires et autolyse microbienne transmissible. CR Soc Biol 83:1293–1295 Bronfenbrenner J (1926) Does bacteriophage respire? Science 63:51–52

400

W. C. Summers

Bronfenbrenner J (1928) Virus diseases of bacteria – bacteriophagy. In: Rivers TM (ed) Filterable viruses. Williams and Wilkins, Baltimore, pp 373–417 Burnet FM (1968) Changing patterns: an atypical autobiography. William Heinemann, Melbourne Chamberland C (1884) Sur un filtre donnant l’eau physiologiquement pure. CR Acad Sci Paris 99:247–248 d’Herelle F (1917) Sur un microbe invisible antagoniste des bacilles dysentériques. CR Acad Sci Paris 165:373–375 d’Herelle F (1926) The bacteriophage and its behavior. Williams and Wilkins, Baltimore De Schweinitz EA, Dorset M (1903) A form of hog cholera not caused by the hog-cholera bacillus, vol 41. US Bureau of Animal Industry Circular, Washington, DC DeKruif P (1962) The sweeping wind: a memoir. Harcourt, Brace and World, New York Duckworth DH (1976) Who discovered bacteriophage? Bact Rev 40:793–802 Elford WJ, Andrewes CH (1932) The sizes of different bacteriophages. Brit J Exptl Path 13:446–456 Ellis EL, Delbrück M (1939) The growth of bacteriophage. J Gen Physiol 22:365–384 Fildes P (1951) Frederick William Twort. 1877–1950. Obit Notice Fellows R Soc 7:504–517 Fleming A (1922) On a remarkable bacteriolytic element found in tissues and secretions. Proc R Soc Lond B: Biol Sci 93:306–317 Gratia A (1921) Studies on the d'Herelle phenomenon. J Exp Med 34:115–126 Gratia A (1930) The Twort phenomenon and bacteriophagy. CR Soc Biol 105:219–222 Holmes FL (2001) Meselson, Stahl, and the replication of DNA: a history of “the most beautiful experiment in biology”. Yale University Press, New Haven Holmes FL (2006) In: Summers WC (ed) Reconceiving the gene: Seymour Benzer’s adventures in phage genetics. Yale University Press, New Haven Hull DL (1988) Science as a process: an evolutionary account of the social and conceptual development of science. University of Chicago Press, Chicago Kuhn T (1962) Historical structure of scientific discovery. Science 136:760–764 Lwoff A (1957) The concept of virus. J Gen Microbiol 17:239–253 Merton RK (1957) Priorities in scientific discovery: a chapter in the sociology of science. Am Soc Rev 22:635–659 Northrop JH (1939) Crystalline enzymes. The chemistry of pepsin, trypsin, and bacteriophage. Columbia University Press, New York Pfeiffer R (1894) Weitere Untersuchungen über das Wesen der Choleraimmunität und über specifisch baktericide Processe. Zeitsch f Hyg u Infekt 18:1–16 Pirie NW (1990) The career of F.W. Twort. Nature 343:504 Schlesinger M (1936) The Feulgen reaction of the bacteriophage substance. Nature 138:508–509 Sertic V, Boulgakov N (1935) Classification et identification des typhi-phages. CR Soc Biol 119:1270–1272 Summers WC (1991) From culture as organism to organism as cell: historical origins of bacterial genetics. J Hist Biol 24:171–190 Summers WC (2011) In the beginning. . .. Bacteriophage 1:50 Twort FW (1915) An investigation on the nature of ultra-microscopic viruses. Lancet 186:1241–1243 Van Helvoort T (1992) Bacteriological and physiological research styles in the early controversy on the nature of the bacteriophage phenomenon. Med Hist 36:243–270 Wollman E (1935) The phenomenon of Twort-d'Herelle and its significance. Lancet 226:1312–1314 Zinsser H, Bayne-Jones S (1938) Textbook of bacteriology, 8th edn. Appleton Century, New York

Early Therapeutic and Prophylactic Uses of Bacteriophages Nina Chanishvili and Zemphira Alavidze

Contents Introduction: A Brief History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Therapy for Wound Treatment, Surgery, and Dermatology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Therapy for Treatment of Enteric Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Prophylactic Use of Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intravenous Staphylococcal Bacteriophage: The Highest Achievement of the Georgian Scientists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

402 406 410 414 416 423 425

Abstract

The long history of the discovery and therapeutic use of bacteriophages, especially non-English-language journals, is often overlooked by modern researchers. While this body of evidence does not provide suitable proof of the safety and efficacy of phage therapy, it nonetheless demonstrates the potential for safety and efficacy and as such is worthy of attention by modern-day researchers in the field. Here, we discuss some of the early work carried out to develop clinical applications of phage therapy for many diseases, with a particular focus on the huge amount of work carried out behind the Iron Curtain, in countries where phage therapy was (and in many cases still is) more commonly used than elsewhere in the world.

N. Chanishvili (*) George Eliava Institute of Bacteriophage, Microbiology and Virology (EIBMV), Tbilisi, Georgia e-mail: [email protected] Z. Alavidze Phage Therapy Center, Tbilisi, Georgia Eliava BioPreparations, Tbilisi, Georgia e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_12

401

402

N. Chanishvili and Z. Alavidze

Introduction: A Brief History It has been argued that Ernest Hanbury Hankin may have been the first to publish on a bacteriophage-related phenomenon, when in 1896 he reported an agent which he demonstrated had antibacterial properties that could reduce titrers of the bacterium Vibrio cholerae in laboratory culture (Hankin 1896). The agent was found in the river Ganges in India, which was regarded by the population as a holy river as it was believed that its waters saved people from diseases (Adhya and Merrill 2006). Approximately at the same time, in 1989 Nikolai Fedorovich Gamaleya (a Ukrainian doctor bacteriologist-epidemiologist) published an article in the Russian Archives of Pathology, Clinical Medicine and Bacteriology (Gamaleya 1898), in which he described bacterial lysis of anthrax in distilled water after which an unknown agent was produced which in turn caused so-called transmissible lysis of other cultures of Bacillus anthracis. Other early observations of the phenomenon of bacterial lysis were made by Kruse and Pansini (cited from Kazarnovskaya 1933) who noticed that old pneumococcal cultures turned transparent due to death of bacterial cells. Later, Eijkman (1901) showed that an extinction of bacterial cells is not associated with exhaustion of media as it was believed before. Eijkman showed that after the death of one bacterial culture, it was not possible to cultivate another one, even of a related bacterial culture, on the same agar plate. According to Eijkman (1901), the extinction of bacterial culture is a result of accumulation of toxic substances inhibiting bacterial growth and in some cases dissolving them. Condari and Kurpjuweit (cited from Kazarnovskaya 1933) found out that the inhibiting substance described by Eijkman (1901), if present in liquid media, was destroyed in a short period of time after heat treatment at 60–70
C. Long-term observations suggested that these substances were produced by bacteria themselves due to their intracellular fermentative activities. They concluded that autolysis of bacterial cells occurs as a result of self-poisoning with the accumulated toxins. The active principle of this phenomenon was called autotoxin. The autotoxins caused lysis of old cultures; however, the autotoxins could not multiply like the d’Herelle phenomenon, and filtered autotoxins did not maintain an ability to lyse bacterial cultures (Kazarnovskaya 1933). Emmerlich and Low (1899) observed formation of an agglutinated mucous pellet of fresh bacterial culture in liquid media. This process was associated with increasing transparency of the media occurring within 2–3 days. Transmission of the transparent media into a tube with growing culture caused lesser pellet formation, and the tube turned transparent. After 2–3 inoculations of the transparent media into the new culture, no more pellet developed. Emmerlich and Low explained this phenomenon by production of bacteriolytic enzymes (Emmerlich and Low 1899). In 1915 the English bacteriologist Frederick Twort (1915), who is recognized to be an original discoverer of bacteriophages, published in The Lancet his observations about the destruction of the bacterial cells of Staphylococcus aureus and that the disrupted glassy areas might be transmitted to another culture and cause the same phenomenon. Twort (1915) believed that this phenomenon was caused by an

Early Therapeutic and Prophylactic Uses of Bacteriophages

403

enzyme secreted by the bacteria and called the contagion “the bacteriolytic agent.” Twort’s work may have been ignored if Jules Bordet and Andre Gratia had not rediscovered his paper (Summers 2012). Two years after Twort’s publication, Felix d’Herelle published his article, Sur un microbe invisible antagoniste des bacilles dysentériques (d’Herelle 1917). D’Herelle started work at the Pasteur Institute in 1911, where he was engaged in the vaccine development and manufacturing process. In his spare time, he examined samples from dysentery patients. From the feces of several patients, he isolated an anti-Shiga “microbe” which was multiplied through many serial passages on its host bacterium and which could produce tiny clear circles on the lawn of the same Shigella culture. In 1917 he presented the results of his work to the French Academy of Sciences. Monitoring the patients with bacillary dysentery, d’Herelle discovered that shortly before the disappearance of blood in stool samples and recovery, some “agent” appears in the intestines, with an ability to dissolve the dysentery bacteria. In patients who have died of dysentery, the agent was never detected. This agent had the ability to reproduce itself only in presence of the host bacteria. D’Herelle gave the name to this agent – “bacteriophage” from Greek “bacteria eater.” He assumed that bacteriophages were tiny creatures, much smaller than bacteria, with a corpuscular structure that parasitized bacteria and acted through production of specific enzymes. It is striking that these conclusions were made solely on the basis of d’Herelle’s empirical observations and intuition, since visualization of bacteriophage became possible only 22 years later (Summers 2012). The potential efficacy of phage preparations as antibacterial agents was demonstrated by tests on chickens using typhoid phages isolated from poultry. Due to the application of these phages, bird mortality was lowered from 95% to 5% (Kazarnovskaya 1933). As d’Herelle was sure that all types of bacteria had corresponding phages circulating in natural sources, the results of his animal experiments led to a decision to commence clinical trials to treat shigellosis at the Hôpital des Enfants-Malades in Paris under the clinical supervision of Professor VictorHenri Hutinel, the hospital’s Chief of Pediatrics (Kazarnovskaya 1933; Summers 1999). To confirm bacteriophage safety prior to the experiment, the phage preparation was ingested by d’Herelle, Hutinel, and some other hospital staff members – as was rather usual in those days. On the next day the bacteriophage preparation was given to a 12-year-old boy with severe dysentery. Even after a single administration, the disease symptoms were reduced, and complete recovery was achieved within a few days. Soon after this, three more patients diagnosed with bacterial dysentery were treated with one dose of the preparation, with improvement observed within 24 h after administration. D’Herelle, however, did not rush with publication of these results, so the first reported clinical application of phages belongs to Richard Bruynoghe and Joseph Maisin (1921), who used bacteriophages to treat skin infections caused by Staphylococcus. Bacteriophages in that study were injected into and around surgically opened lesions. According to this publication, a regression of the infection occurred within 24–48 h. A number of promising studies followed (Rice 1930; Schless 1932; Stout 1933). Encouraged by these early results, d’Herelle and others continued studies of the therapeutic use of phages (e.g., d’Herelle used

404

N. Chanishvili and Z. Alavidze

various phage preparations to treat thousands of people having cholera and/or bubonic plague in India) (Kazhal and Iftimovich 1968; Summers 1999, 2001). An enormous number of publications dedicated to discussions about the nature of bacteriophages were published in the 1930s and 1940s. Scientists were divided into two camps, supporting “precursor” and “viral” theories. In parallel with this, experiments on therapy and prophylaxis had been conducted by scientists working around the world. This information is impossible to embrace in one chapter. Therefore, we will discuss only a tiny number of publications illustrating results of phage treatment described in the early scientific literature. The discussion about the nature of bacteriophages continued for a long time until the invention of the transmission electron microscope by Max Knoll and Ernst Ruska, with the first phage images made in 1939 (Ruska 1940; Ackermann and Dubow 1987; Summers 2012). In 1933, Bayne-Jones and Sandholzen (1933) managed to film the bacterial burst caused by bacteriophage infection, albeit without coming to the right conclusion. According to the precursor theory, bacteriophages are developed endogenously; they exist in bacteria as a precursor which become lytic agents in the same way as trypsinogen is transformed into trypsin. This theory was supported by, among others, Gildemeister (1921), Bordet (1925), Northrop (1937, 1938), and Kruger and Scribner (1941). Another hypothesis was presented by Beckerich and Haudoroy (1922) who considered bacteriophages to be an external parasite which participated in the sexual cycle of bacteria. This theory was shared by the Russian scientists Jukov-Werezhnikov and Fruauf (1934), who considered that bacteriophages were a sexual (hereditary) non-visual form of bacteria. They assumed that bacteriophages fertilize bacteria, similarly as plant pollen fertilizes an ovule, during which phages force the bacterial cell to turn into an invisible stage. Another view was suggested by Gamaleya, who assumed that the bacteriophage phenomenon was associated with the production of a hormone “clustin.” He supposed that overproduction of clustin caused dissolution of the bacterial cell, turning it into a filterable form. The filtered clustin would cause dissolution again and again, over successive passages (cited from Jukov-Werezhnikov and Fruauf 1934). In 1934 a critical review of the available literature on phage therapy, based on 150 references, was published by the Journal of the American Medical Association, which drew conclusions not in favor of the therapy. It was directly stated that, “Experimental studies of lytic agent called ‘bacteriophage’ have not yet disclosed its nature. D’Herelle’s theory that this is a living virus parasite of bacteria has not been proved. On the contrary, the facts appear to indicate that the material is inanimate, possibly an enzyme” (Eaton and Bayne-Jones 1934). Such an assessment likely had a negative impact on investment in major research and production of bacteriophage, at least in the USA (Myelnikov 2018). This fact might have contributed to d’Herelle’s decision to accept George Eliava’s invitation to conduct his research in Georgia. In the history of medicine, little is known about George Eliava, who was a colorful central figure in phage history. Obviously, the discovery of bacteriophages was inevitable; Eliava was one of those scientists who had observed spontaneous lysis of bacteria but could not explain this phenomenon. This is why he was strongly

Early Therapeutic and Prophylactic Uses of Bacteriophages

405

interested in d’Herelle’s research. He supported d’Herelle’s theory about the viral nature of this phenomenon, along with his colleagues Asheshov, Bronfenbrenner, Gratia, Flu, Wollman, etc. Eliava first learned about d’Herelle and his discovery due to vigorous discussions held at the Pasteur Institute, where he had been sent by the Georgian government to advance his knowledge of bacteriology in 1919. Eliava spent several years working at the Pasteur Institute, where he met d’Herelle during one of his visits. This was the start of their collaboration and friendship. Eliava invited d’Herelle to Georgia, where they decided to found a World Center of Bacteriophage Research in that country. Without the support that Eliava provided to d’Herelle, much of the early knowledge of phage therapy may not have been achieved. Unfortunately, with his progressive thinking, tireless activities, and close collaboration with many foreign scientists, including d’Herelle, George Eliava became a victim of Stalin’s regime in 1937. He was pronounced to be a “people’s enemy” and was executed (Georgadze 1974; Chanishvili 2012a, 2012b). D’Herelle spent altogether 18 months during 1933 and 1934 in Georgia, collaborating with Eliava and other Georgian colleagues (Chanishvili 2012a, 2012b). D’Herelle intended to move to Tbilisi permanently (a cottage built for his use still stands on the institute’s grounds); however, his intentions could not be realized. Nevertheless, the seeds of phage therapy found a good ground to grow and develop further in the former Soviet Union (Fig. 1).

Fig. 1 The first conference on phage therapy in Tbilisi, Georgia, 1933–1934 (from left to right: unknown (sitting) Prof. Alexander Tsulukidze (standing), Felix d’Herelle, Prof. Simon Amiradjibi, George Eliava (all sitting), unknown (standing), Vladimir Antadze (standing))

406

N. Chanishvili and Z. Alavidze

Phage Therapy for Wound Treatment, Surgery, and Dermatology Soon after d’Herelle’s discovery of phages, many researchers around the world successfully isolated phages active against a wide range of bacteria, such as Salmonella Typhi (Alessandrini and Doria 1924), Shigella spp. (da Costa Cruz 1924), Streptococcus spp. (Dutton 1926), Corynebacterium diphtheriae (Fejgin 1925; Stone and Hobby 1934), Escherichia coli, Pseudomonas aeruginosa, Pasteurella multocida, Vibrio cholerae, Yersinia pestis (Krupp and Madhivanan 2015), and Neisseria meningitidis (Fruciano and Bourne 2007). The availability of these phages inspired doctors to develop specific treatments against different bacterial diseases. The phages were administered topically and systemically (orally and parenterally) (Rice 1930; Cipollaro and Scheplar 1933). Phage therapy was used experimentally to treat skin (Cipollaro and Scheplar 1933), eye (Town and Frisbee 1932), and bone infections caused by Staphylococcus (Albee 1933). Many doctors and researchers tried to use phages to treat suppurative wound infections using phage-soaked dressings. Over multiple articles, these doctors reported generally successful results (McKinley 1923, Rice 1930; Albee 1933); Albee and Patterson (1930) describe positive results using bacteriophage therapy to treat osteomyelitis, when after removal of infected bones and tissues they applied a plaster soaked with bacteriophages, which was changed several times during 7–9 weeks of treatment. Phage therapy of wounds was attempted during the Finnish Campaign in 1939– 1940. Early reviews of this work were published by Kokin (1941, 1946) and describe the use of mixtures of bacteriophages produced by the Eliava Institute of Bacteriophage, Microbiology and Virology, active against anaerobes, Staphylococcus and Streptococcus, for the treatment of gas gangrene. Phage mixtures were applied to 767 infected soldiers with a lethal outcome observed in 18.8% of cases compared to 42.2% in the control group. Using the same mixture of phages, other researchers reported a 19.2% lethal outcome in a group of soldiers compared to 54.2% of the control group (Lvov and Pasternak 1947 – cited from Krestovnikova 1947). Furthermore, this same phage mixture was used as an emergency treatment for wounds to prevent gas gangrene for soldiers in the Red Army both during and after World War II. Krestovnikova (1947) summarizes the observations of three mobile sanitary brigades carried out over periods of 2–6 weeks after evacuation to front-line hospitals. One brigade treated 2500 soldiers with phages. According to the reports, only 35 soldiers (1.4%) in this group showed symptoms of gas gangrene, while in the control group of 7918 wounded soldiers, 342 (4.3%) became infected. The second brigade applied phages to 941 soldiers, of which only 14 (1.4%) contracted gas gangrene, in contrast to 6.8% of soldiers in the control group. The third brigade treated 2584 soldiers, of these 0.7% developed symptoms of gas gangrene compared to 2.3% in the control group. Overall, the data described by these three independent brigades showed an average 30% decrease in the incidence of gas gangrene due to the use of a mixture of bacteriophages (Kokin 1941; Krestovnikova (1947). One of the pioneers in the application of phages in surgery was Alexander Petrovich Tsulukidze, a professor of medicine who began using such preparations

Early Therapeutic and Prophylactic Uses of Bacteriophages

407

Fig. 2 Historical glass vials of bacteriophage suspensions as stored on the premises of the Eliava Institute, Tbilisi, Georgia

in 1931 for the treatment of various diseases. According to Tsulukidze (1940, 1941), prior to initiation of treatment, the wounds were analyzed bacteriologically. Besides that, blood analyses were carried out before phage therapy and during surgical manipulations (bandages, puncture, etc.). The condition of the wound was thoroughly described and temperature, pulse, breathing rate, etc. were all recorded. These examinations were also performed after each phage application. Initially the phage therapy was used only for the most severe cases where a lethal outcome was expected. Later, a wider group of patients was involved in the study. Since in the majority of cases bacteriological analysis indicated the presence of mixed bacterial infections (Tsulukidze 1940, 1941), Pyo-bacteriophage (a cocktail of phages used for the treatment of common pyogenic infections, active against S. aureus, Streptococcus, E. coli, P. aeruginosa and Proteus spp.) or a mixture of streptococcal and staphylococcal bacteriophages were applied for treatment (Fig. 2). The phage was administered topically or directly to the accessible part of the wound. Subcutaneous injection of phages was performed 3–4 times every second day to avoid the development of anti-phage antibodies. Additionally, phages were sprayed onto the top of wounds each time bandages were changed. All patients with injuries of soft tissues (38.3%) underwent “ordinary therapy,” which in this instance implied treatment with chloramines, rivanolum, and Vishnevsky ointment (all commonly used antiseptics at the time). These patients

408

N. Chanishvili and Z. Alavidze

often had major tissue damage with penetrating or perforating wounds. The wounds were characterized by the accumulation of pus, infections, and surrounding inflammation, sometimes with necrotic foci, etc. A number of cases with an abscess/ phlegmon around a bullet or mine fragment wound underwent surgical cuts performed during first aid in a field hospital prior to the start of phage therapy. After purification of wounds with iodine solution and/or alcohol, followed by washing with 2% saline solution, the phages were sprayed on the top of the wounds. Simultaneously, 5–10 ml of phage (titer unknown) was injected remotely from the wound into the stomach wall, shoulder, or hip. The wound was bandaged with gauze soaked in phage. According to reports from that time, no cases treated with this method required additional cuts or any other surgery (Tsulukidze 1940, 1941). After a couple of phage applications, the body temperature usually normalized. To achieve a complete cure, only three to four procedures were required. Since the recovery from traumatic injuries and numerous lesions required an extended period, the wounds were stitched 6–8 days after phage treatment, so that further infection was unlikely. In general, phage therapy showed improvement within a number of days whereas standard therapy required several weeks (Tsulukidze 1940, 1941). Phage therapy was effective also for patients with bone injuries or open fractures. The wounds were purified with disinfecting solutions and washed with 2% saline before being sprayed with phages. Simultaneously, phages were injected intramuscularly or subcutaneously at a site remote from the wound. Following this, a plaster cast was applied. The treatment with phages resulted in a faster reduction of pain, improvement in patients’ general condition, and healing of wounds beginning after 2 or 3 days (Tsulukidze 1940, 1941, 1942a, 1942b; Arshba 1942; Makashvili et al. 1942; Sirbiladze 1942; Pokrovskaya 1942). It was noted also that by using phage therapy prior to plastering, it became possible to avoid moistening the plaster, which usually necessitated its replacement and could lead to the development of secondary infections. The plaster could remain unchanged for up to 60 days (Tsulukidze 1940, 1941, 1942a, 1942b). Importantly, in cases of severe hip, shin, forearm, and shoulder injuries, which normally require amputation, application of phage therapy avoided amputation if wounds were left for 10–30 days in blind cast plasters. Very often, phages against the main causes of wound infections were isolated from patients that had never received phage therapy before (Tsulukidze 1942a, 1942b). Phage-treated wounds were examined several times before and during the application of phage therapy. These examinations showed resulting complete sterilization of wounds and in cases of clinical improvement a loss of virulence by the pathogen (Tsulukidze 1940, 1941, 1942a, 1942b). Sirbiladze (1942) described morphological changes of clostridial colonies plated on agar. Before phage therapy the colonies of Clostridium perfringens formed entire, round colonies, while the isolates obtained from the same wounds after the treatment formed colonies with ragged edges. Besides that, animals infected with these strains of C. perfringens survived, implying that their toxicity and pathogenic properties were significantly weakened (Tsulukidze 1942b, Arshba 1942; Makashvili et al. 1942). Successful wound treatment with bacteriophages during the Finnish Campaign and World War II was also reported (Pokrovskaya 1942; Fedorovich 1944; Sutin 1947; Fisher 1949).

Early Therapeutic and Prophylactic Uses of Bacteriophages

409

Phage therapy was applied in different fields of medicine, such as stomatology (Ruchko and Tretyak 1936), ophthalmology (Rodigina 1938), urology (Tsulukidze 1957), and gynecology (Purtseladze 1941), among others. The results of phage therapy in dermatology are especially important. The successful treatment of deep forms of dermatitis caused by S. aureus with specific bacteriophages has been described in a number of articles (Beridze 1938; Vartapetov 1941, 1947, 1957; Izashvili 1940, 1957; Khuskivadze 1954; Gvazava 1957; Shvelidze (1970); Vartapetov et al. 1974). The oldest study in the field was carried out by Beridze (1938). The author described 143 cases of purulent skin infection caused by S. aureus divided into 2 major groups exhibiting either deep or superficial forms of the disease. The group with deep infections included 90 patients with furunculosis (multiple boils) (73 cases), abscesses (10), and hidradenitis (7). A group of 53 patients with superficial skin infections included cases of impetigo vulgaris (29), impetigo contagiosa (13), and various other diagnoses (11). The patients’ ages ranged between 1 and 60 years, with the majority being workers between 20 and 35 years of age. The duration of the illness in patients with acute forms varied from 1 to 7 days, whereas the chronic forms lasted from several weeks to 3 years. The methodology of treatment was described in detail. Initially the area around the infected site was cleaned with a disinfectant solution, moving from the periphery to the center. The pus was then released from the infected area in order to decrease the bacterial load and allow access for the bacteriophages. Simultaneously, swabs were taken to isolate the infecting bacteria and assess their susceptibility to phages. An initial dose (0.5 ml) of bacteriophage preparation was injected directly into the wound and surrounding healthy tissue. The wound was then covered with phage and bandaged. If, on the following day, there was no evidence of irritation, swelling, or any allergic reaction, treatment with phage was continued, with the dose of injected phage gradually increased up to 1 ml on the second day, 2 ml on the third day, and so on. The doctors decided whether to increase the dose each day based on the appearance of the wound and skin reaction. Altogether, 4–5 phage injections were given. After this course, the patients were switched to so-called indifferent therapy which included zinc salve or similar medication. After 3 days the patients were examined again. If the infection persisted, phage therapy was continued in the form of applications onto the previously disinfected wound. If no effect was observed, the patients underwent phototherapy. The patients underwent repeated medical examination after 3 weeks, 3 months, and 10 months. In this group of 143 patients, 108 (75.5%) were successfully treated and an improvement was seen in 11 cases (7.7%). No effect was observed in 18 cases (12.6%), and the fate of 7 patients (4.9%) was unknown (Beridze 1938). In his paper, Vartapetov (1957) listed numerous authors studying the clinical efficiency of bacteriophages (staphylococcal and streptococcal phages and Pyo-bacteriophage) for the treatment of furuncles, carbuncles, hidradenitis, abscesses, and so on. Vartapetov (1957) summarized the data of over 6000 patients involved in phage therapy studies, demonstrating that in all cases healing occurred within 4–8 days. Clinical success rates ranged from 70% to 100%. In general, the

410

N. Chanishvili and Z. Alavidze

best results were seen in cases of treatment of abscesses and sycosis caused by staphylococcal infection. An interesting study was performed by Shvelidze (1970) on 161 patients with chronic and frequently relapsing infections. Sixty-two patients were diagnosed with furuncles (boils) and furunculosis, 54 with carbuncles, and 45 with hidradenitis. Despite antibiotic treatment, some patients suffered with chronic infections, in some cases for as long as 20 years. The patients complained of fevers, headaches, weakness, insomnia, and/or movement difficulties. Bacteriological analyses showed that in 82.7% cases the infection was caused by coagulase-positive penicillinresistant Staphylococcus aureus. Phages were administered topically via intradermal injections performed every second day. The phage was administered to patients in increasing doses ranging from 0.1 to 0.5 mL. In total, 7–10 injections were given around the infected site. It is important to underline that the results of phage therapy were compared with the previously performed antibiotic treatment, which was considered as a control. Successful results were achieved in 94.4% (152 cases), 4.3% (7 patients) showed a significant improvement, and only in 1.3% of cases (2 patients) was no improvement observed. The patients were monitored for a 4-year period following treatment. Relapse was observed in 8.5% of cases; however the severity of these subsequent infections was of relatively minor concern, and an additional course of phage therapy resulted in complete cure (Shvelidze 1970). Many authors (Beridze 1938; Gvazava 1957; Vartapetov 1957; Shvelidze 1970; Vartapetov et al. 1974) drew attention to the immunostimulation potential of staphylococcal bacteriophages. Based on their observations, the authors (Beridze 1938; Gvazava 1957; Vartapetov 1957) suggested the following: 1. Phages kill (lyse) the appropriate host bacteria (direct effect of phage therapy). 2. The killed bacteria are present in the bloodstream as pieces of bacterial cell wall and debris (i.e., antigens) which stimulate the immune system. These antigens might trigger the immune system in a manner that could be more diverse and better presented than recombinant vaccines and better folded than other forms of inactivated vaccines, such as heat-killed pathogens (indirect action of phage therapy).

Phage Therapy for Treatment of Enteric Infections In the 1920s and 1940s, intestinal infections caused by Salmonella and Shigella species were a huge problem all over the world (Beckerich and Haudoroy 1922; Alessandrini and Doria 1924, da Costa Cruz 1924; Rolleston 1926; Compton 1929; Karamov 1938; Karpov 1946). Karpov (1946) provided epidemiological data on mortality rates at different times and at various geographic locations. Mortality rates in cases of typhoid fever varied between 7 and 10%. In Baku (Azerbaijan) in 1932, the mortality rate was 5.8%. Similarly, in one of the main hospitals in Leningrad (now St. Petersburg, Russia), the mortality rate in 1931 also attained 5.8%. During an outbreak in Rostov (Russia), in 1926, mortality rates reached 8.2% (Karpov

Early Therapeutic and Prophylactic Uses of Bacteriophages

411

1946). In autumn of 1926, an outbreak of typhoid fever started in Hanover, Germany, where 4220 cases were reported, among which 320 (7.6%) lethal outcomes were recorded. A water supply system was recognized to be a main source of this outbreak (Rolleston 1926; Speigel Online 2011). These figures indicated the urgent need to introduce novel therapeutic means to combat these infections. The first trial to treat human intestinal infections was performed by d’Herelle in 1919 and included only four patients suffering with dysentery. The phage treatment was administered as a single dose, and the symptoms of dysentery disappeared by the next day (d’Herelle 1917). After d’Herelle’s experiment, many scientists started to use phage therapy to treat dysentery, however with varying success (Davison 1922; Spence and McKinley 1924; Compton 1929; Riding 1930; Querangal des Essarts 1933; Kessel and Rose 1933; Haler 1938; Johnston et al. 1933; Karamov 1938; Karpov 1946). For example, Davison (1922) described 12 cases of bacillary dysentery caused by Shigella flexneri, among which 7 patients received phages orally, while 5 were treated by enema. Only 5 out of 12 patients (42%) overcame the infection. Failure in the other cases was explained by the fact that the therapy began too late when the disease was already in its peak (19). According to da Costa Cruz (1924), phage therapy was the best treatment for bacillary dysentery, with the positive results achieved in 24–48 h. These statements, however, were not supported with reliable statistics. One of the most detailed early reports belongs to Compton (1929), who described the treatment of dysentery in the city of Alexandria, Egypt, performed in 1927–1928 with the cooperation of the doctors willing to evaluate the effects of phage therapy in the treatment of dysentery. The polyvalent phage lysate included phages active against S. shiga, S. flexneri, S. hiss, S. sonnei, and S. gay. Only those patients whose diagnosis was confirmed etiologically were included in the study. Each patient received three ampoules containing 2 mL of phage lysate. The patients were provided with instructions on the use of the phage and a questionnaire. In 1927, almost 50 patients were treated and about 150 in 1928. Among 200 patients, 92 did not complete the treatment and returned the ampoules. Of these, only 66 were intact and applicable for subsequent use. Therefore, the authors assumed that 108 cases successfully completed the cure since they did not return for additional visits to their doctors. Thus, coming out of this assumption, Compton (1929) concluded that the cure rate was 108/200 (54%). The remaining ampoules Compton used for treatment of 66 patients diagnosed with dysentery. For this experiment, the author developed a semi-qualitative method of evaluating the recovery of the patients, scoring the success results as “very good” (35 cases), “good” (10 cases), “moderately good” (6 cases), “partial failure” (5 cases), and “failure” (10 cases). Four out of ten negative cases Compton removed from the study, since their treatment started too late when the patients were already severely sick. Thus, summarizing very good and good results, the total success rate achieved was indicated as 72.6% (45/62) (Compton 1929). Due to Compton’s (1929) study, it became possible to analyze in detail the reasons for phage therapy failures. Analysis of the results revealed that the age of the patient, the duration of illness prior to phage treatment, and resident bacteria other than the targeted species were important factors influencing the outcome of the phage treatment. In particular, it was observed that phage treatment was the least

412

N. Chanishvili and Z. Alavidze

successful with children under 1 year old, with the success rate increasing in direct proportion to age. Early start of phage therapy treatment, however, was critical to a successful outcome. It was shown that if the patient had been ill for 3 or fewer days prior to treatment, then the success rate could be as high as 90%. The longer the duration of the period between the onset of illness and start of phage therapy, the lower was the success rate. Later studies conducted by Riding (1930) and others (Querangal des Essarts 1933; Kessel and Rose 1933; Johnston et al. 1933; Haler 1938; Murray 1938; Goodridge 2013) appeared to be less informative, since no information was reported regarding the doses, duration, and the number of times the phage preparation was administered, control groups (Johnston et al. 1933; Murray 1938). The failures of phage therapy sometimes were caused by the fact that no preliminary in vitro susceptibility tests were done prior to clinical trials (Murray 1938). Collectively, these reports reveal much diversity in results and conclusions. Comparisons of the studies are impossible due to lack of information regarding concentration of phage, numbers of different phages employed, method of preparation, method of administration, and the fact that in many of the reports no controls were included. The largest clinical study of therapeutic anti-dysenteric phages was reported by Sapir (1939) who describes a total of 1064 cases of dysentery treated with bacteriophages in two different Moscow clinics. The patient group included 767 men and 297 women ranging in age from newborns to 79 years old. Dysentery was diagnosed using bacteriological tests (362 patients), clinical observations (512 cases), and clinical colitis (190). Bacteriological analysis of 362 patients indicated that dysentery in 289 cases was caused by Shigella dysenteriae type 1 (22 resulting in a lethal outcome), 69 cases of S. flexneri, and 4 cases of S. dysenteriae type 2. A standard phage therapy, using the dysenteric bacteriophage preparation that was developed by the Mechnikov Institute in Moscow, was applied to every age group as described below. The exact content of this preparation has not been specified. A daily dose of phage for an adult was 20 ml and 10 ml for a child (titer 109 to 11 10 using the Appelmans method of serial dilutions in bacteriology media which does not allow to quantify phages in the stock solution) (Appelmans 1921; Chanishvili 2012a; Rohde et al. 2018). The dose was divided into two portions and given to patients at midnight on the day of arrival and at 4 am to minimize the inactivation of phage by any meal residues. The comparable time of phage administration to all patients facilitated the evaluation of the results of the phage therapy. The patients were given a magnesium-soda solution (magnesium 10 g/1 L + sodium bicarbonate 20 g/L), initially 6 h prior to phage therapy and then every 2 h for the following 12 h. Adults were given 100 ml of the solution per dose, and children were given 10–50 ml depending on their age. The solution was given with the aim of providing optimal conditions for phage propagation and also to help clear the intestines. The patients were kept on a strict diet during the first 48 h. The author concluded that the application of this dose of phage divided into two portions and administered over the course of a single day was sufficient and did not need to be repeated (Sapir 1939).

Early Therapeutic and Prophylactic Uses of Bacteriophages

413

According to Sapir (1939) the application of phage therapy significantly decreased the duration of hospital stays (11–20 days), in contrast with symptomatic treatment (43 days) or even with specific (serological) treatment (22 days). The paper highlighted that early use of phage therapy reduced hospitalization time and that usually after 1–2 days of phage therapy a dramatic improvement in the patient’s condition was observed, as evidenced by less frequent and less watery stools with less blood and/or mucus. Sapir (1939) reported also that after one day of phage treatment, the number of patients with bloody stools decreased from 100 to 74, and on the fifth day of treatment, only 4 patients remained with this symptom. A single week of phage therapy resulted in reduction of symptoms such that 95% of patients could be released from hospital. Nevertheless, a lethal outcome was observed in 47 cases (4.4%), although it was noted that these patients suffered with dysenteric pancolitis and other severe degenerative changes of the parenchyma of various organs, as well as colonic ulcers, etc., which are typical for long-term infection, as verified by the postmortem studies. The author concluded that phage preparations should be given to every patient showing symptoms of dysentery, independently of whether the patient is arriving at hospital, being seen by ambulance or asking for medical help at home. This measure would have not only therapeutic but prophylactic effect as well (Sapir 1939). Lipkin and Nikolskaya (1940) performed phage therapy on 100 patients suffering from dysentery. A control group of 50 patients received ordinary medication, such as purgative salts, which were used in most cases. In 21 cases the patients underwent serum therapy. In five severe cases, a combined phage and serum therapy was used. All patients were maintained under the same conditions, in terms of care, diet, etc. Phages produced by the Tbilisi Institute of Vaccine and Sera and the Kuibishev Institute of Epidemiology and Microbiology were used in these studies (An old name of the Eliava Institute of Bacteriophage, Microbiology & Virology). The titers of these phage preparations were 109 to 1011 by the Appelmans method (Appelmans 1921; Rohde et al. 2018). Five ml of phage was given to patients orally together with 2% soda solution 3 times per day. After receiving the phage, the patients fasted all day. In almost every case the phage treatment was performed for one day. Only in 6 cases was the phage treatment at the same dose performed over 2 days. The majority of patients (66%) received the phage within the first 5 days of the start of infection. The development of the disease was evaluated through observations of stool frequency, presence of mucus, blood, cramps, etc. Lipkin and Nikolskaya (1940) reported a significant effect of phage therapy even in cases where the treatment was started rather late. 25% of patients (n ¼ 100) stopped reporting painful symptoms by the second day of treatment. 79% did not show pathological symptoms by the fourth day, and 100% did not by the sixth day, after which stool had normalized. These data are in contrast to the results obtained with standard therapy, where only 2% (1 case out of 50 patients) showed an improvement on the second day of treatment, 14% on the fourth, and 46% on the sixth day. It is noteworthy that patients with more mild symptoms were included in the control group, while those with relatively severe illnesses were included into the experimental group. According to the authors, the fact that these patients showed an improvement as

414

N. Chanishvili and Z. Alavidze

soon as they got phage treatment illustrated the effectiveness of this method. Relief of symptoms in patients treated with serum therapy was recorded in 33% of cases (7/21) on the fourth day of treatment and in 67% on the sixth day, indicating that serum therapy resulted in the slower relief of symptoms than phage therapy. Five patients subjected to serum therapy remained sick over 10 days. These patients later underwent successful phage therapy (without further serum therapy) as well. Phage preparations were generally considered to be particularly efficient for the treatment of intestinal infections (Podvarko 1964). Vlasov and Artemenko (1946) described the results of treating 30 chronic dysentery patients with phages. Many of the patients were exhausted by infection and were bedridden. A dry tablet preparation known as “phage-vaccine” – a combined preparation comprising, after reconstitution in saline, 106 killed cells/ml and 107 (by Appelmans 1921) – was used. The patients had suffered with infections for 1–2 years, and in 70% of cases rectoscopic investigation indicated the presence of bleeding ulcers. Prior to combined phage-vaccine therapy, all of the patients underwent multiple courses (1– 8 times) of therapy with antibiotics and sulfonamide preparations. After the phagevaccine therapy, the authors reported curing 26 patients (86.7%) within 10–20 days. Assessment of the results was based on improvements in the general condition of patients including normalization of stools and recovery of the mucous layer of sigmoid colon and rectum (Vlasov and Artemenko 1946).

Prophylactic Use of Phages Phages have been used extensively in the former Soviet Union for prophylaxis in regions with a high incidence of infections and also in communities where rapid spread of infections might occur, such as kindergartens, schools, military accommodation, etc. (d’Herelle 1935, Belikova 1941, Blankov 1941, Blankov and Zherebtsov 1941, Kagan et al. 1964, Florova and Cherkass 1965, Agafonov et al. 1984, Anpilov and Prokudin 1984, Chanishvili 2012a, 2012b). The application of phages for prophylaxis was carried out in 1929–1930 against the diseases that were the most severe and important at that time, such as dysentery, typhoid fever, staphylococcal infections, etc. The first mass application of dysenteric bacteriophages in the USSR was performed in Alchevsk (Donbas region) in Ukraine in 1930 (Ruchevski, “Vrachebnaya gazeta,” 1931, 21: 1586, cited by Krestovnikova 1947). An experiment on the prophylactic use of phages was later successfully carried out in 1935 on thousands of people in regions with a high incidence of dysentery (Melnik et al. 1935, Belikova 1941, Blankov 1941, Blankov and Zherebtsov 1941, Vlasov and Artemenko 1946). The results were reported at scientific conferences in 1934 and 1936 in Kharkov and in 1939 in Moscow, after which the dysenterial phage preparation was finally approved as a preventive measure for mass application (Krestovnikova 1947). According to Krestovnikova (1947), it was recommended that repeated seasonal prophylactic “phaging” be carried out in areas where dysentery was endemic. Later modifications included

Early Therapeutic and Prophylactic Uses of Bacteriophages

415

formulating the dysenterial phages as dry tablets, which also began to be included in clinical studies. One of the most dramatic examples of prophylactic use of cholera bacteriophages is related to the Stalingrad battle and an outbreak of cholera with many lethal outcomes in 1942 and 1943. The prominent Soviet bacteriologist Yermolieva, with her staff, organized bacteriophage production in the city. Doses of cholera phage were given to 50 thousand people every day during 5 consequent days (bread was given only after “phaging”). The outbreak subsequently ceased (Yermolieva 1939, 1942; Yermolieva and Yakobson 1943, 1949). During the monitoring period, lasting 3 months, the presence of Vibrio cholerae in the feces of convalescents was not registered (Yermolieva and Yakobson 1949). Babalova et al. (1968) describe the results of prophylactic measures carried out in 1963–1964 using phage tablets with an acid-resistant coating. Over 30,000 children from the age of 6 months to 7 years were involved in the study. Prophylactic phage treatment was carried out on children living on one side of the street, while those living on the other side did not get the phage treatment and thus were considered a control group. The phage tablets were administered either before or 2 h after meals. Children from 6 months to 5 years received 1 tablet (equal to 20 ml of phage). The titers of each component of this polyvalent dysenteric preparation was equal to 104–105 by Appelmans (1921), while children over 5 years received 2 tablets. The effect of phage prophylaxis was evaluated on the basis of clinical symptoms and in some cases also based on bacteriological analysis. The incidence of acute dysentery in the control group was 3.8 times higher than in the experimental group (Babalova et al. 1968). Later studies on mass prophylaxis of intestinal diseases by the application of phages were performed in Red Army units by military doctors. For prevention of dysentery and typhoid, epidemic strain-specific phages were also used, with two tablets administered once every 5–7 days during the outbreak season. The authors reported about six- to eightfold decrease of incidences of intestinal infections in the test groups in comparison with the controls (Florova and Cherkass 1965; Agafonov et al. 1984; Anpilov and Prokudin 1984; Kurochka et al. 1987; Chanishvili 2012a, 2012b). Sayamov (1963) reports the results of therapeutic and prophylactic trials using an anti-cholera bacteriophage preparation, performed by Soviet doctors in East Pakistan in 1958 and in Afghanistan in 1960. During a cholera outbreak in Dacca (East Pakistan) in 1958, only 22 patients with severe conditions underwent phage therapy treatment. Each patient was given a single intravenous dose of phage suspension (5– 10 ml) prepared on saline solution and simultaneously an oral dose of phage suspension (30 ml), for three consecutive days (phage titer is unknown). Only two lethal outcomes were registered, in contrast with the fatality rate at the Kandahar hospital, which appeared to be about 50%. At the same time, the Soviet doctors used phage therapy prophylactically in East Pakistan. It was reported that phage prophylaxis was performed on a large group of around 30,000 people. These measures successfully ceased spread of cholera epidemics, since no cases were reported in these areas (Sayamov 1963).

416

N. Chanishvili and Z. Alavidze

Phage prophylaxis was implemented in Afghanistan in 1960, which included patients admitted to the hospital as well as healthy hospital staff (total >1600 persons). The doses and duration of the treatment were the same as described above. Only 4 lethal outcomes out of 119 cholera patients (3.5%) were registered as a result of these measures. No complications were observed as a result of phage therapy. Similar prophylactic measures were reported also in Kataghan Province in the north of Afghanistan in 1960 (Plankina et al. 1961; Sayamov 1963). In this case the majority of the patients were treated at home (about 90%) because of the difficulties associated with transportation of the patients to the hospitals and limited number of bed places. Like in other cases, the patients received a dose of 20–30 ml (or, in particularly severe cases, as much as 50 ml) of phage suspension, accomplished with one or two intramuscular injections (intravenous administration was not used). In addition to phage therapy, a single dose of cholera vaccine was used for preventive purposes. This complex treatment was applied to a healthy part of each village’s population. Sayamov (1963) and Plankina et al. (1961) underlined the role of phage prophylaxis in prevention of new cholera cases in rural populations as the best success was observed only in those villages where the whole population was subjected to the phage treatment. Thus, altogether approximately 270,000 persons in all regions underwent phage treatment. The information about the control groups is missing in this publication which makes it difficult to evaluate the contribution of phages and of the vaccine in the observed prophylactic effect. However, Sayamov tended to conclude that the effect of anti-cholera prophylactic largely was dependent on bacteriophage treatment. He suggested that phage prophylaxis may be used as an important supplement to the standard measures for cholera control (Plankina et al. 1961; Sayamov 1963).

Intravenous Staphylococcal Bacteriophage: The Highest Achievement of the Georgian Scientists The story of phage therapy would be incomplete if we did not mention the development of intravenous anti-staphylococcal phage and its use in humans. Attempts to produce intravenous bacteriophage preparations were started by d’Herelle using crude yeast extract medium phage lysates (Antadze 1957). These experiments continued into the early 1930s in the Soviet Union, and the earliest study dedicated to this to be found in the library of the Eliava Institute was published by Ebert and Shapiro (1938). In this article the authors briefly describe previous animal trials on rats and rabbits, which were performed by intravenous infection of the animals followed by a single intravenous administration of phage. No protective effect against infection was apparent, and the animals died. This was in contrast to the positive impact of topical bacteriophage therapy, as observed by other doctors. Therefore, the experiments of intravenous administration of phages continued. Skvirskyi et al. (1938) used combined intravenous and intramuscular administration of bacteriophage to treat typhoid fever. These authors were among the first to administer phage therapy intravenously (1–2 ml) against typhoid fever. Skvrskyi

Early Therapeutic and Prophylactic Uses of Bacteriophages

417

et al. (1938) observed a lowering of temperature but did not attribute this to the use of phage. In 1930 Ruchko and Melnik (cited by Krestovnikova 1947) reported experiments performed on 69 patients suffering with typhoid fever. Like previous authors, they also used an intravenous mode of administration and also observed a lowering of temperature and a shortening of the duration of illness. However, many authors reported a rise in temperature of 1–2
C prior to lowering (Krestovnikova 1947). During World War II, Yermolieva, in severe cases of cholera, used bacteriophages intravenously. 5 ml of phage lysate diluted in 2 L of saline solution was slowly introduced into the vein twice, with a two-day interval between treatments. The treatment continued orally using the following regimen: 30 ml of cholera phage divided into two doses consisting of 15 ml of phage diluted in 20 ml of boiled (sterile) water administered 3 h apart. The treatment continued for 3 days (Yermolieva 1939; Yermolieva 1942; Yermolieva and Yakobson 1943, 1949). These measures decreased the number of cholera incidences and helped to stop the spread of the epidemic among the civilian population and army units as well. Positive results of intravenous administration of phage therapy among patients with acute septicemia caused by anaerobic infections were described by Arsentieva (1941). The phage was administered as transfusions in doses of 50–100 ml with intervals of 2–3 days. The author refers to shivering and slight rise of temperature which was treated with caffeine and warming of the patient (Arsentieva 1941). Kokin (1941, 1946) used intravenous administration of phages in cases of staphylococcal infections. 30 ml of specific anti-staphylococcal bacteriophage diluted in 300 ml of saline solution was used for transfusions. An intensive reaction following intravenous administration was explained by release of bacterial toxins and activation of phagocytosis. In cases of moderate body reaction, the dose of phage was increased up to 50–60 ml in 300 ml of saline solution (Kokin 1941, 1946). Yukelis (1946) reported a 72% success rate of phage therapy in cases of furunculosis and deep ulcerous pyodermitis using intravenous administration of staphylococcal bacteriophages. The phage was prepared on protein-free bacteriology media, and 10 ml was administered every second day. The author did not observe any side effects and recommended performing intravenous transfusions of 40–50 ml every day (Yukelis 1946). A very interesting study was performed by Manolov et al. (1948) who reported on the intravenous application of bacteriophages for treatment of typhoid fever. The low efficacy of the therapeutic effect after oral administration of bacteriophages led the authors to conclude that intravenous administration could be used. For treatment of typhoid fever, the authors applied 20–25 ml of bacteriophage prepared in saline solution containing minimal amounts of organic contaminants. Safety of this bacteriophage had been proven in experiments on rabbits and white mice. It was shown that intravenous administration of the typhoid bacteriophage in animal models protected the mice from the development of infection when challenged by a dose of the typhoid culture (Manolov et al. 1948). 15–20 min after administration of the intravenous phage, patients complained of shivering. After 2–3 h a rise in temperature was observed which in some cases was followed by a feeling of nausea and often vomiting. 12–14 h after the phage injection, the temperature was normalized.

418

N. Chanishvili and Z. Alavidze

After 24 h, however, it rose again to the same level as seen after 2–3 h. As a result of this, the authors applied several phage injections every day or every other day. Unfortunately, the number of patients is not indicated in the study, but the results led to the following conclusions: 1. Oral administration of bacteriophage was regarded as unsuitable due to the specificity of pathogenesis of typhoid fever. And low doses (2–10 ml) of typhoid bacteriophages administered orally were inefficient. 2. Intravenous use of 20–25 ml of typhoid bacteriophage, prepared in saline solution, every day over 3 days led to a decrease of temperature and shortening of the fever period, improvement of the general condition, and complete cure (Manolov et al. 1948). Although phages were administered intravenously in early studies during the 1930s and 1940s, this type of therapy was rejected due to the unpleasant side effects, including a rise in temperature up to 39
C, shivering, headaches, etc. The negative reactions usually occurred for 40–60 min following treatment. However, no lethal outcomes were reported (Skvirskyi et al. 1938; Yukelis 1946; Kokin 1941; Kokin 1946; Krestovnikova 1947; Manolov et al. 1948; Antadze 1957). The Eliava Institute of Bacteriophage has had particular success in the elaboration of intravenous bacteriophage preparations for the treatment of S. aureus septicemia. Antadze (1957) wrote that the organic residuals emerging during manufacturing of the phage preparations create difficulties for their parenteral use, since they might cause a number of side effects due to their sensitizing impact on the body’s immune system. Taking into consideration the fact that during the manufacturing process (i.e., at the stage of the growth of bacteriophages in the bacteriology media), the host bacteria are unable to completely utilize the substrate, researchers from the Eliava Institute decided to exclude the unessential protein fragments of the media from the phage propagation process. They determined the essential portion of bacteriology media which would be necessary for multiplication of the host bacteria and amplification of the phage particles to allow them to reach sufficient titer. This goal was achieved, and the Eliava staff succeeded in propagating the bacteriophages on media diluted with a saline solution so that they would reach titers of 108–1010 by Appelmans (1921). Successful preparation of anti-pyrogenic staphylococcal bacteriophage for parenteral use was achieved in the 1970s. These studies, including manufacturing of the intravenous staphylococcal bacteriophage preparation and its use in animal and clinical trials, were performed under the leadership of Professor Teimuraz Chanishvili. Research to elaborate a protein-free apyrogenic media suitable for the manufacture of intravenous staphylococcal bacteriophages was carried out simultaneously at the Tbilisi Institute of Vaccine and Sera (Chirakadze and Chanishvili 1964; Chanishvili et al. 1974; Nadiradze 1983) and the Gorky Institute of Epidemiology and Microbiology (Anikina 1982). Anikina (1982) tried to reproduce staphylococcal septicemia in an experimental animal mouse model and carried out a total of 30 experiments on 750 animals.

Early Therapeutic and Prophylactic Uses of Bacteriophages

419

Further studies were performed in rabbit models using a generalized staphylococcal infection in rabbits with three S. aureus strains. Each strain induced infectious processes of different severity and duration. The impact of the intravenous bacteriophage was studied in acute and chronic septic models. The rabbits were given phages intravenously at 5 ml/kg, and injections were repeated three times over one day. The treatment did not prevent lethal outcomes but that occurred on the 18th day, much later than in the control group, and kidney abscesses were not seen. The animals in the control group died between the fourth and tenth days. The next series of experiments was performed using a chronic septicemia rabbit model using two phage preparations: 1. An apyrogenic staphylococcal phage preparation made by the Tbilisi Institute of Vaccine and Sera, prepared on synthetic media with the addition of yeast extract 2. An experimental staphylococcal phage preparation made by the Gorky Institute of Microbiology and Epidemiology, manufactured on media containing aminochlorine hydrolysate Altogether, 35 animals were divided into 2 groups of 10, with 15 left as a control. The animals were infected with the strain S. aureus # 79 (4  106 cfu/kg). A day after infection the rabbits were given 3 ml/kg of the different phage preparations for 5 days. The animals from both experimental groups were cured within 7 days (Anikina 1982). Both preparations appeared to be almost equally effective. In the next experiment, rabbits received combined phage and antibiotic (penicillin) therapies. The animals were infected with the penicillin-resistant strain, S. aureus # 36 (2  106 cfu/kg), which caused prolonged generalized infection. The phage was given intravenously 3–5 ml/kg. Treatment started on the second day after the infection and continued for 5 days. Penicillin was administered intramuscularly, first 4 h after infection (200,000 units per 1 kg of weight) and continued for 5 days. The animals were observed for 14 days, during which bacteriological analyses were performed on the first, second, third, fourth, fifth seventh, and ninth days. After 14 days of observation, all animals from the experimental and control groups were euthanized and their internal organs checked for bacterial loads. Neither of the 5-day treatments performed separately with the phage or antibiotics resulted in eradication of infection. The best results were achieved after the combined treatment with phage and penicillin: Only in one case was S. aureus detected, 3 days after infection. However, dissection of this animal did not show any pathological changes (Anikina 1982). Since high doses of antibiotics may cause various side effects, the author decided to perform the same experiment using low doses of penicillin (25,000 units/kg). 15 rabbits were divided into 3 groups. The first was an untreated control group; the second group of rabbits was treated with injections of 25,000 units of penicillin per kg, with the antibiotic given 4 h after infection and then intramuscular injections continued for 5 days. The third group was given penicillin according to the scheme described above with the addition of the intravenous phage treatment (5 ml/kg) every day for 5 days. The first phage injection was given 24 h after the infection. The

420

N. Chanishvili and Z. Alavidze

animals were observed for 14 days, after which they were euthanized. Blood analysis was performed on days 1, 3, 5, 7, and 10 (Anikina 1982). The control animals demonstrated high temperature, weight loss, and pyuria. Two animals died after 4–5 days. Examination of the surviving animals demonstrated the development of multiple widespread abscesses including on the kidneys. S. aureus was isolated from all organ samples and urine. The 2nd group of rabbits treated with penicillin demonstrated the same symptoms as the control animals, two died on the 9th and 12th days; 5 days after infection all animals from this group started to produce pus in urine. Examination of the animal organs did not show any pathological changes. Bacteriological analysis of the organ samples, however, showed the presence of S. aureus. In the third group of rabbits undergoing combined treatment with penicillin and bacteriophage, all animals survived. A rise in temperature was observed as late as 3 days. After 7 days the blood samples appeared to be sterile. Dissection of the animals did not show any development of abscesses, and the presence of S. aureus was not seen by bacteriological analysis (Anikina 1982). Following animal trials, the phage was tested on 20 human volunteers with various types of acute and chronic staphylococcal infections. No side effects were observed. As a result of this, special permission was granted on April 11, 1979, allowing clinical trials on 250 patients per hospital, simultaneously in several hospitals. Altogether 653 patients were involved in the clinical trials, 355 men and 298 women, with 345 patients in the experimental group and 308 in the control group. 130 patients in the experimental group were treated with the intravenous phage preparation only; the other 215 received a combined treatment with the phage and antibacterial preparations commonly used in the medical practice (antibiotics, etc.). Patients in the control group were only treated with antibiotics. Usually intravenous staphylococcal phage (IVSP) was applied to patients having contraindications against antibiotics such as allergies, pregnancy, multiply resistant forms of staphylococcal infections, etc. The IVSP treatment was often applied after unsuccessful antibiotic therapy. The IVSP was mainly used intravenously (Meladze et al. 1981; Bochorishvili 1984; Chkhetia 1984; Samsygina and Boni 1984; Samsygina 1985). Only in cases of traumatic osteomyelitis was the phage applied intra-arterially (Tavberidze 1993). For intravenous use the IVSP was used in a dose of 0.5–1 ml per 1 kg of weight as transfusions combined with blood replacing compounds (saline solution, etc.). Higher doses (2 ml per 1 kg of weight) were applied rarely, e.g., in the cases of osteomyelitis. During the intravenous phage administration of the IVSP, the doctors did not observe any life-threatening side effects (Meladze et al. 1981; Chkhetia 1984). The results of the clinical studies were published by a group of doctors from the Institute of the Clinical and Experimental Surgery (Tbilisi, Georgia) – one of the organizations implementing clinical studies (Chkhetia 1984; Meladze et al. 1981). They performed a thorough analysis of the immune changes occurring during phage therapy. In total, 340 patients (253 males and 87 females) with unspecific festering diseases of the pleura and lungs were under observation. An experimental group received a complex treatment with the IVSP and antibiotics. A control group was treated with antibiotics only. Stable remission in the experimental group was attained on average in 53.5% of cases in contrast with the control group at

Early Therapeutic and Prophylactic Uses of Bacteriophages

421

22.0%. In the experimental group, sequelae and lethal outcomes were observed in 2% of cases, while in the control, this figure was 4%. The authors indicated that 43 patients received intravenous phage transfusions together with antibiotics and topical treatment with bacteriophages. The dose for intravenous application was 0.5– 1.0 ml per kg of weight. No side effects were observed in the cases of intravenous phage transfusions; only in the cases of use of phages without antibiotics was a mild rise of temperature of 0.3–0.6
C observed. Local irritation of the bronchi epithelium was not noted. The duration of phage therapy was determined according to the X-ray analysis and clinical observations (Meladze et al. 1981; Chkhetia 1984). Interestingly, the microbiological analysis of the lung flora demonstrated dynamic changes in antibiotic susceptibility patterns. At the start of treatment, 64.4% of strains isolated from the experimental group showed resistance to antibiotics, compared to 66.7% in the control group. After implementation of phage therapy, drug resistance in the experimental group decreased to 60%, while in the control group, it increased to 73.8%. Resistance to staphylococcal bacteriophage was determined in 13.4% of cases (Chkhetia 1984). Dr. Nugzar Chkhetia (1984) described the results of treating 152 patients recovering from lung operations. A control group of 107 patients was treated with antibiotics alone, and 45 patients received antibiotics together with phages. Remission and stabilization of the suppuration process in the experimental group was observed in 93% of cases, while in the control group, it was 80%. The frequency of post operational reinfection of the pleural cavity was 23% in the experimental group, compared to 67% in the control group. No lethal outcomes were observed in the experimental group, while a lethal outcome was observed in 8.4% of the group treated with antibiotics alone (Meladze et al. 1981; Chkhetia 1984). Phages were administered through various routes including local administration (tampons, bathing of cavities), inhalation, oral and parenteral, including intramuscular and intravenous injections. Phage preparations were applied as either liquid preparations or aerosols, with doses varying between 10 ml and 150 ml. The author recommended performing post-operational treatment by intra-pleural administration of phages. In this case the phage could be administered either via a drainage tube or by puncture. The dose of the liquid phage to be used could be determined by the remaining volume of pleural cavity, which usually varies between 10 and 300 ml. Prior to phage administration, the pleural cavity should be released of exudates by aspiration and then washed with a sterile saline solution mixed with a painkiller. Topical application of bacteriophage therapy did not cause any side effects. The duration of the topical phage therapy depended on the speed of the healing process. According to the author, however, it should not exceed 20 days (Chkhetia 1984). The IVSP was successfully applied to treatment and prophylaxis of staphylococcal post-traumatic infections of the long bones. The study refers to 45 cases in the experimental group, 6 of which were treated with the IVSP alone and 39 with IVSP in combination with antibiotics, and 50 in the control treated with antibiotics alone (Tavberidze 1993). After the clinical trials, Dr. Levan Tavberidze (1993) from the Institute of Traumatology (Tbilisi, Georgia) continued with the application of IVSP. Altogether 125 patients were treated, 109 of whom had been diagnosed with post-

422

N. Chanishvili and Z. Alavidze

traumatic osteomyelitis, and of these, 62 had unconsolidated fractures. 69 patients received phage therapy (topically or intra-arterially) combined with antibiotics; 59 were treated with antibiotics alone. Prior to phage therapy, all patients had undergone unsuccessful generalized antibiotic and topical disinfecting therapies (Tavberidze 1993). The presence of S. aureus was proved by bacteriological analysis in 76% of cases. 13.1% cases were mixed infections with Proteus sp. In the majority of cases, the isolated strains showed resistance to multiple antibiotics. At the same time, these strains were susceptible to the IVSP (90.4%). The phage treatment was performed topically by introduction of the preparation into the fistula, festering cavities, soft tissues, etc. Prior to application of the phage preparation, the infected site was washed with 2% sodium bicarbonate solution. 69 patients received phage therapy prior to operation. Dynamic changes in the wounds were studied regularly by bacteriological analysis (105 tests). In 90% of cases, the wounds were cleaned of infection, and the festering and inflammation was reduced in comparison with the initial indexes. After the operation, phage therapy continued, but it was combined with antibiotic treatment. A mixture containing 1 g of antibiotic diluted in 20 ml of phage was administered into the wound by irrigation or by injection into the soft tissues localized around the wound. Altogether 155 patients received phage therapy in the post-operational period, among them 124 cases with osteomyelitis of the ankle where the phage administration was performed intra-arterially. Other patients were treated topically. Intra-arterial phage therapy was combined with antibiotics. Phage preparation was administered directly into the hip arteries. The effect of intra-arterial administration was understood as being more efficient than other methods. The clinical effect was obvious after 4–5 injections and was shown as a decrease of pain, normalization of temperature, reduction of inflammation, cleaning of the wound, improvement of the blood test parameters, etc. This effect was explained by concentration of high doses of the preparations in the soft tissues, which could not be achieved by other methods. Side effects were observed in two cases as a rise in temperature up to 39
C, which was easily released with painkillers. The most positive outcome was shown in the group receiving combined phage and antibiotic therapies (Bochorishvili 1984). Along with the improvement of clinical symptoms, a normalization of blood analysis, such as SOE (sedimentation of erythrocytes), and immune indices (phagocytic activity of leukocytes) occurred due to the phage therapy (Samsygina and Boni 1984; Samsygina 1985). Samsygina (1985) concluded that the immunomodulating properties of the IVSP should have been considered as one of the most important properties of this preparation. Interestingly, after application of IVSP, the symptoms of the bacteremia in the majority of cases disappeared within 5 days. Among 149 children the bacteremia was observed in 53 cases. After the phage therapy, bacteremia remained in eight cases and after 7–10 days in only three cases. Meanwhile, in the control group of patients, the bacteremia remained in 50% of cases even after 10 days of antibiotic therapy. Thus, the doctors concluded that the optimal duration of phage therapy in children was 7–10 days (Samsygina and Boni 1984; Samsygina 1985). Forty-four children with mild symptoms of located festering inflaming disease (LFID) received intravenous phage treatment combined with topical therapy, without

Early Therapeutic and Prophylactic Uses of Bacteriophages

423

antibiotics. Ten patients were suffering with dacryocystitis, 12 with conjunctivitis combined with omphalitis, and 22 with omphalitis in combination with pyodermatitis. The IVSP was transfused every day for 4–5 days, and a positive outcome was observed in all 44 cases, with complete cure seen on days 5 through 77 depending on severity of infection and underlying diseases. The effect was observed earlier than in the control group (how much earlier is not reported). No relapses were seen. Infection was cured in the experimental group with mild LFID in 6.3 days and in the control group in 8.2 days. Intravenous phage therapy applied to the severe generalized forms of septicemia leads to a complete cure after around 17 days compared with 24 days in the control groups. At the same time, the doctors underlined that the entire duration of the hospital stay in both groups was usually about 40 days. This was explained by other medical factors affecting newborns (e.g., neurology problems, etc.). It is important to note that out of a total of 257 treated cases, 9 lethal outcomes were observed: 1 out of 149 in the experimental group (0.7%) and 8 out of 98 in the control group (8.2%) (Samsygina and Boni 1984; Samsygina 1985). No significant side effects of phage therapy were observed. After application of IVSP for treatment in the age group of 1 month to 1 year (38 children), only 2 (3%) reacted with an increase in temperature up to 38–39
C. This reaction was observed during transfusion and was alleviated by painkillers. These patients demonstrated a strong reaction to skin testing with staphylococcal anatoxin. The authors presumed that the reaction to IVSP was caused by sensitizing with derivatives of the bacterium, Staphylococcus. They suggested performing the skin tests with staphylococcal anatoxin and/or IVSP prior to the intravenous application of the bacteriophage preparation when treating young children (Samsygina 1985). In summary, the authors concluded that the IVSP preparation could be used as a supplement to antibiotic therapy for the treatment of staphylococcal infections of adults and children. The IVSP was recommended for use on patients with contraindications to antibiotic therapy. The IVSP lessened observations of toxic shock, shortened the healing period, improved sanitation of the initial infectious sites, and decreased the number of fatalities (Samsygina 1985). Following these studies, the staphylococcal phage was produced at the industrial production plant which was part of the Eliava Bacteriophage Institute. It was successfully provided, for intravenous use (infusions, transfusions, injections), in many clinics throughout the former Soviet Union for 20 years, until the collapse of the Soviet Union in 1990. Staphylococcal phage was mostly used for treatment of chronic septicemia, for treatment and prophylaxis of eye, ear, throat, and lung diseases, for healing burns, to treatment of the bacterial consequence of surgical operations on the bones and skull and for female infertility problems related to bacterial inflammation, etc.

Conclusions According to Larry Goodridge (2013), who reviewed old Western literature dedicated to use of phage therapy, those studies revealed much diversity in the drawn-out results and conclusions. Goodridge (2013) underlines that comparisons of these

424

N. Chanishvili and Z. Alavidze

studies are impossible due to lack of information regarding concentration of phage, numbers of different phages employed, method of preparation, diversity of administration routes, and the fact that in many of the reports no controls were included. A similar conclusion was made by Pavlova et al. (1973) who referred to prophylactic use of phages carried out in the Soviet Union in sites of high infection. In their review article, Pavlova et al. (1973) summarized the results of phage prophylaxis experiments carried out from 1934 to 1971 in 100 publications, 93 of which were published by Soviet researchers. The authors found that the results obtained by different researchers varied greatly. Thus, some researchers refer to different levels of decrease of disease in experimental groups varying from 1.2 times (Belikova 1941) to 46 times (Fisher 1949) when compared with the control group. Pavlova al (1973) explained such diversity of results as being due to methodological errors in the organization of the prophylactic measures. One of the major principles of such experiments was to provide quantitative and qualitative equity (i.e., similarity of epidemic conditions) in the establishing of experimental and control groups. This could be achieved by randomly selecting the people for the control group, carrying out of the same measures as in the experimental group but using a placebo, or by obligatory coding of the preparations used both in the experimental and control groups (Pavlova et al. 1973). The use of placebos and the coding of preparations was absent from most of the studies described in the previous sections. In addition, in some publications, such as Florova and Cherkass (1965), no control groups were used, and the efficiency of phage prophylaxis was estimated on the basis of comparing the results of prophylactic phage treatment with figures from outbreaks obtained in previous years, when no prophylactic measure was carried out (e.g., Kagan et al. 1964). This approach was inappropriate primarily due to the epidemic cycles of this disease. According to Pavlova et al. (1973), in some publications, the numbers in the experimental and control groups are missing, which makes the results obtained insufficient for proper statistical analysis. In other articles (e.g., Belikova 1941; Antadze 1957), the numbers of children receiving bacteriophage preparation and those that formed the control group differed significantly. The same erroneous approach was made by many other authors (Pavlova et al. 1973). In this review article, the authors explained the lack of efficiency of the dysenteric phage preparations as being due to bad organization of prophylactic measures. One of the main reasons for the skeptical attitude to phage prophylaxis was the lack of reliable data on the phage preparation, which was used at the time the epidemics arose. The same errors were found also in designing and implementation of therapeutic experiments. No randomized, double-blind, placebo-supported experiments have been done before. This may be explained by the fact that the mass applications were associated with war times when any treatment was considered as valuable and applicable. However, from the large amount of publications dedicated to use of phages for disease prevention, it is possible to conclude that, if the prophylactic measures are organized properly, the outcome could well be positive. For this purpose, consolidation of the efforts of the international scientific and medical communities is absolutely essential especially in the light of the background of growing antibiotic resistance.

Early Therapeutic and Prophylactic Uses of Bacteriophages

425

References Ackermann HW, Dubow MS (1987) Viruses of prokaryotes. CRC Press, Boca Raton Adhya S, Merrill C (2006) The road to phage therapy. Nature 443:754–755 Agafonov BI, Khokhlov DT, Zolochevsky MA (1984) Epidemiology of typhoid- paratyphoid infections and their prophylactics. Military-Med J 6:36–40 Albee FH (1933) The treatment of osteomyelitis by bacteriophages. J Bone Surgery 15:58 Albee FH, Patterson MB (1930) The bacteriophage in surgery. Ann Surg 91(6):855–874 Alessandrini A, Doria R (1924) Bacteriophagum antityphicum polyvalens. Med Klin 20:1447 Anikina TA (1982) Staphylococcal bacteriophage on the protein-free media with aminocrolin. Cand Diss, Gorky Anpilov LI, Prokudin AA (1984) Prophylactic effectiveness of the dry polyvalent dysentery bacteriophage in organized communities. Military-Med J 5:39–40 Antadze VS (1957) The main trends in the research of the issues related to bacteriophages and the ways of their resolution. In: Bacteriophages, Selected Articles of the Inter-Institutional Conference held in Tbilisi, October 26–29, 1957, pp 5–18 Appelmans R (1921) Le dosage du bactériophage. Compt Rend Soc Biol 85:1098–1099 Arsentieva VA (1941) Intravenous use of high doses of bacteriophages in surgery. Sov Med 9:18–20 Arshba SY (1942) Phage treatment of the open gun wounds and bone and joint injuries. In: Phage therapy of wounds in the conditions of evacuation hospital. GruzMedGiz, Tbilisi, pp 18–26 Babalova EG, Katsitadze KT, Sakvarelidze LA, Imnaishvili NS, Sharashidze TG, Badashvili VA, Kiknadze GP, Meipariani AN, Gendzekhadze ND, Machavariani EV, Gogoberidze KL, Gozalov EI, Dekanosidze NG (1968) On the issue of prophylactic importance of the dry dysenteric bacteriophage. J Microbiol Epidemiol Immunol (JMEI) 2:143–145 Bayne-Jones S and Sandholzen LA (1933) Changes in the shape and size of bacterium coli and Bacillus megaterium under the influence of bacteriophage. A motion photomicrographic analysis of mechanism of lysis. J Exp medicine 57:279 Beckerich A, Haudoroy P (1922) Le bacteriophage dans le treatment de la fever typhoid. Comp Rend Soc Biol 86:168 Belikova MA (1941) Experience of phage prophiyaxis of summer dysentery among the young children performed in the city Stalingrad. J Microbio Epidemiol Immunol (JMEI) 5-6:142 Beridze MA (1938) Role of bacteriophage therapy in combating purulent skin infections. Medgiz, Tbilisi Blankov BI (1941) Analysis of the results of phage prophylaxis of dysentery among the contacting people (report # 1). J Microbiol Epidemiol Immunol (JMEI) 5-6:125–131 Blankov BI, Zherebtsov ID (1941) Experience on the multiple phaging of the contacting population in the fight against dysentery (report # 2). J Microbiol Epidemiol Immunol (JMEI) 5-6:131–136 Bochorishvili TV (1984) Clinical evaluation of effectiveness of various methods of immune and intravenous therapies of the generalized forms of staphylococcal infections. Tbilisi, Cand Diss Bordet J (1925) La theorie de l’autolyse transmissible et les objections de d’Herelle. Compt Rend Soc Biol 93:143 Bruynoghe R, Maisin J (1921) Essais de thérapeutique au moyen du bacteriophage du Staphylocoque. Compt Rend Soc Biol 85:1120–1121 Chanishvili N (2012a) Phage therapy-history from Twort and d'Herelle through soviet experience to current approaches. Adv Virus Res 83:3–40. https://doi.org/10.1016/B978-0-12-3944382.00001-3 Chanishvili N (2012b) A literature review of the practical application of bacteriophages. Nova Science Publishers, New York City Chanishvili TG, Giorgadze TV, Chirakadze IG, Chkhenkeli NK (1974) Experience of obtaining of apyrogenic preparation of Staphylococcus phage on the SSM and study of exotoxins appearing in the course of phage lysis. In: materials of the symposium dedicated to the 50th anniversary of Tbilisi Institute of Vaccine and Sera. Tbilisi, pp 263–265

426

N. Chanishvili and Z. Alavidze

Chirakadze IG, Chanishvili TG (1964) Study of the potential of obtaining Staphylococcus phage for intravenous administration. In: Abstracts of the Scientific Session dedicated to Staphyloccocus infection, Leningrad, pp 56–57 Chkhetia N (1984) Treatment of lung diseases. Tbilisi, Cand Diss Cipollaro AC, Scheplar AE (1933) Therapeutic uses of bacteriophages in the pyodermias. Arch Dermat 25:280 Compton A (1929) Anti-dysentery bacteriophage in the treatment of bacillary dysentery: record of 66 cases treated, with inferences. Lancet 2:2735. https://doi.org/10.1016/S0140-6736(01) 02171-7 d’Herelle F (1917) Sur un microbe invisible antagoniste des bacilles dysentériques. Comp Rend Acad Sci Paris 165:373–375 d’Herelle F (1935) Bacteriophage and phenomenon of recovery. TSU Press, Tbilisi da Costa Cruz J (1924) Le traitment des dysentéries bacillaires par le bactériophage. Compt Rend Soc Biol Paris 91:845 Davison WC (1922) The bacteriolysant therapy of bacillary dysentery in children: therapeutic application of Bacteriolysants; d’Herelle's phenomenon. Am J Dis Child 23:531–534 Dutton LO (1926) Role of bacteriophage in Streptococcus infection. J Inf Dis 39:48 Eaton MD, Bayne-Jones S (1934) Bacteriophage therapy: review of the principles and results of the use of bacteriophage in the treatment of infections. JAMA 103:1769–1776 Ebert BP, Shapiro FI (1938) Bacteriophage as a therapeutic and prophylactic factor in case of staphylococcal infections. J Microbiol 214(10):53–57 Eijkman C (1901) Ueber Enzyme bei Bakterien und Schiemmelpilzen. Centralbl Bakt 29:841–848 Emmerich R, Löw O (1899) Bakteriolytische Enzyme als Ursache der erworbenen Immunität und die Heilung von Infectionskrankheiten durch dieselben. Z Hyg Immun 31:1 Fedorovich DP (1944) Bacteriophage treatment of gunshot wounds. Soviet Medi 6:26–27 Fejgin B (1925) Sur le principle lytique anti-diphtherique. Comp Rend Soc Biol Paris 93:365 Fisher L (1949) Bacteriophage, contemporary knowledge about the nature and mechanisms of action. In: Bacteriophage: selected guides for application of phage preparations, 2nd edn, Tomsk, Russia Florova NN, Cherkass FK (1965) Results of mass application of polyvalent dysenteric bacteriophage. J Microbiol Epidemiol Immunol (JMEI) 3:118–125 Fruciano DE, Bourne S (2007) Phage as an antimicrobial agent: d’Herelle’s heretical theories and their role in the decline of phage prophylaxis in the west. Can J Inf Dis Med Microbiol 18(1):19–26 Gamaleya NF (1898) Bacteriolysins – ferments destroying bacteria. Russ Arch Pathol Clin Med Bacteriol 6:607–613 Georgadze IA (1974) Fifty Years of the Tbilisi Scientific-Research Institute of Vaccine and Sera of the Ministry of Health of the USSR. In: Selected Articles of the Jubilee dedicated to 50th Anniversary of the Tbilisi Institute of Vaccine and Sera. TIVS: Tbilisi Gildemeister E (1921) Uber das d’Herellesche Phenomenon. Berlin Klin Wschr 58:1355 Goodridge LD (2013) Bacteriophages for managing Shigella in various clinical and non-clinical settings. Bacteriophage 3(1):e25098 Gvazava AI (1957) Results of therapy of deep forms of pyodermitis with the expired phage preparations. Cand Dis, Tbilisi Haler D (1938) Use of the bacteriophage in an outbreak of dysentery. Br Med J 2:698–670. https:// doi.org/10.1136/bmj.2.4056.698 Hankin ME (1896) Les microbes des riviíres de l’Inde. Ann Inst Pasteur (Paris) 10:175–176 Izashvili NP (1940) Clinical observation upon the therapeutic effect of phage therapy in purulent processes. Tbilisi, Cand Dis Izashvili NP (1957) Combined use of penicillin and bacteriophage for treatment of some purulent processes. In: Bacteriophage Research Selected Articles of Inter-Institutional Conference held in Tbilisi on October, 26–29, 1955, pp 397–405 Johnston MM, Ebbs JH, Kaake MJ (1933) Bacteriophage therapy in acute intestinal infection (summer diarrhea). Can Public Health J 24:443–446

Early Therapeutic and Prophylactic Uses of Bacteriophages

427

Jukov-Werezhnikov NN, Fruauf VA (1934) Zur frage uber Wesen und Bedeutung der Bakteriophagie. II Morphologische Vwranderungen Bac. Typhi abdominalis unter Einfluss des Bakteriophags. Revue De Microbiologie D’Epidemiologie et de Parasitologe Institut de Microbiologie et Epidemiologie du Sud-Est de RSFSR a Saratov 13(4):263–272 Kagan MI, Kuznetsova EV, Teleshevskaya EA (1964) To the issue of epidemic effectiveness of the planned phaging in the day-nurseries. J Microbiol Epidemiol Immunol (JMEI) 7:89–102 Karamov S (1938) Experience of phage therapy for treatment of typhoid fever. Selected Articles of Azerbaijani Institute of Epidemiology and Microbiology 6(1):101–105 Karpov SP (1946) The specific bacteriophage in relation to the issue of combating typhoid and paratyphoid diseases. J Microbiol Epidemiol Immunol (JMEI) 1–2:40–44 Kazarnovskaya SS (1933) Bacteriophagyia. USSR Academy of Sciences Press, Leningrad Kazhal N, Iftimovich R (1968) From the history of the fight against bacteria and viruses. Bucharest, Nauchnoe Izdatelstvo Kessel JF, Rose EJ (1933) Bacteriophage therapy in bacillary dysentery of the Flexner type. Ann Intern Med 6:1193–1199. https://doi.org/10.7326/0003-4819-6-9-1193. Accessed on Sept 2019 Khuskivadze ZF (1954) To the issue of phage therapy of the deep forms of pyodermitis. Tbilisi, Cand Dis Kokin GA (1941) Use of phage therapy in surgery. Soviet Med 9:33–35 Kokin GA (1946) Phage therapy and prophylaxis of gas gangrene. In: military medicine during the Great Patriotic War Moscow, vol 3. 56–63 Krestovnikova VA (1947) Phage treatment and phage prophylactics and their approval in the works of the soviet researchers. J Microb Epidemiol Immunol (JMEI) 3:56–65 Krueger AP, Scribner EJ (1941) The bacteriophage: its nature and its therapeutic use (I) JAMA 116:2160–2167 Krupp K, Madhivanan P (2015) Antibiotic resistance in prevalent bacterial and protozoan sexually transmitted infections. Indian J Sex Transm Dis AIDS 36(1): 3–8. https://doi.org/10.4103/25890557.156680. Accessed on 2Sept 2019 Kurochka VK, Karniz AF, Khodyrev AP (1987) Experiences of implementation of preventive antiepidemic measures in the center of intestinal infections with water transmission mechanism of morbidity. Military-Med J 7:36–37 Lipkin NM and Nikolskaya II (1940) Experience of phage therapy of dysentery. In: Selected articles, Kuibishev Red Army Military-Medical Academy, Kuibishev, Issue 4, pp 193–198 Makashvili EG, Bagdoeva A, Tarasashvili N, Anjaparidze TG (1942) Aerobic microflora of wound infections. In: Phage therapy of wounds in the conditions of evacuation hospital. GruzMedGiz, Tbilisi, Georgia, pp 54–63 Manolov DG, Sekunova VN, Somova EE (1948) Experience of therapy of typhoid fever by intravenous administration of the phage. J Microbiol Epidemiol Immunol (JMEI) 4:33 McKinley EB (1923) Bacteriophage in the treatment of infections. Arch Intern Med 32:899–910 Meladze DG, Mebuke MG, Chkhetia NS, Kiknadze NY, Koguashvili GG, Timoshuk II, Larionova NG, Vasadze GK (1981) Effect of the staphylococcal bacteriophage for treatment of purulent infections of lungs and pleura. Breast Surg (“Grudnaya Khirurgia”) 1:53–56 Melnik MI, Khastovich RI, Mitelman MM (1935) Phage therapy of dysenteric patients. In: Selected articles of the Mechnikov institute, Kharkov, Issue 1, v1, pp 89–92 Murray JE (1938) The treatment of bacillary dysentery with bacteriophage. Practitioner 141:199–201 Myelnikov D (2018) An alternative cure: the adoption and survival of bacteriophage therapy in the USSR, 1922–1955. J Hist Med Allied Sci 73(4):385–411 https://doi.org/10.1093/jhmas/jry024. Accessed on 2 Sept 2019 Nadiradze MM (1983) Method of preparation of the apyrogenic intravenous Staphylococcus phage. In: Selected articles bacteriophages – theoretical and practical issues. Mechnikov Institute of Microbiology, Moscow, pp 215–218 Northrop J (1937) Chemical nature and mode of formation of pepsin, trypsin and bacteriophage. Science 86:479–483 Northrop J (1938) Concentration and purification of bacteriophage. J Gen Philos 21(3):335–366

428

N. Chanishvili and Z. Alavidze

Pavlova LI, Sumarokov AA, Solodovnikov YP, Nikitiuk NM (1973) Use of dysentery bacteriophage as a means of preventing dysentery (review of literature). J Microbiol Epidemiol Immunol (JMEI) 7:27–32 Plankina ZA, Nikonov AG, Sayamov RM, Kotliarova RI (1961) Control of cholera in Afghanistan. J Microbiol Epidemiol Immunol 32:202–204 Podvarko AG (1964) Bacterial intestinal infections and respiratory tract infections, coccoidal infections. In: Multivolume guide for Microbiology, Clinics and Epidemiology of Infectious diseases, Meditsina, Moscow, vol 6. pp 245–254 Pokrovskaya VP (1942) Treatment of wounds with bacteriophages. Medgiz, Moscow Purtseladze MD (1941) Treatment of purulent forms of mastitis. Obstet Gynecol (“Akusherstvo i ginekologia”) 11–12:22–24 Querangal des Essarts J (1933) Le bactériophage dans une épidémie de dysenterie bacillaire (Applications thérapeutiques et prophylactiques). Bull Soc Path Exot 26:979–981 Rice TB (1930) Use of bacteriophage filtrates in treatment of suppurative conditions: report of 300 cases. Am J Med Sci 179:345–360 Riding D (1930) Acute bacillary dysentery in Khartoum province, Sudan, with special reference to bacteriophage treatment: bacteriological investigation. J Hyg (Lond) 30:387–401. https://doi. org/10.1017/S0022172400010512. Accessed on Sept 2019 Rodigina AM (1938) Pneumococcal bacteriophage: its application for treatment of the ulcerous corneal serpens. Sov Stomatol (“Sovetskaya Stomatologia”). Perm, Russia 1:11–20 Rohde C, Resch G, Pirnay JP, Blasdel BG, Debarbieux L, Gelman D, Górski A, Hazan R, Huys I, Kakabadze E, Łobocka M, Maestri A, Almeida G, Makalatia K, Malik DJ, Mašlanová I, Merabishvili M, Pantucek R, Rose T, Štveráková D, Van Raemdonck H, Verbeken G, Chanishvili N (2018) Expert opinion on three phage therapy related topics: bacteriophage resistance, phage training and Prophages in bacterial production strains. Viruses 10:178–193 Rolleston JD (1926) Vorläufiger Bericht der staatlichen Untersuchungs-kommission über die Typhus-Epidemie in Hannover Herbst 1926. Klin Wochenschr 5:2412–2414. https://www. cabdirect.org/cabdirect/abstract/19272701009. Accessed on 2 Sept 2019 Ruchko I, Tretyak K (1936) Therapeutic effect of Staphylococcus phage for oral and dental infections. Sov Stomatol 6:35–40 Ruska H (1940) Die Sichtbarmachung der bakteriophagen Lyse im Ubermikroskop. Naturwissen 28:45–46. https://doi.org/10.1007/BF01486931. Accessed on 2 Sept 2019 Samsygina GA (1985) Purulent-inflammatory infections of newborns (etiology, risk factors, clinical-immunology criteria for diagnostics and therapeutic tactics). Moscow, Doct Dis Samsygina GA, Boni EG (1984) Bacteriophages and phage therapy in pediatric practice. Pediatrics 4:67–68 Sapir IB (1939) Observations and recommendations related to phage therapy of dysentery. In: Proceedings of the Moscow Institute of Infectious Diseases after II Mechnikov, pp 135–151 Sayamov RM (1963) Treatment and prophylaxis of cholera with bacteriophage. Bull World Health Organ 28:361–367 Schless RA (1932) Staphylococcus aureus meningitis: treatment with specific bacteriophage. Am J Dis Child 44:813–822 Shvelidze KD (1970) Treatment of deep forms of Staphylococcal dermatitis with the staphylococcal bacteriophage and some aspects of immune reaction. Cand Dis” Tbilisi Sirbiladze NY (1942) Anaerobic micro-flora of wounds and their phage therapy. In: Phage therapy of wounds in the conditions of evacuation hospital. GruzMedGiz, Tbilisi, pp 66–71 Skvirskyi PV, Sinitskyi AA, Lisyanskaya SM (1938) About specific effectiveness of bacteriophages therapy of typhoid fever and Para-typhi. Sov Med Gazette 17-18:764–781 Spence RC, McKinley EB (1924) Therapeutic value of bacteriophage in treatment of bacillary dysentery. South Med J 17:563–568. https://doi.org/10.1097/00007611-192408000-00005. Accessed on 2 Sept 2019 Stone FM, Hobby GL (1934) A coccoid form of C. diphtheriae susceptible to bacteriophage. J Bact 27:403–417

Early Therapeutic and Prophylactic Uses of Bacteriophages

429

Stout BF (1933) Bacteriophage therapy. Texas State J Med 29:205–209 Summers WC (1999) Felix d’Herelle and the origins of molecular biology. Conn: Yale University Press, New Haven Summers WC (2001) Bacteriophage therapy. Annu Rev Microbiol 55:437–451 Summers WC (2012) The strange history of phage therapy. Bacteriophage 2(2):130–133. https:// doi.org/10.4161/bact.20757. Accessed on 2 Sept 2019 Sutin IA (1947) Bacteriophage and its use in medical practice. Stalingrad Tavberidze LN (1993) Post-traumatic treatment of the long bones. Tbilisi, Cand Dis Town AE, Frisbee FC (1932) Bacteriophage in ophthalmology. Arch Ophthalmol 8(5):683-689 https://doi.org/10.1001/archopht.1932.00820180055005. Accessed on 2 Sept 2019 Tsulukidze AP (1940) Phage treatment in surgery. Surgery (“Khirurgia”) 12:132–133 Tsulukidze AP (1941) Experience of use of bacteriophages in the conditions of war traumatism. Medgiz, Tbilisi, pp 334–418 Tsulukidze AP (1942a) Bacteriophage treatment of anaerobic infection. In: Phage therapy of wounds in the conditions of evacuation hospital. GruzMedGiz, Tbilisi, pp 3–18 Tsulukidze AP (1942b) To the issue of bacteriophage therapy of the soft tissue injuries. In: Phage therapy of wounds in the conditions of evacuation hospital. GruzMedGiz, Tbilisi, pp 45–53 Tsulukidze AP (1957) Results of phage application for treatment of surgical infections. In: Bacteriophages. Tbilisi, Georgia, pp 99–108 Twort F (1915) An investigation on the nature of ultramicroscopic viruses. Lancet 11:1241 Typhus-Epidemie in Hannover Killerkeime aus dem Wasserhahn (2011). Speigel Online. http:// www.spiegel.de/einestages/typhus-epidemie-in-hannover-a-947264.html. Accessed on 2 Sept 2019 Vartapetov AY (1941) Predictive significance of correspondence of the results of in vitro and in vivo lyses in cases of treatment of pyodermatitis. In: Selected Articles of the Institute of Dermatology and Veneric Diseases, Tbilisi, vol 2–3, pp 93–97 Vartapetov AY (1947) Results of phage therapy of deep forms of pyodermitis. In: Selected articles of the Institute of Dermatology and Veneric Diseases, Tbilisi, pp 205–207 Vartapetov AY (1957) Bacteriophage therapy of deep forms of staphylococcal dermatitis, In: Bacteriophage Research In: Selected Articles of Inter-Institutional Conference held in Tbilisi on October, 26–29, 1955, pp 411–426 Vartapetov AY, Gogolashvili RA, Tsaava AA, Gilels A, Soboleva VA (1974) Experience of introduction of phage therapy into practice of some industrial manufacturing facilities for prophylaxis of deep forms of Staphylo-dermitis. In: Selected articles of the Jubilee symposium dedicated to the 50th anniversary of the Tbilisi Institute of Vaccine and Sera, Tbilisi, pp 147–149 Vlasov KF, Artemenko EA (1946) Treatment of chronic dysentery. Sov Med (“Sovetskaya Medicina”) 10:22–28 Yermolieva ZV (1939) About bacteriophage and its application (review). J Microb Epidemiol Immun (JMEI) 9:9–17 Yermolieva ZV (1942) Cholera (monograph). Moscow, Medizdat Yermolieva ZV, Yakobson NM (1943) Diagnostics of cholera and effect of phage prophylaxis during the cholera outbreaks. Achievements of the Soviet Medicine during the years of Patriotic War Moscow 1:50–64 Yermolieva ZV, Yakobson NM (1949) Bacteriophage. In: Microbiology research methods for infectious diseases. Moscow Yukelis II (1946) Treatment of patients with furuncles and other pustular diseases with Di-phage. Military Med J 6:14–16

Part IV Bacteriophage Technology

Isolation of Bacteriophages Frits van Charante, Dominique Holtappels, Bob Blasdel, and Benjamin H. Burrowes

Contents Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Basic Phage Isolation Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Biases in Isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spot Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Increasing Phage Concentrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Generation of Pure Isolates and Phage Stocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Storing Isolated Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Where to Hunt for Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isolation from the Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Common Sources for Isolation of Therapeutic Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isolation from Lysogens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isolation from Diverse Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Resulting Phage Phenotypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Altered Procedures and Their Effect on Host Range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plaque Morphology Indicates Diverse Phage Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Isolation for Biotechnology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Implications for the Use of Phages as Antimicrobials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . High-Throughput (HTP) Phage Isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

434 436 438 439 440 440 442 443 444 445 445 446 447 449 450 451 453 454 455

F. van Charante Ghent University, Ghent, Belgium D. Holtappels Laboratory of Gene Technology, Leuven, Belgium B. Blasdel Laboratory of Gene Technology, Leuven, Belgium Vésale Pharma, Noville-Sur-Mehaigne, Belgium B. H. Burrowes (*) Evolution Biotechnologies, Georgetown, TX, USA e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_14

433

434

F. van Charante et al.

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456 Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457

Abstract

Before any phage can be studied, or used for its biological properties, it must first be isolated. As such, isolation is a critical step – indeed, the critical step – in many explorations of phage biology and biotechnology. There are several techniques, both classical and modern, by which phages can be isolated, and selection of the proper method often depends on the intended use of the phage. In this chapter, we discuss the general principles of phage isolation and techniques to obtain pure phage isolates from a variety of sources, with a particular focus on the isolation of therapeutic phages.

Glossary Bacteriocins Burst size Chronic infection

Double agar overlay

Enrichment

Halo

Host strain Indicator strain

Toxic proteins produced by bacteria that kill or inhibit the growth of other, usually closely related, bacteria. The average number of progeny phages produced from the lytic infection of a single bacterial host cell. A phage life cycle in which the host cell does not lyse but instead releases progeny phages continuously by extrusion from the cell membrane. A method used for plaque visualization in which a thin layer of soft agar is inoculated with host cells and phages and then poured over a thicker layer of solid agar. The soft agar allows phage diffusion, while the hard agar provides support and nutrients to the bacterial lawn in the soft agar. A process whereby a phage-containing sample is incubated in the presence of host bacteria and growth substrates to amplify any phages present in the sample that can propagate on the provided host bacterium. A region encircling a plaque center that is less turbid than the surrounding lawn but more turbid than the plaque center. Halos are usually distinct and are not a general, graduated reduction of turbidity at the plaque periphery. Bacterial strain used for the propagation of a given phage strain. Bacterial strain used to support the generation of plaques such as in the course of double agar overlays. Usually this is also the host strain, but not always.

Isolation of Bacteriophages

Latent period Lysate

Lysogen Mesophile

Phage stock

Plaque assay

Plaque purification

Plaque

435

Time between the infection of a cell and the release of progeny phages. Phage-containing liquid medium as present after viral propagation. Usually this will be the spent liquid medium after mixing phages and host cells and allowing the reaction to produce progeny phages. However, if phages are grown in solid or semisolid media, then the lysate is usually the liquid buffer into which the phages and other materials are extracted from the semisolid medium. Lysates are usually centrifuged and/or filtered to remove viable bacterial cells and large bacterial debris. A bacterial strain that harbors a viable (or active) prophage (see “Lysogeny” chapter). A bacterial strain that grows best at moderate temperatures, typically (although not categorically) between 20 and 45  C. A sample of phage, usually but not always having been plaque purified, that is used for multiple ongoing experiments or purposes. Using the same phage stock ensures uniformity of the downstream processes. A means of enumerating phages in a sample by serially diluting the sample in known increments and plating measured aliquots of the dilutions with indicator bacteria in order to count the number of plaques. The volume of dilution used, the amount of dilution, and the number of plaques counted are then used to calculate the phage titer in the original sample (see ▶ “Detection of Bacteriophages: Phage Plaques” chapter). A process of obtaining a pure, clonal sample of a phage. A plaque is picked either by removing the agar of the plaque itself or by touching the surface of the plaque with a sterile inoculating loop. Either this sample can then be directly struck onto the surface of solid media and overlaid with a lawn of host bacteria, or it can be diluted and mixed with host cells in a lawn of bacteria. After incubation, the process is repeated from a resulting plaque. A minimum of three rounds of purification is usually required for the resulting phage plaque to be considered clonally pure. A visible clearing in a lawn of bacteria caused by a single phage or phage-infected cell. Plaques are discussed in detail (see ▶ “Detection of Bacteriophages: Phage Plaques” chapter).

436

Prophage

Psychrophile Psychrotroph Soft agar

Superinfection immunity

Temperate

Thermophile Virulent

F. van Charante et al.

A temperate phage genome residing within a bacterial genome. While in the prophage state, most phage genes are switched off, and the phage’s genome is replicated along with the host. When induced, the prophage becomes active and viral replication occurs. Bacteria that grow and live at cold temperatures, below the typical mesophilic range. Cold-tolerant bacteria that can grow at low temperatures but have maximal growth rates in the mesophilic range. Sometimes called semisolid agar or top agar. Soft agar contains enough agar to set but not enough to form a solid layer, typically 0.2–0.7% agar. Soft agar allows phage diffusion while minimizing bacterial movement and is therefore used for plaque formation. A property of a lysogen in which the prophage blocks subsequent infection by the same or a closely related phage (see “Lysogeny” chapter). A phage that can exist in a prophage state. In phage life cycle terms, the prophage state is ultimately followed by prophage induction, progeny generation, and host cell lysis, although other fates such as mutation to a defective state are also possible (see “Lysogeny” chapter). Bacteria that grow at high temperatures, above the mesophilic range. A phage with an obligately lytic life cycle, in which successful infection is only followed by progeny production and cell lysis. There is no prophage state.

Introduction The first bacteriophages were isolated more than a hundred years ago, early in the twentieth century by Frederick Twort and Félix d’Hérelle (see ▶ “The Discovery of Bacteriophages and the Historical Context” chapter). The first time bacteriophages were mentioned, though not by that name, appears to have been in 1915 when Twort identified “glassy areas” caused by samples of “cultivations from glycerinated calf vaccinia” through filtration of extracts of this material. With his efforts, he intended to discover a method to independently cultivate these ultramicroscopic viruses that passed through the finest filters available at the time. This was not successful, though he noted instead that the filtrate was able to turn Micrococcus (at the time a morphological description and not a phylogenetic identification) colonies transparent. The glassy clearings were no longer able to grow, and new colonies or cultures

Isolation of Bacteriophages

437

that were infected with some of the glassy areas (plaques, as we now call them) resulted again in glassy clearings and transparent cultures (Twort 1915). A more complete understanding of bacteriophages came from Félix d’Hérelle, a microbiologist at the Pasteur Institute who was sent to investigate an outbreak of dysentery among French mounted infantrymen (d’Hérelle 1916). In his investigations of the bacteria over the next 18 months, he found that some seemingly sterile Chamberland filtrates were capable of lysing dysentery bacilli (likely Shigella). In two short pages, d’Hérelle (1917) described the experiments that he performed showing that this lytic property could be serially passed from one culture to the next by transferring 10 6 dilutions 50 times. Similarly, he showed that there was no dilution of these lysed cultures that would produce hazy subinhibitory growth when plated over a lawn of bacteria, like any antibacterial toxin would, but instead would display a number of clear plaques equal to the concentration that would lyse a liquid culture. From these observations, d’Hérelle (1917) radically intuited that he had discovered “un microbe invisible antagoniste des bacilles dysentériques,” describing it as “un bactériophage obligatoire.” What he discovered was that the plaques he observed were a clearing in his bacterial lawns that was left behind by the consuming activity of an invisible but self-amplifying entity. Impossible to amplify in the absence of their bacterial hosts, he found that phages could be readily isolated by exposing lawns of a specific bacterium to environmental samples. In this way, any phages present in those samples, capable of infecting the indicator bacteria, would grow at the expense of their host, forming the visible lysis originally observed by Twort. d’Herelle was able to collect individual plaques and propagate the phages on new bacterial lawns. Bacteriophage isolation, especially for therapeutic purposes, has been reviewed before (e.g., Gill and Hyman 2010, Seeley and Primrose 1982, and WeberDabrowska et al. 2016). Furthermore, several books give detailed methodological outlines and considerations for isolating phages from a variety environmental samples (e.g., Carlson 2005, Van Twest and Kropinski 2009, and Wommack et al. 2009), including from anaerobes (García-Aljaro et al. 2018) and fastidious bacteria (Matsuzaki et al. 2018), or for therapeutic use (Łobocka et al. 2014; Sillankorva 2018). Łobocka and colleagues discuss in some detail the many issues faced when it comes to isolating and selecting the most suitable therapeutic phages (Łobocka et al. 2014). Methods are also available for the isolation of phages based on more specific requirements, such as thermophilic hosts (Fulton et al. 2009), cyanobacterial hosts (Millard 2009), and temperate phages (Carlson 2005; Raya and Hébert 2009). While the methods show considerable variety, in each case the authors highlight the importance of selecting the right bacterial host(s), methods, and isolation conditions in order to optimize the likelihood of isolating phages that are fit for the intended purpose. Ultimately, in terms of phage isolation, as with anything, you get what you select for, intentionally or otherwise. In this chapter we introduce some of the various techniques and considerations used for phage isolation, considering the different purposes that isolated phages can serve. In particular, we discuss which environments are suited to hunt for phages and

438

F. van Charante et al.

ways to search for phages adapted to therapeutic contexts as well as the methods best adapted to perform isolations from these samples. We pay special attention to the effects that different methods can have on the properties of resulting phages. We also discuss the particular methods used to isolate phages from lysogens, to isolate difficult to culture phages, and to isolate phages when it is the bacterial host that is difficult to culture. We then conclude by discussing the implications of phage isolation methods for phage therapy.

Basic Phage Isolation Techniques The techniques developed by d’Herelle still form the basis for much of phage isolation today (Fig. 1). Phages are generally present in any environment where their hosts are found, which should serve as the first principle behind any phage hunting effort (see ▶ “Bacteriophage Discovery and Genomics” chapter). However, not all target bacterial species (or even strains) are created equal when it comes to the ease with which phages can be isolated (e.g., Mattila et al. 2015), so perseverance and adaptability by the phage hunter are often necessary. As with d’Herelle’s original discovery, phage isolation typically ends with a plaque assay (see ▶ “Detection of Bacteriophages: Statistical Aspects of Plaque Assay” chapter). In many cases, a phage concentration or propagation strategy is first used, such as use of enrichment cultures or concentration of phages onto a filter membrane, for the isolation of phages. In this section we focus on these latter strategies – concentration,

2 Plaque Assay 1 Environmental Sample

4 Identification and Harvesting of Plaques

6 Production of Phage Stock(s)

5. Plaque Purification 3 Optional Step: Enrichment Desired Optimal Conditions and Host(s)

Fig. 1 Flow chart illustrating the basic techniques for phage isolation. An environmental sample (1) can be either directly plated (2) onto a lawn of isolation host bacteria to provide minimal selective pressure and thereby a maximally representative sampling of environmental phage or instead first be enriched (3), followed by (2) to select for optimal growth in vitro as well as potentially other specific properties. The plaques obtained in the plaque assay (4) are then plaque purified at least three times (5) before a pure plaque is used to produce a purified phage stock (6)

Isolation of Bacteriophages

439

enrichment, and plaquing – and in the following section on phage isolation sources. First, though, we provide an introduction to the issue of biases in isolation.

Biases in Isolation Ideally, the purpose intended for a phage would be considered before its isolation, since the specific methods employed, as well as the source used for the phage isolation, can have a large impact on which phages are isolated and which properties might intentionally or inadvertently be selected for. For example, an effort that seeks to isolate a maximally environmentally representative sample of phages present in an ecosystem should contain as few biasing steps as possible. However, an effort to isolate phages for therapeutic or other antimicrobial purposes should select strongly for phages with useful properties such as broad host range, ease of amplification, and capacity to infect in the specific contexts in which they will be used. Therefore, it is important to consider the specific ecology from and for which the phage will be isolated as this can relate to the final purpose of the phage. Isolation biases might either enrich or deplete phages of interest even before the user has a chance to identify and purify them from the sample. Bacterial host, media (including nutrient composition, pH, salinity type and concentration, osmolarity, etc.), incubation temperature and time, oxygenation, and other variables may result in selection biases during the enrichment and isolation processes. Another issue in this regard is one of inadvertent biasing of isolations against different phage virion types. Before a plaque assay (process 2 in Fig. 1) or a possible enrichment step (process 3 in Fig. 1), bacteria and other debris are often cleared from samples through a variety of physical and chemical methods. These include centrifugation, filtration, or treatment with chloroform. Both filtration and centrifugation can be employed to remove or reduce bacteria and debris from the sample. While excessive centrifugation can remove larger phages, filtration and centrifugation generally do not heavily impact the types of phages isolated. The same, however, is not true for various chemicals used to sanitize samples. A good example of such an effect is treatment with chloroform. Although chloroform can be very effective at lysing bacteria, thereby removing viable bacteria and potentially releasing phages, and is often employed in phage amplification protocols, it can also be harmful to certain phages (Kuo et al. 1969; Lopez et al. 1977). For example, while they are relatively rare, any phage with lipid capsid components will experience virion damage and be left unable to infect their host (Ellis and Schlegel 1974; Lopez et al. 1977; Vidaver et al. 1973). Additionally, it has been reported that chloroform can damage the tail of certain phages, which also leads to a severe loss of viable phage titer (Chilar et al. 1978; Fay and Bowman 1978). While for some applications it may indeed be desirable to select for chloroform-insensitive phages, as chloroform can be a convenient addition to many phage protocols, this will reduce the diversity of the phages isolated. For example, phage PM2, having a spherical proteinaceous lipid core, was isolated from seawater by Espejo and Canelo (1968) using standard techniques, but without the use of chloroform. The same approach was used to isolate other membrane-

440

F. van Charante et al.

containing phages, such as φ6 (Vidaver et al. 1973) and PRD1 (Olsen et al. 1974). Use of chloroform in phage isolation can thus result in an underappreciation of the prevalence of chloroform-sensitive phage types that are found within environments. Simple storage of samples, freezing, or even exposure to light also can bias what phage types are present at the point of enrichment or sampling. This is because different phages will display different levels of durability either over time or under differing physical conditions (Jończyk et al. 2011).

Spot Testing Prior to plaque isolation, with either untreated or enriched samples, the sample can be added to bacterial lawns by dropping a small aliquot (e.g., 5–20 μl) of the sample onto a lawn to test for the presence of potential phages of interest – this process is usually called spotting. Confluent spots are not the same as plaques (see below) in that such confluence is the result of bacterial lysis initiated by multiple phages (or other bactericidal compounds such as secondary metabolites present in the sample) (see ▶ “Detection of Bacteriophages: Phage Plaques” chapter), but they are a simple way of determining whether or not a sample might contain phages that lyse the target strain(s) without the need to dilute or streak out the sample. This test also allows the user to test multiple phage-containing samples on a single bacterial lawn (as shown in Fig. 2). Spotting can even replace the plaque assay step (process 2) shown in Fig. 1 if phage titers are sufficiently low as to generate isolated plaques (as indicated by the white arrow in Fig. 2) that the user can reasonably harvest without disturbing other plaques. Given that only a small aliquot of sample is used, certain low-titer, unenriched, or unconcentrated samples may not produce lysis after spotting. Similarly, the absence of lysis could be taken as an indicator that the sample requires enrichment or concentration prior to further testing. It is also important to note that false-positive results can be obtained as other bactericidal compounds can cause clearing of the bacterial lawn. As such, the presence of phages in the clearing should be evaluated by an additional plaque assay.

Increasing Phage Concentrations As mentioned above, one commonly used method to isolate phages from samples where they exist in low abundance is to include an enrichment step (Carlson 2005; Sillankorva 2018; Van Twest and Kropinski 2009). This involves incubation of the sample along with the desired host(s) in liquid culture, often overnight depending on the host, to amplify any phages present in the sample. Creative forms of enrichment can be used to specifically select for phages with desired properties by constructing an enrichment step where those properties will be advantageous to any desired phages in the sample. An enrichment step especially favors phages specific for the species and strain of bacterium used during enrichment (Jensen et al. 1998), as well

Isolation of Bacteriophages

441

Fig. 2 Phage isolation after enrichment step using a spot assay. Here, different samples are tested against an Agrobacterium sp. strain. Undiluted samples are spotted on top of the bacterial lawn that contains the host of interest. The assay measures the phages’ ability to kill the host without necessarily measuring the phages’ ability to replicate at the expense of the host. Spot 15, indicated by the white arrow, appears to show the presence of individual plaques indicating lower numbers of phages active against the bacterial strain compared to the spots numbered 2, 4, 5, 11, 12, and 16

as the specific conditions of enrichment such as temperature, medium, etc. Therefore, such a step is ideal for when a certain phage is desired or when it is not necessary for isolated phages to reflect the diversity of the sampled environment. A common alternative to enrichment using a well-defined enrichment bacterial strain, or strains, involves inoculating a bacteria-rich environmental sample in medium for a certain length of time, again often overnight depending on the sample. Bacteria present in the sample will propagate and act as hosts for phages in the sample. This represents a middle ground as there will still be some selection of phages that propagate on the cultivatable bacteria, but not necessarily toward one specific strain. This method will increase the phage titer and could therefore be used if a direct plaque assay does not yield sufficient isolates (Auling et al. 1977; Karumidze et al. 2013). Another alternative to the enrichment step during phage isolation from environmental samples is based on phage precipitation. Here, samples are centrifuged to remove debris and filtered through a 0.22 μm membrane to sterilize the solution. Next, ZnCl2 is added, and the solution is incubated at 37  C. During the incubation, the phages precipitate and are then concentrated by centrifugation. The phages are resuspended in a small volume of buffer, and the resulting phage concentrate can then be used for plaque assays. The method has been tested for various soil and plant samples and was reported to outperform an enrichment procedure for the same

442

F. van Charante et al.

samples, resulting in successful phage isolations where the standard enrichment method failed (Czajkowski et al. 2016). Another method for concentrating phages, but without the use of a precipitant, is to use tangential flow filtration (TFF) (Van Twest and Kropinski 2009). TFF has the advantage of being effective at low phage concentrations, such as fresh- or saltwater samples where the phage concentration is too low to permit efficient precipitation. Yet another variant on standard isolation using enrichment and plaque assay has been developed recently involving bacterial hosts on a filter membrane. In this method, the bacterial hosts are fixed on a 0.45 μm filter instead of in agar. The sample containing phages is pretreated by filtration through a 0.22 μm membrane before the phage suspension is run through the bacterial filter. Phages active against the bacterial host will adhere to their filter-bound hosts. Afterward, this filter is used to start enrichment of only the host-specific phages. This method was demonstrated to be effective with small volumes of sample (Ghugare et al. 2016).

Generation of Pure Isolates and Phage Stocks A plaque, being the visual indicator of the presence of a single phage or infected cell that initiated the plaque, is the standard endpoint of phage isolation (see ▶ “Detection of Bacteriophages: Phage Plaques” chapter). That is, for the user, following the initial isolation steps, individual plaques from the culture may then be obtained through diluting the sample and plating it together with the host strain (process number 4 in Fig. 1). This can be done in the form of a plaque assay, where the sample is diluted and aliquoted with the host strain using the double agar overlay method, where the bacteria and sample are suspended in molten soft agar and poured over a standard solid agar plate, followed by cultivation (Adams 1959; Kropinski et al. 2009). Alternatively, the sample is drawn across the surface of solid agar medium and overlaid with a lawn of bacteria in semisolid media before incubation. The effect is similar, and, if phages infect the target bacteria with sufficient efficiency, and phage particles can migrate through the agar rapidly enough, then plaques will form that can be collected (Lin et al. 2010). Plaque collection can be by removing a plug of agar from the plaque or by touching the plaque with a sterile inoculation loop or toothpick in case of small plaques. The sample is then usually suspended in buffer. In order to obtain phage stocks that can be reasonably presumed to hold one genetic clone, the original plaque should be subjected to at least three rounds of plaque purification (process number 5 in Fig. 1), in which the sample is serially diluted and overlaid or struck out and overlaid as described above (Carlson 2005). It is most common to perform the rounds of purification by using an inoculation loop and then to pick the plug of agar from the final, purified plaque and suspend it in buffer. The resulting phage suspension is considered to be a pure phage suspension and is ready to be used to generate phage lysate (process 6 in Fig. 1) or for any downstream purpose. Usually, a small volume of lysate is prepared first, as discussed below.

Isolation of Bacteriophages

443

Fig. 3 Screening of environmental samples that contain phages infecting Agrobacterium sp. Plaques were prepared by means of the double agar overlay infused with dilutions of three phages, each with a distinct plaque morphology. The black arrows indicate clear plaques of 2 mm diameter, the white arrows indicate plaques that are characterized by the formation of a halo (see below), and the red arrows show small, turbid plaques

During the process of plaque purification, it is worth noting the morphology of all plaques because uniformity of plaque morphology is an indicator of a successful purification. As can be seen in Figs. 3 and 4, by way of example, distinct plaque morphologies (Fig. 3) indicate insufficiently purified plaques, whereas highly similar morphologies (Fig. 4) typically, although certainly not always, indicate a monoclonal phage population.

Storing Isolated Phages Once purified, it is advisable to prepare phage stocks for storage and ongoing use of the isolated phage. Generic protocols are typically employed when preparing initial stocks, but depending on the resulting phage titer and other nuances of host or phage growth conditions, small-scale propagation can help in optimizing a phage-/hostspecific amplification protocol. Once a lysate has been obtained with sufficient titer, the virions in the crude lysate can then be purified in several ways to remove bacterial debris and spent medium. Purification can enhance the stability of the phage solution and remove unwanted components of the lysate such as exogenous DNA, endotoxins, bacteriocins, etc. Density gradient ultracentrifugation is generally considered as the gold standard to purify phage solutions (Boulanger 2008). However, this process is cumbersome and not scalable for high volume needs. Cheaper alternatives include anion-exchange chromatography (Adriaenssens et al. 2012) or precipitation of phages using polyethylene glycol (PEG) prior to high-speed centrifugation (Yamamoto et al. 1970). Although in most cases phage stocks are stored as

444

F. van Charante et al.

Fig. 4 Uniform plaque morphology seen after plaque purification of a clear, 2 mm plaque (black arrows in Fig. 3). Although this plate was prepared by the plaque assay method, note that the overlapping plaques mean that this plate could not be used for titering purposes (see ▶ “Detection of Bacteriophages: Statistical Aspects of Plaque Assay” chapter). However, isolated plaques (white arrows) are available for the purposes of plaque purification

crude lysates, only centrifuged and/or filtered in the spent medium in which they were produced.

Where to Hunt for Phages While selecting the proper method for phage isolation increases the likelihood of finding desired phage, the source is equally important. Generally, the best place to find phages is the environment or niche in which their host can be found. Nevertheless, finding phages can sometimes be more complicated than expected. Owens et al. (2013), for example, originally intended to isolate Campylobacter jejuni phages from poultry kept indoors, but were unsuccessful. Phages were successfully obtained, however, when they attempted the same isolation from free-range chickens. The exact reason for this discrepancy was not identified. Although the bacteria are generally more likely to be present in free-range poultry compared to poultry kept indoors (Heuer et al. 2001), both indoor and free-range flocks were reported to be colonized by C. jejuni (Owens et al. 2013). Similarly, other researchers have isolated C. jejuni phages from free-range (Loc Carillo et al. 2007) and organic poultry (El-Shibiny et al. 2005). Evidently, subtle differences in sampling location can greatly impact phage isolation success, emphasizing the value in sampling a variety of locations and/or conditions, particularly when the ecology of the target phage and host bacteria is not well understood. In this section we walk through the issue of isolation sources for a variety of circumstances and phage isolation goals.

Isolation of Bacteriophages

445

Isolation from the Environment Bacteriophages are most often isolated from the same source as the host, typically, soil, sewage, or water samples. Generally, isolation techniques from such samples are quite similar with some small differences. Sewage and water samples are first clarified and often filtered to get rid of unwanted debris and bacterial contamination (Maal et al. 2015; Uchiyama et al. 2008; Van Twest and Kropinski 2009). This is sometimes followed by an enrichment step as described above, especially if phages are desired against specific bacterial strains such as clinical isolates (Anand et al. 2015; Pootjes et al. 1966; Sillankorva 2018). Next, the samples can be tested against the hosts of interest using double agar overlays (Czajkowski et al. 2014). Soil and water samples can be directly incubated with the desired host to allow phages to amplify and be subsequently isolated (Rombouts et al. 2016). As phages can adhere to soil particles, it can be advantageous to extract the viruses from these particles. This can be done through a variety of methods. One popular procedure makes use of sterile water or a buffer solution in which the sample is incubated overnight to free phages from the soil particles. Low-speed centrifugation is used to remove the environmental debris, and the supernatant is filtered (Adriaenssens 2011; Barnet 1972). During the overnight incubation step, it is advisable to gently shake the mixture to facilitate the extraction of the phages into the buffer (Salifu et al. 2013; Yehle and Doi 1967). Enrichment and then plating follow, as above.

Common Sources for Isolation of Therapeutic Phages When it comes to phages to be used for the treatment of human patients, the most common sources are sewage or infected individuals, depending on the bacterial host. Often, however, such sources do not always provide the most suitable phages (Łobocka et al. 2014), and sewage contains high concentrations of bacteria, especially those that are present in the alimentary canal of humans, and as such also contains a high concentration of potentially medically relevant phages. A good example of phage isolation from sewage can be found in Tbilisi, Georgia, where river water contaminated with hospital sewage has been used to successfully isolate therapeutic phages against Klebsiella bacteria (Karumidze et al. 2013), among many other bacterial host species and strains (Parfitt 2005). With respect to isolation from sewage, the location where the sample is taken can influence which phages are obtained. Generally, sewage samples are taken at either waste treatment plants (Kwiatek et al. 2017; Markel and Eklund 1974; Wittmann et al. 2014) or directly at the source of the sewage (Anand et al. 2016; Cao et al. 2015). Attention should be paid to the nature of your municipality’s sewage treatment process. For a mixed system that includes stormwater runoff, it will be advantageous to collect samples during dry periods to isolate human rather than soil-relevant phages. It is also recommended to collect samples from after the pretreatment sedimentation process for ease of handling and before the primary treatment process to maximize microbial diversity.

446

F. van Charante et al.

The ecology of the target host should also be considered, especially for ubiquitous bacterial species. Acinetobacter baumannii, for example, is found in a great diversity of environments. But although phages for this species can be readily isolated from these diverse environments, such as marine sediment (Yang et al. 2010), far more therapeutically relevant phages have been isolated from clinical samples and hospital waste (Lin et al. 2010; Popova et al. 2012; Zhou et al. 2018). Similarly, when phages against antibiotic-resistant bacteria are desired, many authors have used hospital sewage as their source (e.g., Cao et al. 2015, Ceyssens et al. 2008, and Nivas et al. 2015). Difficulties with isolation from hospital sewage have also been reported, which were hypothesized to be due to high concentrations of disinfectants present in this source (Melo et al. 2014). It is also unclear whether the effort involved in obtaining hospital rather than municipal sewage results in phages that are better adapted to antibiotic-resistant bacteria. Another common source of phages for human treatment are the infected patients themselves. As these patients have the target bacterium, there is also a possibility that phages are also present that infect these pathogens. Samples are taken from a large variety of sources, depending on the target bacterium, and can range from stool samples and saliva (Bachrach et al. 2003; Machuca et al. 2010) to skin or throat swabs (Chang et al. 2015; Steinberg et al. 1976) depending on context. Even dental plaques have been successfully used for phage isolation (Tylenda et al. 1985; Yueng and Kozelsky 1997). As expected, stool samples are taken for bacteria responsible for gastrointestinal diseases such as E. coli or Salmonella. As bacteria and phage concentrations in such samples are generally high, these samples can be used for direct plaque assays (Chibani-Chennoufi et al. 2004; Cornax et al. 1994). Besides isolations for specific bacteria, indicator strains can also be used. E. coli strain K803 is known to be highly susceptible to coliphages, making it an excellent strain for general phage isolation (Chibani-Chennoufi 2004). However, use of such strains will nonetheless bias the isolated phages toward those best suited to growth on that indicator strain, especially if there are enrichment steps prior to isolation. With respect to phage isolation for phage therapy for animals and plants, the same principles apply: phages will be most easily isolated from places where their host resides. For animals this is often at farms and especially the sewage from farms (see ▶ “Bacteriophage Utilization in Animal Hygiene” chapter). However, not only sewage but fecal samples, animal bedding, intestine samples, and the surrounding soil have been used as well (Augustine et al. 2013; Burchard and Dworkin 1966; Endersen et al. 2013; Verma et al. 2013). Similarly, phage isolation for crop uses typically starts with plant or soil samples (see ▶ “Crop Use of Bacteriophages” chapter).

Isolation from Lysogens Temperate phages have the ability to insert their genetic code into their host’s genome or to form stable extrachromosomal elements such as plasmids or episomes (see ▶ “Temperate Phages, Prophages, and Lysogeny” chapter). In both cases, they

Isolation of Bacteriophages

447

can reside as a prophage with their hosting bacterial lysogen for many generations. These prophages, as found within intact, isolated bacterial lysogens, can act as sources of phages as well. Obtaining these phages, though, typically will require some form of prophage induction. The induction of temperate phages can occur spontaneously, that is, without any inducing agent. However, this is usually ineffective as many (often the majority) of bacterial cells are not induced, resulting in low titers and minimal diversity, which can be an issue depending on the purpose of the isolation. Low phage titers may be less suitable for ongoing analysis such as DNA sequencing, and low diversity reduces the chances of isolating phages with the desired property (unless this property is spontaneous induction itself) or of providing a complete picture of the prophage diversity in the original sample. Other, more reliable strategies to induce lysogens are based on directly damaging their DNA by exposing the bacteria to UV light or to sublethal concentrations of DNA-damaging chemicals such as fluoroquinolones or mitomycin C (Bondy-Denomy et al. 2016; Lamont et al. 1989). Care should be taken to optimize these strategies so that the inducing agent induces prophages but does not otherwise kill the lysogenic cells before they can produce phages (Carlson 2005; Raya and Hébert 2009). After prophage induction, there are several ways to isolate the phages. Most commonly, the induced lysate can be plated with an indicator strain for plaque purification as described above. Alternatively, where superinfection immunity (see “Glossary”) does not occur, the lysate can be plated with the lysogen under non-inducing conditions to yield plaques, although these plaques may be very turbid and difficult to observe.

Isolation from Diverse Environments Phages have not only been isolated from common, relatively mesophilic sources like soil or sewage but also from other, more extreme environments like hot springs, glaciers, and saltwater lakes. Phage isolation from such environments is similar to regular phage isolation. However, special care should be taken to mimic the conditions of the source in order to properly isolate phages as the environmental conditions dictate the growth of bacteria and therefore the phage. For example, bacteriophages that propagate at cold temperatures were initially identified from the marine environment by Spencer (1955). Isolation was successfully carried out using media based on seawater with the addition of Lemco broth and peptone, with cultivation at 20  C. Similarly, cyanophages can be isolated from marine environments using artificial seawater media (Millard 2009). A detailed study of the temperature range of psychrophilic phages, comparing psychrotrophic and mesophilic phages of Pseudomonas, was performed by Olsen et al. (1968). They isolated phages from sewage enriched at 20–25 and 37  C. Both higher and lower growth temperature phages were used to infect mesophilic and psychrotrophic Pseudomonas, and the resulting infections were compared. The coldactive phages were able to infect both the psychrotrophic and mesophilic hosts, yet

448

F. van Charante et al.

were not able to propagate beyond 32  C and showed larger plaques at lower temperatures (Olsen et al. 1968). Delisle and Levin (1969) then used seawater to isolate Pseudomonas phages, with similar results. Psychrotrophic phages were plated at 2 and 20  C, and plaque size was reported to be significantly larger for the 2  C infections. For one phage, infection was possible at 2  C but not detected at 20  C (Delisle and Levin 1969). These results demonstrate that although phages can be isolated without closely mimicking the specific ecology of the target bacterium, there is a risk of preferentially isolating phages less well adapted to that specific ecology. More recently, psychrophilic bacteriophages have been isolated from more extreme environments, such as sea ice (Luhtanen et al. 2014), glaciers (Ji et al. 2015; Li et al. 2016), and semi-frozen sea ice slurry (Wells and Deming 2006). As above, media were chosen to be representative of the sampling environment. The glacier samples were enriched prior to plating, and the sea ice samples were concentrated by centrifugation. The major difference between these two approaches was the incubation temperature: ranging from 3 to 10  C for the sea ice isolation and 15  C for the glacier isolation (Ji et al. 2015; Li et al. 2016; Luhtanen et al. 2014). Since agar-based methods are not compatible with temperatures below 0  C (when the medium freezes solid and is no longer a gel), new techniques were developed making use of silica gel which allows plaque formation at 1  C. This technique was reported to yield approximately five times the number of plaques compared to agarbased plating at higher temperatures (Dietz and Yayanos 1978; Wells and Deming 2006). The same basic techniques, with modifications to the media and culture conditions, are also used to isolate thermophilic phages and archaeal viruses (Fulton et al. 2009). Many isolation protocols for viruses of extremophiles have been developed for Archaea (e.g., Fulton et al. 2009 and Rice et al. 2001). As expected, archaeal viruses are isolated by adapting standard protocols to the growth conditions of the host. In the case of thermophiles, gellan gum (Gelrite) gels can be used instead of agar due to their higher melting temperature than agar gels (Stedman et al. 2009). However, archaeal viruses are not bacteriophages (Abedon and Murray 2013), and while there are similarities in their isolation protocols, they will not be considered further in this review. As with the cold-active phages, the existence of thermophilic phages was demonstrated early in the history of phage research, with the first report in 1926 (Koser 1926). Since then, hyperthermophilic phages have also been isolated, the first being against Thermus thermophilus. The phage was able to propagate at up to 78  C (Sakaki and Oshima 1975). Hot springs are a common source of such phages as these waters contain thermophilic or hyperthermophilic bacteria (Hjorleifsdottir et al. 2014; Lin et al. 2011; Yu et al. 2006). Other sources include geothermal vents (Liu et al. 2006) and compost (Cheepudom et al. 2015; Marks and Hamilton 2014). Generally, no major modifications to the isolation methodology are required aside from variations in media and incubation temperature (Hjorleifsdottir et al. 2014; Sakaki and Oshima 1975; Yu et al. 2006). Agar plates can still be used for these phages as the incubation temperature is not sufficient to warrant the use of Gelrite

Isolation of Bacteriophages

449

plates as is necessary for some archaeal viruses (Rice et al. 2001). What is of note, however, is the early observation that the temperature tolerance of thermophilic phages can depend on the solution in which they are suspended. During the first hyperthermophilic phage isolation by Sakaki and Oshima (1975), it was noted that the maximum stability of the phage particles at high temperatures was achieved when the viruses were suspended in either the growth medium, HB8 broth, or in the water from the hot spring itself, while buffered solutions showed varied but rather low stability. Therefore, it was suggested that some environmental factor influenced the stability of the phage particles at these high temperatures. Hence it is relevant to underline the importance of keeping the culture conditions close to those of the environment from which the phages are isolated (Sakaki and Oshima 1975). Similarly, the same techniques can be used for the isolation of other challenging phages, like those for extremely slow-growing bacteria. Phages against Mycobacterium avium, with a doubling time of 24 h, were isolated using enrichment and the double agar overlay method. The major differences here compared to normal phage isolation were incubation times, as one would expect. Notably, it is sometimes possible to isolate and culture phages against difficult to culture bacteria with an alternative but closely related strain, such as (in the case of M. avium and M. tuberculosis) M. smegmatis (Basra et al. 2014). These examples all show that phages can be isolated from diverse environments and for difficult to culture hosts using standard isolation protocols, though it is important to mimic the natural conditions from which the phage was isolated to ensure the stability and successful propagation of the phage particles. Difficulties can, however, arise when the conditions for cultivation are not conducive to the use of standard materials, as is the case with some extreme environments like very high or low temperatures. This overall message should also be borne in mind when isolating phages from other extreme environments not discussed in detail here, such as high salinity (Muruga et al. 2013) or high acidity (Ward et al. 1993), all of which may need to be considered when choosing media, culture conditions, gelling agent, etc.

Resulting Phage Phenotypes Certain phage phenotypes that are easily observable via plaques can be highly relevant to phage isolation. The first is phage host range, which although it can be easily manipulated (Burrowes et al. 2019; Mapes et al. 2016) should nonetheless be selected for in the course of phage isolation schemes (Mattila et al. 2015; Sillankorva 2018). Second is the morphology of resulting phage plaques, which can be used as a first-approximation means of distinguishing among phage isolates, particularly toward increasing the diversity of phages that are isolated from a single environment. Isolating multiple phages possessing the same plaque morphology, that is, likely will supply lower phage diversity than isolating multiple phages possessing different plaque morphologies. Furthermore, plaque morphology can be used to infer

450

F. van Charante et al.

phenotypic and physical traits of the isolated phages, such as their size, life cycle, and ability to degrade bacterial polymers.

Altered Procedures and Their Effect on Host Range Multiple factors influence the ability of a phage to infect a potential host, such as the specificity of the phage receptor binding proteins, the presence of related prophages, and bacterial resistance mechanisms (discussed in detail by Hyman and Abedon (2010) and Ross et al. (2016)). Hence, phages are often described as having a very narrow host range. However, over the years, broad host range phages with the ability to infect multiple strains of the same bacterial species (Anand et al. 2015; Vinod et al. 2006), multiple species (Greene and Goldberg 1985; Mirzaei and Nilsson 2015), and even multiple genera (Paolozzi and Ghelardini 2006) have been described. Often these phages are discovered serendipitously when isolated phages are tested against multiple hosts. For example, Greene and Goldberg (1985) isolated phages from a large array of soil samples by enrichment with Streptomyces avermitilis; however, some isolated phages were also able to infect other Streptomyces species, namely, S. venezuelae, S. parvulus, and S. azureus (Greene and Goldberg 1985). More recently, methods have been developed specifically to isolate broad host range phages. The simplest way of generating broad host range phages is by using multiple hosts during the isolation protocol. Jensen et al. (1998) were one of the first groups to experiment with the isolation of broad host range phages. They use in their enrichment step two different bacterial hosts simultaneously and were able to isolate phages that infect both P. aeruginosa and E. coli, as well as Sphaerotilus natans. However, not all attempts proved to be successful (Jensen et al. 1998). Ross and colleagues also found host range broadening when they isolated bacteriophages using two strains of Enterococcus faecalis as the isolation hosts (Ross et al. 2016). More recently, an alternative method has been proposed in which samples are incubated with different potential hosts sequentially in both solid and liquid medium (Yu et al. 2016). Using solid media, the phages were first propagated with the isolation strain (host 1) starting from a mixed phage stock derived from an environmental sample. Next, all the plaques were collected, pooled, enriched using a taxonomically distinct strain (host 2), and subsequently plated with host 2. This process was repeated with two other hosts. In theory, phages that can form plaques with the final host should be able to form plaques with all previous hosts. Similarly, using liquid medium, a phage stock is added to the first host and given time to adsorb. Unbound phages are washed away, and the adsorbed phages are allowed to propagate on host 1. This procedure is repeated for a number of hosts until the sample is plated on the final host strain. Again, only phages that are able to lyse the final host should be able to infect all the different hosts. Both methods produced a cocktail of phages, and plaque-based isolation was still needed to identify individual broad range phages able to infect a number of Pseudomonas species as well as some phages also able to infect E. coli (Yu et al. 2016).

Isolation of Bacteriophages

451

Plaque Morphology Indicates Diverse Phage Characteristics A plaque can be viewed as the macroscopic representation of a bacteriophage. It is the three-dimensional clearing in a bacterial lawn initiated by an infectious bacteriophage or a phage-infected bacterium leading to a reduction in the surrounding cell density within a bacterial lawn (see ▶ “Detection of Bacteriophages: Phage Plaques” chapter). In the formation of a plaque, four phases can be defined: adsorption of the viral particle, first rounds of viral replication, enlargement phase of the lysis zone, and, finally, maturation of the plaque (Abedon and Yin 2008, 2009). As phages are highly diverse and plaque morphology is intrinsically related to the nature of a phage, the morphology of a plaque can give clues to properties of the phage itself. Indeed, during a standard isolation process, plaque morphology will be the first observable property of any isolated phage. It should be noted that external factors such as growth medium, osmolarity, agar type and concentration (Bronfenbrenner and Korb 1925; Elford and Andrewes 1932), thickness of the agar layers, incubation time and temperature, humidity, and cell inoculum (both cell number and physiological state) among others can affect plaque morphology and should therefore be kept as constant as possible if meaningful comparisons are to be made. Nevertheless, different features can be considered when looking at plaque morphologies, such as plaque size, clarity, and the presence of halos. However, no single plaque feature can be thought of as an absolute indicator of a phage’s physiological or genetic properties, but plaque morphology can be useful to the user as a rough approximation of the possibility of certain traits. Further study will always be required to determine any specific properties of a given phage. Nevertheless, plaque morphology can be especially useful when working with otherwise clonal phage populations, where (as long as the environmental conditions are held constant) different plaque morphologies can be a useful way to detect mutant or contaminant phages. This was demonstrated when a temperate M. smegmatis phage that produced turbid plaques on M. abscessus was engineered to be virulent for use in treating a patient with disseminated, terminal M. abscessus infection (Dedrick et al. 2019). After removing the phage’s repressor gene, the modified phage produced larger, clearer plaques on M. abscessus while also increasing the efficiency of plating ~100-fold. This modified phage was included in a phage cocktail that was used to successfully treat the patient. The size of a plaque depends on a range of variables, both intrinsic to the phage and its host, as well as being environmentally determined, as mentioned above. Intrinsic factors such as virion size, shape and hydrophobicity, and physiological or genetic properties like the receptor affinity, latent period, and burst size can all have implications for plaque size. The rate of phage adsorption to its host is determined by the binding kinetics of the phage and its bacterial receptor (see chapters ▶ “Detection of Bacteriophages: Statistical Aspects of Plaque Assay” and ▶ “Phage Infection and Lysis” chapters). When adsorption rates are low, phages will tend to infect less efficiently and hence can diffuse further from their point of origin before infecting a host cell, resulting in larger plaques. Conversely, high adsorption rates can have a negative influence on plaque size since the phages do not diffuse far before binding

452

F. van Charante et al.

Fig. 5 Differing plaque morphologies on a lawn of Agrobacterium sp. Phages that produce clear plaques (right panel) are typical strictly lytic phages that can easily infect the host of interest. Turbid plaques (left panel) on the contrary are an indication that the infection is not efficient due to inefficient adsorption, infection, or lysogeny

to their receptor (Abedon and Yin 2009). The second step in plaque formation, the enlargement phase of a plaque, is influenced by the speed by which the phage particles diffuse through the agar medium. Both internal and external factors influence the diffusion of phages. Among the intrinsic factors of the phage particles are their size and the burst size of the phage, which, to a lesser extent, can influence the expansion of the plaque. Smaller particles can diffuse more easily, generating larger plaques, while large particles have a harder time getting through the agar matrix. The clarity of a plaque gives an indication of the efficiency at which a phage can lyse its host, meaning that a clear plaque indicates efficient host lysis, while a turbid plaque can be an indication of inefficiency since not all the bacteria are lysed within the plaque (Abedon and Yin 2008) (Fig. 5). The efficiency of phage infection depends on the overall efficiency of the individual steps involved in the infection, such as adsorption, host takeover, and lysis. In the case of temperate phages, a certain population of infected cells will become lysogens (Fortier and Sekulovic 2013) (see ▶ Temperate Phages, Prophages, and Lysogeny chapter) that will continue to grow, resulting in a turbid plaque (Abedon and Yin 2008), and superimmunity means that lysogenic cells will not be reinfected. Hence, turbid plaques may be an indication of a temperate phage, especially when the center of the plaque is the most turbid; the superimmune lysogens in the plaque center will have more time to grow than those at the periphery. However, a turbid plaque is neither an especially specific or sensitive indicator of a phage being temperate. Indeed, many temperate phages have either low or variable rates at which they undergo a temperate life cycle and may produce visibly clear plaques as a result (Bertani 1951). Further studies, usually by sequencing the phage or the lysogen, are required. Another important feature of plaque morphology is the formation of a halo. A halo is a region around the plaque center that is less turbid than the surrounding lawn but more turbid than the plaque center (see Fig. 3 (white arrows) and Fig. 6). Halos are usually clearly demarcated and are not simply a gradation of turbidity at the plaque edge. Halos are often correlated with the presence of exopolysaccharide

Isolation of Bacteriophages

453

Fig. 6 Phages that have a depolymerase activity typically form a halo around the plaque. Here, Agrobacterium sp. phages produce plaques in which a clear lysis zone (white arrows) is surrounded by an expanding halo (black arrows), indicating the presence of EPS-degrading enzymes

(or EPS) depolymerases. These enzymes may be encoded by the phage to degrade external matrices like biofilms in which the bacteria are embedded as a strategy to reach their phage receptor (Yan et al. 2014; Pires et al. 2016), but they can also be induced from the host cells (discussed in Harper et al. 2014). In this respect, depolymerases can be part of the capsid of the phage, e.g., tail spikes of Pseudomonas putida phage φ15 and AF have been shown to have depolymerase activity (Cornelissen et al. 2011, 2012). The halo itself is the result of the depolymerization of the EPS, which increases the lawn transparency. This degradation can occur after the maturation of the plaque: while the boundaries of a plaque are restricted by the stationary growth of the surrounding bacteria, the diffusion of phage particles and unbound EPS depolymerases is not and can expand after the plaque is formed (Cornelissen et al. 2011, 2012). However, not every phage-containing EPS-degrading depolymerase forms this typical halo structure, e.g., Salmonella phage phi PVP-SE1 (Santos et al. 2009).

Phage Isolation for Biotechnology One important biotechnological phage application is their use as antibacterial agents, e.g., as for phage therapy. This too can require phage isolation approaches which bias toward certain phage phenotypes. So, too, it can be helpful when seeking phages with certain properties to isolate and characterize large numbers of phages, as can be achieved via high-throughput approaches.

454

F. van Charante et al.

Implications for the Use of Phages as Antimicrobials One important characteristic of a phage intended for therapeutic use is an exclusively virulent life cycle (Łobocka et al. 2014; Loc-Carrillo and Abedon 2011; Pirnay et al. 2018). Temperate phages can encode for often poorly characterized pathogenicity islands, toxins, and other virulence factors that could potentially cause harm to patients when expressed. Indeed, the evolution of negative traits in many bacterial pathogens, from antibiotic resistance to virulence itself, is linked to genetic material carried by temperate phages (Fortier and Sekulovic 2013). Even without the concern of introducing harmful genetic material to a patient’s microbiota directly through treatment, the ability of temperate phages to mediate horizontal gene transfer from virulent target pathogens to bystander bacteria by transduction is another notable concern (Kutter et al. 2010) that may be variable among temperate phages (Krylov et al. 2012). Other selectable factors that can be broadly important for the suitability of an isolated phage for phage therapy include ease of amplification, ability to infect a minimally virulent production host, in vivo persistence, ability to adsorb under in vivo conditions, and ability to productively infect under the metabolic conditions found in vivo (Łobocka et al. 2014; Merabishvili et al. 2009). Any practical application for phage antimicrobial activity will benefit from maximally practical amplification of high-titer stocks, which can be selected for by using an enrichment step that mimics manufacturing conditions as much as possible. Similarly, especially phage therapy for humans can require phages to be manufactured in minimally virulent and minimally lysogenic bacterial strains (Pirnay et al. 2015), and it will be important to select for the ability to infect such a host early on in the isolation process. Another important quality is the ability to persist in the environment being treated, whether it is the dry surface of a plant exposed to the sun or an inflamed human tissue, which can be selected for by exposing environmental samples to similar conditions before or during isolation. Additionally, many applications being considered for phage therapy contain factors known to interfere with phage adsorption, such as milk components during treatment of bovine mastitis (Gill et al. 2006; O’Flaherty et al. 2005), resistance to which can be selected for by enriching for phages in milk media (Breyne et al. 2017). However, it is becoming increasingly clear that more complex factors, like the ability to infect bacteria in specific metabolic states, may be advantageous to select for. Indeed, very little is currently known about the ability of bacteriophages to infect anaerobically growing bacteria (Ceyssens et al. 2010; Kutter et al. 1994) or bacteria in stationary phase (Bryan et al. 2016), despite their relevance to many wound systems. Finally, care should be taken that phages are not evolved away from the environmental conditions of phage therapy. As phages cultivated in Erlenmeyer flasks over multiple generations will evolve toward optimization for those conditions, they may become less effective when used for phage therapy as the environment on or in a patient is quite different. This can easily be avoided by creating phage seed lots

Isolation of Bacteriophages

455

directly after isolation which are used to generate phages for therapeutic or experimental use.

High-Throughput (HTP) Phage Isolation Even when a phage discovery approach is exquisitely designed, it is still unlikely that all or even most phages isolated will be those best suited for the downstream purpose. And, as noted above, some host species and strains are far more difficult to isolate phages for than others (Mattila et al. 2015). So it is often still necessary to isolate and analyze many phages or carry out many failed attempts before the most suitable candidate(s) is found. This is now far less true when it comes to many ecological studies that use HTP DNA sequencing rather than phage isolation per se (e.g., Hurwitz and Sullivan 2013; Reyes et al. 2012), where the considerations turn to molecular concerns relating to DNA amplification and sequencing biases rather than sampling or enrichment biases (e.g., Ercolini 2013; Solonenko and Sullivan 2013). One approach to resolving the sampling biases is to simply power through them with sheer throughput of isolation and analysis to identify the most commercially viable isolates, so throughput itself is a valuable objective, especially commercially. Robotic liquid handlers now offer the potential to carry out isolation and analysis of individual phages on a much larger scale than is feasible by hand. Microtiter platebased isolation approaches (Harper and Blake 2018; Xie et al. 2018) when combined with liquid handlers can process all or most of the steps required for phage isolation once a sample has been obtained, and they can do this continuously, with high repeatability, and with minimal operator input. Besides HTP liquid handlers, there are HTP incubators (Henry et al. 2012), and HTP image analysis can also be used to increase the rate and accuracy of analysis of phage plaques (Yakimovich et al. 2015). Even incubation times could be greatly minimized using techniques that are currently being explored and commercialized for rapid bacterial diagnostics (e.g., Kang et al. 2019, Li et al. 2017, and Lu et al. 2013) while also obviating the need for plaque-based identification of candidate phages. Even if agar-based plaque isolation is required, automated colony pickers are already available and could be combined with plaque analysis algorithms to enable rapid plaque purification. It is still early days for HTP phage isolation, although commercial interest is already present. For example, in July 2018, Locus Biosciences, Inc. acquired EpiBiome, Inc. in order to obtain their HTP phage discovery technology for isolation and analysis of therapeutic phages (Locus Biosciences 2018). Rapid phage analytical approaches are also being commercialized (e.g., Adaptive Phage Technologies 2018 and Henry et al. 2012). HTP phage isolation allows not only for rapid isolation and analysis but also for precise control of environmental parameters with high repeatability. Combined with data on the suitability of phages for their desired purpose, HTP techniques could and most likely will be used to optimize all steps of phage isolation in the future.

456

F. van Charante et al.

Conclusions The isolation of phages, either for specific purposes such as therapy or to answer broader questions about their biology and diversity, is an exercise that rewards ingenuity in a way that is presently under-explored. Indeed, if phages are being isolated for their antimicrobial activity, then the clever adaptation of the source used for isolation, the growth media, and many other aspects of the isolation procedure will allow for the preferential selection of properties most relevant to the pathology being addressed. If, for example, phages are needed that will replicate in the human bladder or on the leaves of leek plants, it is naturally wise to use environmental samples that will contain phages that already succeed in these niches. It will also be wise to incorporate the challenges they will face, such as low pH and uric acid stress, or UV irradiation and dehydration, as selective pressures in enrichment pathways that are specially adjusted to the purpose in mind. This will allow the isolation and investment of effort into phages that are more likely to be useful than a more randomly selected isolate. In the same way, depending on the questions they would like to ask, a researcher working to isolate a study object can manipulate the conditions of their isolation process either to obtain the phages that best represent the diversity of an environmental sample or to bias for phages with any number of specific properties or phenotypes. For example, high titers from either liquid or solid media amplification could be selected for from a mixed population of phages through enrichment in these media, diluting the lysate to extinction and only picking from plates with the fewest plaques. Long-term persistence in either a laboratory refrigerator or a field condition could be selected for by storing a mixed culture enrichment in these circumstances for an extended period of time prior to isolation. Phage isolation is a critical step in many phage experiments. When it comes to biotechnological applications, it can be argued that isolation is often the critical step. Phages are now being explored for commercial uses beyond therapy. Despite this need, general techniques used for most commercial and research phage isolation appear to be based on convenience rather than a focused process, optimized to bias the output phage heavily toward the desired properties: host range, which can be either broad or very narrow; temperate vs virulent; high titer; adaptation to in situ conditions, and even, conceivably; more commercializable properties such as ease of formulation or purification could be selected during enrichment and isolation steps. The added power of high-throughput isolation, sequencing and annotation, and other HTP analysis will only further facilitate our ability to rapidly isolate and identify phages with enhanced suitability for any given objective.

Cross-References ▶ Bacteriophage Discovery and Genomics ▶ Phage Infection and Lysis ▶ Bacteriophage Utilization in Animal Hygiene

Isolation of Bacteriophages

457

▶ Crop Use of Bacteriophages ▶ Detection of Bacteriophages: Phage Plaques ▶ Detection of Bacteriophages: Statistical Aspects of Plaque Assay ▶ Temperate Phages, Prophages, and Lysogeny ▶ The Discovery of Bacteriophages and the Historical Context

References Abedon ST, Culler RR (2007) Optimizing bacteriophage plaque fecundity. J Theor Biol 249 (3):582–592 Abedon ST, Murray KL (2013) Archaeal viruses, not archaeal phages: an archaeological dig. Archaea 2013:1–10 Abedon ST, Yin J (2008) Impact of spatial structure on phage population growth. In: Abedon ST (ed) Bacteriophage ecology population growth, evolution and impact of bacterial viruses. Cambridge, UK: Cambridge University Press, pp 94–113 Abedon ST, Yin J (2009) Bacteriophage plaques: theory and analysis. In: Clokie MR, Kropinski AM (eds) Bacteriophages: methods in molecular biology, vol 1. Humana Press, New York, pp 161–172 Abedon ST, Hyman P, Thomas C (2003) Experimental examination of bacteriophage latent-period evolution as a response to bacterial availability. Appl Environ Microbiol 69(12):7499–7506 Ackermann HW (2011) The first phage electron micrographs. Bacteriophage 1(4):225–227 Adams MH (1959) Bacteriophages. Interscience, New York Adaptive Phage Technologies (2018) The Science. http://www.aphage.com/the-science/. Accessed on 10th May 2019 Adriaenssens EM, Ceyssens PJ, Dunon V, Ackermann HW, Van Vaerenbergh J, Maes M, De Proft M, Lavigne R (2011) Bacteriophages LIMElight and LIMEzero of Pantoea agglomerans, Belonging to the “phiKMV-Like Viruses”. Appl Environ Microbiol 77(10):3443–3450 Adriaenssens EM, Lehman SM, Vandersteegen K, Vandenheuvel D, Philippe DL, Cornelissen A, Clokie MRJ, García AJ, De Proft M, Maes M, Lavigne R (2012) CIM® monolithic anionexchange chromatography as a useful alternative to CsCl gradient purification of bacteriophage particles. Virology 434(2):265–270 Anand T, Vaid RK, Bera BC, Barua S, Riyesh T, Virmani N, Yadav N, Malik P (2015) Isolation and characterization of a bacteriophage with broad host range, displaying potential in preventing bovine diarrhea. Virus Genes 51(2):315–321 Anand T, Vaid RK, Bera BC, Singh J, Barua S, Virmani N, Rajukumar K, Yadav NK, Nagar D, Singh RK, Tripathi BN (2016) Isolation of a lytic bacteriophage against virulent Aeromonas hydrophila from an organized equine farm. J Basic Microbiol 56(4):432–437 Augustine J, Louis L, Varghese SM, Bhat SG, Kishore A (2013) Isolation and partial characterization of ΦSP-1, a Salmonella specific lytic phage from intestinal content of broiler chicken. J Basic Microbiol 53(2):111–120 Auling G, Mayer F, Schlegel HG (1977) Isolation and partial characterization of normal and defective bacteriophages of gram-negative hydrogen bacteria. Arch Microbiol 115(3):237–247 Bachrach G, Leizerovici-Zigmond M, Zlotkin A, Naor R, Steinberg D (2003) Bacteriophage isolation from human saliva. Lett Appl Microbiol 36(1):50–53 Barnet YM (1972) Bacteriophages of Rhizobium trifolii I. Morphology and host range. J Gen Virol 15(1):1–15 Basra S, Anany H, Brovko L, Kropinski AM, Griffiths MW (2014) Isolation and characterization of a novel bacteriophage against Mycobacterium avium subspecies paratuberculosis. Arch Virol 159(10):2659–2674 Bertani G (1951) Studies on lysogenesis I. The mode of phage liberation by lysogenic Escherichia coli. J Bacteriol 62:293–300

458

F. van Charante et al.

Bondy-Denomy J, Qian J, Westra ER, Buckling A, Guttman DS, Davidson AR, Maxwell KL (2016) Prophages mediate defense against phage infection through diverse mechanisms. ISME J 10:2854–2866 Boulanger P (2008) Purification of bacteriophages and SDS-PAGE analysis of phage structural proteins from ghost particles. In: Clokie MR, Kropinski AM (eds) Bacteriophages: methods and protocols, vol 2. Molecular and applied aspects Humana Press, New York, pp 277–238 Breyne K, Honaker RW, Hobbs Z, Richter M, Żaczek M, Spangler T, Steenbrugge J, Lu R, Kinkhabwala A, Marchon B, Meyer E, Mokres L (2017) Efficacy and safety of a bovineassociated Staphylococcus aureus phage cocktail in a murine model of mastitis. Front Microbiol 8:2348–2348 Bronfenbrenner JJ, Korb C (1925) Studies on the bacteriophage of D’Herelle. J Exp Med 42:483–497 Bryan D, El-Shibiny A, Hobbs Z, Porter J, Kutter EM (2016) Bacteriophage T4 infection of stationary phase E. coli: life after log from a phage perspective. Front Microbiol 7:1391 Burchard RP, Dworkin M (1966) A bacteriophage for Myxococcus xanthus: isolation, characterization and relation of infectivity to host morphogenesis. J Bacteriol 91(3):1305–1313 Burrowes BH, Molineux IJ, Fralick JA (2019) Directed in vitro evolution of therapeutic bacteriophages: the appelmans protocol. Viruses 11:241 Cao Z, Zhang J, Niu YD, Cui N, Ma Y, Cao F, Jin L, Li Z, Xu Y (2015) Isolation and characterization of a “phiKMV-like” bacteriophage and its therapeutic effect on mink hemorrhagic pneumonia. PLoS One 10(1):e0116571 Carlson K (2005) Appendix: working with bacteriophages: common techniques and methodological approaches. In: Kutter E, Sulakvelidze A (eds) Bacteriophages: biology and applications. CRC Press, Washington, D.C., pp 437–487 Ceyssens PJ, Hertveldt K, Ackermann HW, Noben JP, Demeke M, Volckaert G, Lavigne R (2008) The intron-containing genome of the lytic Pseudomonas phage LUZ24 resembles the temperate phage PaP3. Virology 377(2):233–238 Ceyssens PJ, Brabban A, Rogge L, Lewis MS, Pickard D, Goulding D, Dougan G, Noben JP, Kropinski A, Kutter E, Lavigne R (2010) Molecular and physiological analysis of three Pseudomonas aeruginosa phages belonging to the “N4-like viruses”. Virology 405:26–30 Ceyssens PJ, Glonti T, Kropinski NM, Lavigne R, Chanishvili N, Kulakov L, Lashkhi N, Tediashvili M, Merabishvili M (2011) Phenotypic and genotypic variations within a single bacteriophage species. Virol J 8(1):134 Chang Y, Shin H, Lee JH, Park CJ, Paik SY, Ryu S (2015) Isolation and genome characterization of the virulent Staphylococcus aureus bacteriophage SA97. Viruses 7(10):5225–5242 Cheepudom J, Lee CC, Cai B, Meng M (2015) Isolation, characterization, and complete genome analysis of P1312, a thermostable bacteriophage that infects Thermobifida fusca. Front Microbiol 15(6):959 Chibani-Chennoufi S, Sidoti J, Bruttin A, Dillmann ML, Kutter E, Qadri F, Sarker SA, Brüssow H (2004) Isolation of Escherichia coli bacteriophages from the stool of pediatric diarrhea patients in Bangladesh. J Bacteriol 186(24):8287–8294 Cihlar RL, Lessie TG, Holt SC (1978) Characterization of bacteriophage CP1, an organic solvent sensitive phage associated with Pseudomonas cepacia. Can J Microbiol 24(11):1404–1412 Cornax R, Moriñigo MA, Gonzalez-Jaen F, Alonso MC, Borrego JJ (1994) Bacteriophages presence in human faeces of healthy subjects and patients with gastrointestinal disturbances. Zentralblatt Bakteriol 281(2):214–224 Cornelissen A, Ceyssens PJ, T’Syen J, Van Praet H, Noben JP, Shaburova OV, Krylov VN, Volckaert G, Lavigne R (2011) The T7-related Pseudomonas putida phage φ15 displays virion-associated biofilm degradation properties. PLoS One 6(4):e18597 Cornelissen A, Ceyssens PJ, Krylov VN, Noben JP, Volckaert G, Lavigne R (2012) Identification of EPS-degrading activity within the tail spikes of the novel Pseudomonas putida phage AF. Virology 434(2):251–256 Czajkowski R, Ozymko Z, Lojkowska E (2014) Isolation and characterization of novel soilborne lytic bacteriophages infecting Dickeya spp. biovar 3 (‘D. solani’). Plant Pathol 63 (4):758–772

Isolation of Bacteriophages

459

Czajkowski R, Ozymko Z, Lojkowska E (2016) Application of zinc chloride precipitation method for rapid isolation and concentration of infectious Pectobacterium spp. and Dickeya spp. lytic bacteriophages from surface water and plant and soil extracts. Folia Microbiol 61(1):29–33 d’Herelle F (1916) Sur un bacille dysentérique atypique. Ann Inst Pasteur 30:145 d’Herelle F (1917) Sur un microbe invisible antagoniste des bacilles dysentériques. C R Acad Sci Paris 165:373–375 Dedrick R et al (2019) Engineered bacteriophages for treatment of a patient with a disseminated drug-resistant Mycobacterium abscessus. Nat Med 25:730–733 Delisle AL, Levin RE (1969) Bacteriophages of psychrophilic pseudomonads. II. Host range of phage active against Pseudomonas putrefaciens. Antonie Van Leeuwenhoek 35(1):318–324 Dietz AS, Yayanos AA (1978) Silica gel media for isolating and studying bacteria under hydrostatic pressure. Appl Environ Microbiol 36(6):966–968 Elford WJ, Andrewes CH (1932) The sizes of different bacteriophages. Br J Exp Pathol 13 (5):446–456 Ellis LF, Schlegel RA (1974) Electron microscopy of Pseudomonas φ6 bacteriophage. J Virol 14 (6):1547–1551 El-Shibiny A, Connerton PL, Connerton IF (2005) Enumeration and diversity of campylobacters and bacteriophages isolated during the rearing cycles of free-range and organic chickens. Appl Environ Microbiol 71(3):1259–1266 Endersen L, Coffey A, Neve H, McAuliffe O, Ross RP, O’Mahony JM (2013) Isolation and characterisation of six novel mycobacteriophages and investigation of their antimicrobial potential in milk. Int Dairy J 28(1):8–14 Ercolini D (2013) High-throughput sequencing and metagenomics: moving forward in the cultureindependent analysis of food microbial ecology. Appl Environ Microbiol 79:3148–3155 Espejo RT, Canelo ES (1968) Properties of bacteriophage PM2: a lipid-containing bacterial virus. Virology 34(4):738–747 Fay D, Bowman BU (1978) Structure of native and chloroform-methanol-treated mycobacteriophage R1. J Virol 27(2):432–435 Fortier L-C, Sekulovic O (2013) Importance of prophages to evolution and virulence of bacterial pathogens. Virulence 4:354–365 Fulton J, Douglas T, Young M (2009) Isolation of viruses from high temperature environments. In: Clokie MRJ, Kropinski AM (eds) Bacteriophages. New York, NY, USA: Springer, pp 43–54 García-Aljaro C, Muniesa M, Jofre J (2018) Isolation of bacteriophages of the anaerobic bacteria bacteroides. In: Azaredo J, Sillankorva S (eds) Bacteriophage therapy. New York, NY, USA: Humana Press, pp 11–22 Ghugare GS, Nair A, Nimkande V, Sarode P, Rangari P, Khairnar K (2016) Membrane filtration immobilization technique – a simple and novel method for primary isolation and enrichment of bacteriophages. J Appl Microbiol 122(2):531–539 Gill JJ, Hyman P (2010) Phage choice, isolation, and preparation for phage therapy. Curr Pharm Biotechnol 11:2–14 Gill J, Sabour P, Leslie K, Griffiths M (2006) Bovine whey proteins inhibit the interaction of Staphylococcus aureus and bacteriophage K. J Appl Microbiol 101:377–386 Greene J, Goldberg RB (1985) Isolation and preliminary characterization of lytic and lysogenic phages with wide host range within the Streptomycetes. Microbiology 131(9):2459–2465 Harper D, Blake K (2018) Therapeutic bacteriophage compositions. Patent. International Publication Number WO 2013/164640 A1 Harper DR, Parracho HMRT, Walker J, Sharp R, Hughes G, Werthén M, Lehman S, Morales S (2014) Bacteriophages and biofilms. Antibiotics 3:270–284 Henry M, Biswas B, Vincent L, Mokashi V, Schuch R, Bishop-Lilly KA, Sozhamannan S (2012) Development of a high throughput assay for indirectly measuring phage growth using the OmniLog™ system. Bacteriophage 2:159–167 Heuer OE, Pedersen K, Andersen JS, Madsen M (2001) Prevalence and antimicrobial susceptibility of thermophilic Campylobacter in organic and conventional broiler flocks. Lett Appl Microbiol 33(4):269–274

460

F. van Charante et al.

Hjorleifsdottir S, Aevarsson A, Hreggvidsson GO, Fridjonsson OH, Kristjansson JK (2014) Isolation, growth and genome of the Rhodothermus RM378 thermophilic bacteriophage. Extremophiles 18(2):261–270 Hurwitz BL, Sullivan MB (2013) The Pacific Ocean Virome (POV): a marine viral metagenomic dataset and associated protein clusters for quantitative viral ecology. PLoS One 8:e57355 Hyman PY, Abedon ST (2010) Bacteriophage host range and bacterial resistance. Adv Appl Microbiol 70:217–248 Jensen EC, Schrader HS, Rieland B, Thompson TL, Lee KW, Nickerson KW, Kokjohn TA (1998) Prevalence of broad-host-range lytic bacteriophages of Sphaerotilus natans, Escherichia coli, and Pseudomonas aeruginosa. Appl Environ Microbiol 64(2):575–580 Ji X, Zhang C, Fang Y, Zhang Q, Lin L, Tang B, Wei Y (2015) Isolation and characterization of glacier VMY22, a novel lytic cold-active bacteriophage of Bacillus cereus. Virol Sin 30 (1):52–58 Jończyk E, Kłak M, Międzybrodzki R, Górski A (2011) The influence of external factors on bacteriophages. Folia Microbiol 56:191–200 Kang W, Sarkar S, Lin ZS, McKenney S, Konry T (2019) Ultra-fast parallelized microfluidic platform for antimicrobial susceptibility testing of gram positive and negative bacteria. Anal Chem 91(9):6242–6249 Karumidze N, Kusradze I, Rigvava S, Goderdzishvili M, Rajakumar K, Alavidze Z (2013) Isolation and characterisation of lytic bacteriophages of Klebsiella pneumoniae and Klebsiella oxytoca. Curr Microbiol 66(3):251–258 Koser SA (1926) Action of the bacteriophage on a thermophilic Bacillus. Proc Soc Exp Biol Med 24(1):109–111 Kropinski AM, Mazzocco A, Waddell TE, Lingoh E, Johnson RP (2009) Enumeration of bacteriophages by double agar overlay plaque assay. In: Kutter E, Sulakvelidze A (eds) Bacteriophages: biology and applications. CRC Press, Washington, D.C., pp 69–76 Krylov V, Shaburova O, Krylov S, Pleteneva E (2012) A genetic approach to the development of new therapeutic phages to fight Pseudomonas aeruginosa in wound infections. Viruses 5:15–53 Kuo TT, Huang TC, Chow TY (1969) A filamentous bacteriophage from Xanthomonas oryzae. Virology 39(3):548–555 Kutter EM, Kellenberger E, Carlson K, Eddy S, Neitzel J, Messinger L, North J, Guttman B (1994) Effects of bacterial growth conditions and physiology on T4 infection. In: Karam J (ed) Molecular biology of bacteriophage T4. American Society of Microbiology, Washington, DC, pp 406–418 Kutter E, De Vos D, Gvasalia G, Alavidze Z, Gogokhia L, Kuhl S, Abedon ST (2010) Phage therapy in clinical practice: treatment of human infections. Curr Pharm Biotechnol 11:69–86 Kwiatek M, Parasion S, Rutyna P, Mizak L, Gryko R, Niemcewicz M, Olender A, Łobocka M (2017) Isolation of bacteriophages and their application to control Pseudomonas aeruginosa in planktonic and biofilm models. Res Microbiol 168(3):194–207 Lamont I, Brumby AM, Egan JB (1989) UV induction of coliphage 186: prophage induction as an SOS function. PNAS 86(14):5492–5496 Li M, Wang J, Zhang Q, Lin L, Kuang A, Materon LA, Ji X, Wei Y (2016) Isolation and characterization of the lytic cold-active bacteriophage MYSP06 from the Mingyong Glacier in China. Curr Microbiol 72(2):120–127 Li Y, Yang X, Zhao W (2017) Emerging microtechnologies and automated systems for rapid bacterial identification and antibiotic susceptibility testing. SLAS Technol Translating Life Sciences Innovation 22:585–608 Lin NT, Chiou PY, Chang KC, Chen LK, Lai MJ (2010) Isolation and characterization of ϕAB2: a novel bacteriophage of Acinetobacter baumannii. Res Microbiol 161(4):308–314 Lin L, Han J, Ji X, Hong W, Huang L, Wei Y (2011) Isolation and characterization of a new bacteriophage MMP17 from Meiothermus. Extremophiles 15(2):253–258 Liu B, Wu S, Song Q, Zhang X, Xie L (2006) Two novel bacteriophages of thermophilic bacteria isolated from deep-sea hydrothermal fields. Curr Microbiol 53(2):163–166 Łobocka M, Hejnowicz MS, Gagała U, Weber-Dabrowska B, Wegrzyn G, Dadlez M (2014) Phage Therapy: Current Research and Applications. In: Borysowski J, Miedzybrodzki R, Górski A

Isolation of Bacteriophages

461

(eds) The first step to bacteriophage therapy – how to choose the correct phage phage therapy: current research and applications, Norfolk, UK: Caister Academic Press pp 23–69 Loc Carrillo CM, Connerton PL, Pearson T, Connerton IF (2007) Free-range layer chickens as a source of Campylobacter bacteriophage. Antonie Van Leeuwenhoek 92(3):275 Loc-Carrillo C, Abedon ST (2011) Pros and cons of phage therapy. Bacteriophage 1:111–114 Locus Biosciences (2018) Locus biosciences acquires EpiBiome’s high-throughput discovery platform to enhance its global leadership in CRISPR-engineered phage therapeutics. https:// www.locus-bio.com/locus-biosciences-acquires-epibiomes-high-throughput-discovery-platformto-create-the-worlds-leading-crispr-engineered-bacteriophage-company/. Accessed on 9th May 2019 Lopez R, Ronda C, Tomasz A, Portoles A (1977) Properties of “diplophage”: a lipid-containing bacteriophage. J Virol 24(1):201–210 Lu Y, Gao J, Zhang DD, Gau V, Liao JC, Wong PK (2013) Single cell antimicrobial susceptibility testing by confined microchannels and Electrokinetic loading. Anal Chem 85:3971–3976. https://doi.org/10.1021/ac4004248 Luhtanen AM, Eronen-Rasimus E, Kaartokallio H, Rintala JM, Autio R, Roine E (2014) Isolation and characterization of phage–host systems from the Baltic Sea ice. Extremophiles 18(1):121–130 Maal KB, Delfan AS, Salmanizadeh S (2015) Isolation and identification of two novel Escherichia coli bacteriophages and their application in wastewater treatment and coliform's phage therapy. Jundishapur J Microbiol 8(3):e14945 Machuca P, Daille L, Vinés E, Berrocal L, Bittner M (2010) Isolation of a novel bacteriophage specific for the periodontal pathogen Fusobacterium nucleatum. Appl Environ Microbiol 76(21):7243–7250 Mapes AC, Trautner BW, Liao KS, Ramig RF (2016) Development of expanded host range phage active on biofilms of multi-drug resistant Pseudomonas aeruginosa. Bacteriophage 6(1): e1096995 Markel DE, Eklund C (1974) Isolation, characterization, and classification of three bacteriophage isolates for the genus Levinea. Int J Syst Evol Microbiol 24(2):230–234 Marks TJ, Hamilton PT (2014) Characterization of a thermophilic bacteriophage of Geobacillus kaustophilus. Arch Virol 159(10):2771–2775 Matsuzaki S, Uchiyama J, Takemura-Uchiyama I, Ujihara T, Daibata M (2018) Isolation of bacteriophages for fastidious bacteria. In: Azaredo J, Sillankorva S (eds) Bacteriophage therapy. New York, NY, USA: Humana Press, pp 3–10 Mattila S, Ruotsalainen P, Jalasvuori M (2015) On-demand isolation of bacteriophages against drug-resistant bacteria for personalized phage therapy. Front Microbiol 6. https://doi.org/ 10.3389/fmicb.2015.01271 Melo LD, Sillankorva S, Ackermann HW, Kropinski AM, Azeredo J, Cerca N (2014) Isolation and characterization of a new Staphylococcus epidermidis broad-spectrum bacteriophage. J Gen Virol 95(2):506–515 Merabishvili M, Pirnay JP, Verbeken G, Chanishvili N, Tediashvili M, Lashkhi N, Glonti T, Krylov V, Mast J, Van Parys L, Lavigne R, Volckaert G, Mattheus W, Verween G, De Corte P, Rose T, Jennes S, Zizi M, De Vos D, Vaneechoutte M (2009) Quality-controlled small-scale production of a well-defined bacteriophage cocktail for use in human clinical trials. PLoS One 4(3):e4944 Millard AD (2009) Isolation of cyanophages from aquatic environments. In: Clokie MRJ, Kropinski AM (eds) Bacteriophages: Methods and Protocols. New York, NY, USA: Humana Press, pp 33–42 Mirzaei MK, Nilsson AS (2015) Isolation of phages for phage therapy: a comparison of spot tests and efficiency of plating analyses for determination of host range and efficacy. PLoS One 10(3): e0118557 Muruga BN, Wagacha J, Kabaru J, Amugune N, Duboise M (2013) Isolation of bacteriophage infecting haloalkaliphilic bacteria in Lake Magadi, Kenya International. J Innov Res Dev 2:10 Nivas D, Ramesh N, Krishnakumar V, Rajesh P, Solomon EK, Kannan VR (2015) Distribution, isolation and characterization of lytic bacteriophages against multi-drug resistant and extended-

462

F. van Charante et al.

spectrum of Β-lactamase producing pathogens from hospital effluents. Asian J Pharm Clin Res 8(2):384–389 O'flaherty S, Coffey A, Meaney W, Fitzgerald G, Ross R (2005) Inhibition of bacteriophage K proliferation on Staphylococcus aureus in raw bovine milk. Lett Appl Microbiol 41:274–279 Olsen RH, Metcalf ES, Todd JK (1968) Characteristics of bacteriophages attacking psychrophilic and mesophilic pseudomonads. J Virol 2(4):357–364 Olsen RH, Siak JS, Gray RH (1974) Characteristics of PRD1, a plasmid-dependent broad host range DNA bacteriophage. J Virol 14(3):689–699 Owens J, Barton MD, Heuzenroeder MW (2013) The isolation and characterization of Campylobacter jejuni bacteriophages from free range and indoor poultry. Vet Microbiol 162(1):144–150 Paolozzi L, Ghelardini P (2006) The bacteriophage Mu. In: Calendar R (ed) The bacteriophages, 2nd edn. Oxford University Press, New York, pp 469–496 Parfitt T (2005) Georgia: an unlikely stronghold for bacteriophage therapy. Lancet 365:2166–2167 Pires DP, Oliveira H, Melo LD, Sillankorva S, Azeredo J (2016) Bacteriophage-encoded depolymerases: their diversity and biotechnological applications. Appl Microbiol Biotechnol 100(5):2141–2151 Pirnay JP, Blasdel BG, Bretaudeau L, Buckling A, Chanishvili N, Clark JR, Corte-Real S, Debarbieux L, Dublanchet A, De Vos D, Gabard J, Garcia M, Goderdzishvili M, Górski A, Hardcastle J, Huys I, Kutter E, Lavigne R, Merabishvili M, Olchawa E, Parikka KJ, Patey O, Pouilot F, Resch G, Rohde C, Scheres J, Skurnik M, Vaneechoutte M, Van Parys L, Verbeken G, Zizi M, Van den Eede G (2015) Quality and safety requirements for sustainable phage therapy products. Pharm Res 32(7):2173–2179 Pirnay JP, Verbeken G, Ceyssens P-J, Huys I, De Vos D, Ameloot C, Fauconnier A (2018) The magistral phage. Viruses 10(2):e64 Pootjes CF, Mayhew RB, Korant BD (1966) Isolation and characterization of Hydrogenomonas facilis bacteriophages under heterotrophic growth conditions. J Bacteriol 92(6):1787–1791 Popova AV, Zhilenkov EL, Myakinina VP, Krasilnikova VM, Volozhantsev NV (2012) Isolation and characterization of wide host range lytic bacteriophage AP22 infecting Acinetobacter baumannii. FEMS Microbiol Lett 332:40–46. https://doi.org/10.1111/j.1574-6968.2012.02573.x Raya RR, Hébert EM (2009) Isolation of phage via induction of lysogens. In: Clokie MRJ, Kropinski AM (eds) Bacteriophages: Methods and Protocols. New York, NY, USA: Humana Press, pp 23–32 Reyes A, Semenkovich NP, Whiteson K, Rohwer F, Gordon JI (2012) Going viral: next-generation sequencing applied to phage populations in the human gut. Nat Rev Microbiol 10:607 Rice G, Stedman K, Snyder J, Wiedenheft B, Willits D, Brumfield S, McDermott T, Young MJ (2001) Viruses from extreme thermal environments. Proc Natl Acad Sci 98(23):13341–13345 Rombouts S, Volckaert A, Venneman S, Declercq B, Vandenheuvel D, Allonsius CN, Van Malderghem C, Jang HB, Briers Y, Noben JP, Klumpp J, Van Vaerenbergh J, Maes M, Lavigne R (2016) Characterization of novel bacteriophages for biocontrol of bacterial blight in leek caused by Pseudomonas syringae pv. Porri. Front Microbiol 7:279 Ross A, Ward S, Hyman P (2016) More is better: selecting for broad host range bacteriophages. Front Microbiol 7:1352 Sakaki Y, Oshima T (1975) Isolation and characterization of a bacteriophage infectious to an extreme thermophile, Thermus thermophilus HB8. J Virol 15(6):1449–1453 Salifu SP, Casey SA, Foley S (2013) Isolation and characterization of soilborne virulent bacteriophages infecting the pathogen Rhodococcus equi. J Appl Microbiol 114(6):1625–1633 Santos SB, Carvalho CM, Sillankorva S, Nicolau A, Ferreira EC, Azeredo J (2009) The use of antibiotics to improve phage detection and enumeration by the double-layer agar technique. BMC Microbiol 9:148 Seeley ND, Primrose SB (1982) The isolation of bacteriophages from the environment. J Appl Bacteriol 53:1–17 Sillankorva S (2018) Isolation of bacteriophages for clinically relevant bacteria. In: Azaredo J, Sillankorva S (eds) Bacteriophage therapy. New York, NY, USA: Humana Press, pp 23–30

Isolation of Bacteriophages

463

Solonenko SA, Sullivan MB (2013) Preparation of metagenomic libraries from naturally occurring marine viruses. In: DeLong EF (ed) Methods in enzymology, vol 531. Cambridge, MA, USA: Elsevier, pp 143–165 Spencer R (1955) A marine bacteriophage. Nature 175(4459):690–691 Stedman K, Porter M, Dyall-Smith M (2009) The isolation of viruses infecting Archaea. In: Wilhelm S, Weinbauer M, Suttle C (eds) Manual of aquatic viral ecology. American Society of Limnology and Oceanography, Waco, TX, USA, pp 57–64 Steinberg VI, Hart EJ, Handley J, Goldberg ID (1976) Isolation and characterization of a bacteriophage specific for Neisseria perflava. J Clin Microbiol 4(1):87–91 Twort FW (1915) An investigation on the nature of ultra-microscopic viruses. Lancet 186(4814): 1241–1243 Tylenda CA, Calvert C, Kolenbrander PE, Tylenda A (1985) Isolation of Actinomyces bacteriophage from human dental plaque. Infect Immun 49(1):1–6 Uchiyama J, Rashel M, Maeda Y, Takemura I, Sugihara S, Akechi K, Muraoka A, Wakiguchi H, Matsuzaki S (2008) Isolation and characterization of a novel Enterococcus faecalis bacteriophage φEF24C as a therapeutic candidate. FEMS Microbiol Lett 278(2):200–206 Van Twest R, Kropinski AM (2009) Bacteriophage enrichment from water and soil. In: Clokie MRJ, Kropinski AM (eds) Bacteriophages: Methods and Protocols. New York, NY, USA: Humana Press, pp 15–21 Verma H, Pramod D, Abbas M, Prajapati A, Ramchandra D, Rawat M (2013) Isolation and partial characterization of lytic phage against Salmonella Abortusequi. Veterinary World 6 (2):72–75 Vidaver AK, Koski RK, Van Etten JL (1973) Bacteriophage φ6: a lipid-containing virus of Pseudomonas phaseolicola. J Virol 11(5):799–805 Vinod MG, Shivu MM, Umesha KR, Rajeeva BC, Krohne G, Karunasagar I, Karunasagar I (2006) Isolation of Vibrio harveyi bacteriophage with a potential for biocontrol of luminous vibriosis in hatchery environments. Aquaculture 255(1):117–124 Ward TE, Bruhn DF, Shean ML, Watkins CS, Bulmer D, Winston V (1993) Characterization of a new bacteriophage which infects bacteria of the genus Acidiphilium. J Gen Virol 74 (11):2419–2425 Weber-Dabrowska B, Jonczyk-Matysiak E, Zaczek M, Lobocka M, Lusiak-Szelachowska M, Gorski A (2016) Bacteriophage procurement for therapeutic purposes. Front Microbiol 7:1177. https://doi.org/10.3389/fmicb.2016.01177 Wells LE, Deming JW (2006) Characterization of a cold-active bacteriophage on two psychrophilic marine hosts. Aquat Microb Ecol 45(1):15–29 Wittmann J, Dreiseikelmann B, Rohde C, Rohde M, Sikorski J (2014) Isolation and characterization of numerous novel phages targeting diverse strains of the ubiquitous and opportunistic pathogen Achromobacter xylosoxidans. PLoS One 9(1):e86935 Wommack KE, Williamson KE, Helton RR, Bench SR, Winget DM (2009) Methods for the isolation of viruses from environmental samples. In: Clokie MRJ, Kropinski AM (eds) Bacteriophages: Methods and Protocols. New York, NY, USA: Humana Press, pp 3–14 Xie Y, Wahab L, Gill J (2018) Development and validation of a microtiter plate-based assay for determination of bacteriophage host range and virulence. Viruses 10:189 Yakimovich A, Andriasyan V, Witte R, Wang I-H, Prasad V, Suomalainen M, Greber UF (2015) Plaque2. 0 – a high-throughput analysis framework to score virus-cell transmission and clonal cell expansion. PLoS One 10:e0138760 Yamamoto KR, Alberts BM, Benzinger R, Lawhorne L, Treiber G (1970) Rapid bacteriophage sedimentation in the presence of polyethylene glycol and its application to large-scale virus purification. Virology 40(3):734–744 Yan J, Mao J, Xie J (2014) Bacteriophage polysaccharide depolymerases and biomedical applications. BioDrugs 28:265–274 Yang H, Liang L, Lin S, Jia S (2010) Isolation and characterization of a virulent bacteriophage AB1 of Acinetobacter baumannii. BMC Microbiol 10:131

464

F. van Charante et al.

Yehle CO, Doi RH (1967) Differential expression of bacteriophage genomes in vegetative and sporulating cells of Bacillus subtilis. J Virol 1(5):935–947 Yeung MK, Kozelsky CS (1997) Transfection of Actinomyces spp. by genomic DNA of bacteriophages from human dental plaque. Plasmid 37(2):141–153 Yu MX, Slater MR, Ackermann HW (2006) Isolation and characterization of Thermus bacteriophages. Arch Virol 151(4):663–679 Yu P, Mathieu J, Li M, Dai Z, Alvarez PJ (2016) Isolation of polyvalent bacteriophages by sequential multiple-host approaches. Appl Environ Microbiol 82(3):808–815 Zhou W, Feng Y, Zong Z (2018) Two new lytic bacteriophages of the myoviridae family against carbapenem-resistant Acinetobacter baumannii. Front Microbiol 9:850–850. https://doi.org/ 10.3389/fmicb.2018.00850

Bacteriophage Use in Molecular Biology and Biotechnology Nathan Brown and Chris Cox

Contents Early Contributions to Molecular Biology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Early Studies of Phage Chemical Composition and Physical Structure . . . . . . . . . . . . . . . . . . . . Spontaneous Mutation and Heredity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Growth and Plaque Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Typing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Amplification and MALDI-TOF MS for Bacterial Identification . . . . . . . . . . . . . . . . . . . The Discovery of Lysogeny and Prophages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The First Unequivocal Demonstration of Lysogeny . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Discovery of Phage λ and the Beginning of Prophage Genome Mapping . . . . . . . . . . . . The Discovery of Phage P1 Used in Bacterial Transduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Discovery of Site-Specific Recombination and Its Applications . . . . . . . . . . . . . . . . . . . . . . . . . . The Campbell Model of Phage Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Determination of Attachment Sites on Host and Phage DNA Molecules . . . . . . . . . . . . . The Development of Site-Specific Integrating Plasmids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discovery of the Cre/LoxP System in Phage P1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Applications of the Cre/Lox System in Genetically Engineered Mice . . . . . . . . . . . . . . . . . . . . . Discovery of Bacterial DNA Restriction and Modification Systems . . . . . . . . . . . . . . . . . . . . . . . . . . Discovery of the Restriction/Modification Phenomenon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discovery of the Role of Methylation in Restriction/Modification . . . . . . . . . . . . . . . . . . . . . . . . The First Identification of a Specific Restriction Site in Bacterial DNA . . . . . . . . . . . . . . . . . . . The Development of Molecular Cloning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Development of Restriction Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Discovery of Viral Metabolic Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Discovery of Thymineless Death and Its Implications for Chemotherapy Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Role of Phages in Understanding Gene Structure, Expression, and Regulation . . . . . . . . . .

466 467 469 470 470 472 473 473 475 475 476 476 477 479 479 480 481 481 481 482 483 484 485 485 486

N. Brown Department of Infection, Immunity, and Inflammation, University of Leicester, Leicester, UK e-mail: [email protected] C. Cox (*) Cobio Diagnostics, Golden, CO, USA e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_15

465

466

N. Brown and C. Cox

Luria and Delbruck’s “Fluctuation Test” Shows That Spontaneous Mutations Occur in Bacterial Genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hershey and Chase Demonstrate That DNA Is the Hereditary Material . . . . . . . . . . . . . . . . . . . Benzer Maps the Fine Structure of Genes in the Phage T4 rII Region . . . . . . . . . . . . . . . . . . . . The Discovery that Messenger RNA Is a Direct Product of Genes . . . . . . . . . . . . . . . . . . . . . . . . The Determination of the Genetic Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The First Observation of Gene Regulation Due to a Repressor Factor (The PaJaMo Experiment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The First Isolation and Characterization of a Repressor Factor Involved in Gene Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isolation of the ρ Termination Factor and Characterization of Termination/Antitermination in λ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Construction of a Synthetic Genetic Regulatory Circuit that Exhibits Complex Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Refactoring Phages to Learn About Complex Genetic Regulatory Circuits at the Organismal Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Systems Biology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ecology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

486 487 488 491 492 492 494 496 497 498 498 499 499 500 501 501

Abstract

Since their discovery in the early twentieth century, bacteriophages (phages) have played a central role in understanding many key principles in molecular biology. In particular, they were essential model organisms in the search for the physical nature and function of the gene, beginning with the establishment of the American Phage Working Group by Max Delbrück and extending to explication of Francis Crick’s central dogma of molecular biology through studies of RNA transcription and protein expression in phage λ. Beyond illuminating fundamental principles of molecular biology, phages have also been used extensively in biotechnology. Phage biology is a rich source of methods used in recombinant DNA technology, clinical diagnostics, and synthetic biology. Although phage biology is often criticized as passé, there are still compelling reasons to study phages. New frontiers of complexity in biology call for fresh research into phage biology that promises to yield important advances in our understanding of ecology and evolution, our ability to manipulate genetic material, and our investigations into emergent phenomena in systems biology. Here we trace the fundamental and applied discoveries enabled by the study of bacteriophage biology from the early twentieth century till today.

Early Contributions to Molecular Biology First described in 1917, bacteriophages (phages) were given their name as a result of Felix d’Herelle’s discovery of their antagonistic behavior towards bacteria. Early in his studies he reported the tendency of phages to form distinct plaques on bacterial lawns and lyse liquid cultures (d’Herelle 1917). Though his hypotheses on the nature

Bacteriophage Use in Molecular Biology and Biotechnology

467

of phages and their infectious cycle were not widely accepted at the time, he correctly interpreted that these observations were the result of viral infection and multiplication at the expense of the bacterial host. D’Herelle continued to gather evidence into the 1930s to support the idea of phages as discrete, organized obligate infectious bacterial viruses. Ultimately, he defined the host range and antigenic properties of a large number of phages, forming the idea of distinct phage “races” and he described the existence of bacterial mutants with phage resistance (d’Herelle 1931). In doing so, this work paved the way for the use of phages in a myriad of future discoveries in the fields of biochemistry, genetics, and microbiology. Indeed, in the decades to follow, the advances made possible by the study of phages led to the early formation of the field of Molecular Biology. In the late 1930s, Max Delbrück (a physicist by training) had an interest in advancing the physical understanding of genes, which were largely unexplored at the time. During this time, he was introduced to Emory Ellis, and Ellis’s use of phages as a model of viral oncogenesis. They recognized the potential for using phage biology to study genes and their transmission (heredity) and began collaborative studies aimed at further understanding the phage replication cycle, and host interactions. This partnership ultimately led to several key discoveries and greatly advanced our understanding of cell physiology and molecular biology. Together, and drawing on the early work of d’Herelle, they developed methods for phage quantification and defined the prototypical phage lytic lifecycle with the one-step phage growth curve, which is still widely used today (Ellis and Delbrück 1939). In 1945, along with Salvador Luria and Alfred Hershey, Delbrück formed the American Phage Working Group, which largely consisted of a collection of phage researchers focused on a specific set of E. coli strains and phages (which would become known as the T phages) with the idea that all their studies would use the same standardized suite of phages and experimental protocols. Under Delbrück’s supervision, this essentially unified the field with the goal of creating an opensource community of researchers that could easily compare and replicate bacterial and phage genetic data.

Early Studies of Phage Chemical Composition and Physical Structure During the period from the early 1930s through the early 1940s, the exact chemical and structural nature of phages was unknown. In 1936, Max Schlesinger discovered the nucleoprotein nature of viruses (Schlesinger 1936). By employing the Feulgen reaction, a hydrolytic staining technique borrowed from histology for visualization of DNA in tissues, he observed semi-quantitatively that phages consisted of approximately equal quantities amounts of protein and nucleic acids. In 1940, one of the earliest applications of the newly invented electron microscope was to determine phage morphology and visualize how they interacted with their bacterial hosts. Early electron micrographs (Fig. 1) taken independently by Pfankuch and Kausche (1940) as well as Ruska (1940) in 1940, and Luria and Anderson in 1942 allowed for the direct observation of phages. These studies

468

N. Brown and C. Cox

showed the particulate morphology of phages (Fig. 1a), and for the first time allowed a glimpse of the association of coliphages with the cell surface of E. coli (Fig. 1b, c). With this knowledge, phage researchers were then able to begin to piece together the elements of phage absorption and infection.

Fig. 1 Electron micrographs showing the particulate nature and morphology of coliphages in association with the E. coli cell surface. (a) Pfankuch and Kausche 1940; (b) Ruska 1940; (c) Luria and Anderson, 194

Bacteriophage Use in Molecular Biology and Biotechnology

469

Spontaneous Mutation and Heredity In 1952, Joshua and Esther Lederberg demonstrated the influence of spontaneous mutation on the heritability of bacterial phage and antibiotic resistance (Lederberg and Lederberg 1952). This was accomplished by making replicates of E. coli colonies from agar plates and transferring the same spatial pattern of growth from an initial plate to fresh plates (Fig. 2). To transfer colonies without disturbing their spatial relationship, they stretched sterile velveteen across a wood block and imprinted the surface of the master plate onto the fabric. The imprint was then used to replicate the original colonies on fresh plates consisting of either sterile media containing streptomycin or a surface lawn of T1 phage. Replicas to agar

Fig. 2 Demonstration of spontaneous mutation and clonal selection in heritable phage resistance by replica plating

470

N. Brown and C. Cox

containing phage or streptomycin showed that resistant mutants existed in clones from the initial plates. Following several rounds of enrichment, each resistance phenotype was isolated in pure culture. This gave the first definitive evidence of the genetic transmission of spontaneous mutations and would hint at the genetic basis of evolution.

Phage Growth and Plaque Formation Dating back to d’Herelle’s plating experiments, formation of phage plaques (observable areas of clearing on a bacterial lawn) has long served as a means of visualizing phage amplification in a host-specific fashion. This method typically consists of infection of a suspected bacterial host with a well-characterized, broadly speciesspecific phage, followed by plating on nutrient agar (Cherry et al. 1954; Stewart et al. 1998). If the bacterial host is susceptible to infection, lysis proceeds, resulting in the release of a large number of progeny phage and subsequent infection of other nearby bacteria. This in turn leads to the formation of visible plaques (Fig. 3). Traditional plaque assays have been used in the identification of numerous bacterial phylotypes including Bacillus anthracis (Thal and Nordberg 1968; Abshire et al. 2005), Enterococcus faecalis (Pleceas and Brandis 1974), Escherichia coli (Nicolle et al. 1952), Listeria monocytogenes (Shcheglova and Neidbailik 1968), Salmonella enterica (Felix 1956), and Staphylococcus aureus (Wallmark and Laurell 1951). See ▶ “Detection of Bacteriophages: Statistical Aspects of Plaque Assay” for additional discussion of the phage plaque assay.

Phage Typing In the early 1950s, nursery units in Australian, United States, Canadian, and British hospitals reported repeated outbreaks of highly virulent Staphylococcus aureus strains (Hillier 2006). The outbreak strains proved to be resistant to the wonderdrug penicillin and would recur. It wasn’t understood how the strains persisted or spread because there was no way to easily distinguish between the different virulent, resistant strains of Staphylococcus. That changed when Drs. Clair Isbister and Beatrix Durie at the Royal North Shore Hospital in Sydney contacted Dr. Phyllis Rountree, a bacteriologist at the Royal Prince Alfred Hospital, about a S. aureus outbreak in their nursery (Hillier 2006). Rountree had an interest in bacteriophages dating to the 1930s and was connected to the international bacteriophage typing network headquartered at the Staphylococcal Reference Laboratory at Colindale in London, UK. Phage typing is the method of plating a collection of different bacteriophages known to infect a bacterial genus on a clinical isolate that belongs to the same bacterial genus. The pattern of phages that can or cannot form plaques on the clinical strain gives a specific multivariate signature that identifies the strain more precisely than could any other method at the time. Fisk was the first to publish about

Bacteriophage Use in Molecular Biology and Biotechnology

471

Fig. 3 Classic plaque assay performed by addition of phage to a bacterial lawn on the surface of nutrient agar (A). If target bacteria are susceptible to infection, then phage attachment leads to insertion of phage genetic material (B), which reprograms bacterial replication machinery to produce numerous progeny phage (C). This is followed by lysis of the host bacterium and release of new phages for subsequent infection (D), resulting in development of zones of clearance in the lawn termed plaques. (Adapted with permission from Cox (2012))

phage typing in 1942, so the method had been recently established and simply needed to be adapted to this particularly important clinical use case (Fisk 1942). Rountree found that none of the phages in the central phage collection from Colindale could infect the strain from the Royal North Shore Hospital nursery outbreak. Only one of the phages that she had isolated in Australia, “Phage 80,” could infect the breakout strain. This suggested that the S. aureus strain was new and she named it S. aureus strain 80, after the indicative phage. When she phage-typed S. aureus strains from other outbreaks and compared them to strains sent to her from the UK, Australia, or the United States, she found that many of them were only susceptible to infection by “Phage 80” and were likely the same strain. Phage typing was precise enough to allow Rountree to track strain 80 as it colonized patients and returned home with them after discharge. The spread of the “hospital” strain to the

472

N. Brown and C. Cox

wider community explained its spread across multiple continents and its recurrence in the same nursery units. Rountree was the first to observe this phenomenon, thanks to the precision of her method. Her success prompted hospitals in the UK and Australia to include bacteriologists such as her on infection control committees and to adopt phage typing for bacterial epidemiology, which led to a sharp drop in the number of nosocomial infections, including the eradication of S. aureus strain 80 from hospital nurseries in the 1960s. Phage typing was superseded in precision and simplicity by pulsed-field gel electrophoresis and PCR-based methods, but for several decades it was the most precise method available for tracking bacterial strains involved in disease outbreaks.

Phage Amplification and MALDI-TOF MS for Bacterial Identification While historically useful, plaque formation has a number of limitations that can result in false negatives when used for bacterial identification. One concern, although somewhat rare, is the potential for some bacterial strains within a given susceptible phylotype to develop phage resistance (which means that the resistant subpopulation will not form plaques). In addition, plaque formation typically requires at least overnight incubation as well as bacterial cultures that also typically require a minimum of 18–24 h of incubation prior to phage infection. While conventional plaque assays remain tractable as a confirmatory measure of bacterial detection by other methods, such characteristics detract from the utility of their use as an approach to rapid bacterial identification. However, by combining the basic tenets of phage amplification with the powerful capabilities of modern mass spectrometry for protein detection, Voorhees and others have exploited the same characteristics of host specificity and phage replication at play in the basic plaque assay to develop new methodologies for rapid bacterial identification using matrix assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS). MALDI-TOF MS is an increasingly used technique for analysis of a range of analytes including peptides, proteins, carbohydrates, hydrocarbons and polymers. In recent years, MALDI-TOF MS protein profiling has gained widespread acceptance for diagnostic bacterial identification (Holland et al. 1996; Lay 2001; Seng et al. 2009). While useful for analysis of enriched bacterial cultures, the need for development of colonies lengthens the time to answer. However, by combining MALDITOF MS with phage infection of bacterial samples, it can significantly reduce the need for time consuming culturing, with time to answer in as little as an hour (Cox et al. 2012). By focusing on mass spectrometric detection of phage proteins produced during the natural course of amplification in the host rather than on direct bacterial analysis, the technique greatly expands the capability of MALDI-TOF MS by effectively lowering the amount of starting bacteria necessary to elicit a detectable signal. This was first demonstrated in 2003 for E. coli (Madonna et al. 2003; Madonna et al. 2007) and Salmonella enterica (subsp. enterica serovar typhimurium) (Rees and Voorhees 2005). Madonna et al. measured a MALDITOF MS limit of detection for intact E. coli cells of 1.0
105 colony forming

Bacteriophage Use in Molecular Biology and Biotechnology

473

units (cfu)/mL and decreased this by two orders of magnitude by exploiting MS2 phage amplification (Madonna et al. 2003). By exploiting phage-host specificity Rees and Voorhees extended this in 2005 to achieve reproducible, simultaneous detection of E. coli and S. enterica with MS2 and MPSS1 in mixed cultures (Rees and Voorhees 2005).

The Discovery of Lysogeny and Prophages The lysogeny phenomenon had been recorded as early as 1921 by Jules Bordet with Mihai Ciucă, and independently by E. Gildemeister (Gildmeister and Herzberg 1924). Newly isolated bacterial cultures were seen to lyse and produce plaques without any known cause, giving rise to the term “lysogeny” (or more precisely, “lysogenesis” or “lysogenicity”) (see ▶ “Temperate Phages, Prophages, and Lysogeny” chapter) (Bertani 2004). It was unclear at the time whether lysogeny was due to a strain within an isolate that carried extrinsic phage or if it was an intrinsic trait of an individual bacterium in the culture (known today as a “lysogen”) (Lwoff 1953). Further confusing the matter, Bordet insisted – in opposition to Felix d’Herelle – that all bacteriophage phenomena, including lysogeny, were due to a physiological trait of bacteria and not a virus (Lwoff 1953). In 1924, Gildemeister and Herzberg hypothesized and tried to show that a particular lysogenic strain of Escherichia coli, E. coli 88, could generate bacteriophages that would cause lysis without the addition of extrinsic bacteriophages. Their hypothesis was confirmed in E. coli 88 the following year by Bail (1922). Bail showed that a clonal bacterial isolate could be serially reisolated six times and still produce lysis, indicating that the lysogeny phenomenon was vertically inherited by the clone and not due to infection by an extrinsic phage carried by some constituent strain in a nonclonal bacterial population. Questions remained regarding how phages are released from a lysogen, which environmental stimulus gives rise to lysis in a lysogen, and how lysogens interact with each other and non-lysogens in a bacterial population. Although the phenomenon had been observed, named, and confirmed to not require the addition of extrinsic phages, it wasn’t until the late 1940s and early 1950s that experiments were done to convincingly describe the environmental causes and genetic attributes of lysogeny.

The First Unequivocal Demonstration of Lysogeny André Lwoff at the Pasteur Institute in Paris, France, sought to determine how phages were released from a lysogen. He decided to use individual bacteria to study this problem because, as he said, “I dislike mathematics, for which I am not gifted, and I wanted to avoid formulas, statistical analysis and, more generally, calculations as much as possible” (Lwoff 1966), This approach also eliminated the possibility of a carrier strain within a bacterial population that might introduce extrinsic phage as a possible cause of lysis or any other unforeseen interaction

474

N. Brown and C. Cox

between heterogenous cells in a population. He selected Bacillus megaterium as the model bacterium due to its relatively large size (a rod approximately 4 um long by 1.5 um in diameter) and used specially designed equipment built with the help of a microforge to manipulate single B. megaterium cells. His experiments were reminiscent of those of Lazarro Spallanzani in the eighteenth century. Individual B. megaterium cells were suspended in media droplets on glass slides, subjected to different conditions, observed under the microscope for lysis, and plated on a sensitive indicator strain of B. megaterium which would yield plaques. Lwoff and his colleagues first established that individual B. megaterium lysogens could be seen, under the microscope, spending most of the time growing and dividing without liberating virions. Adding lysozyme to forcibly lyse the cells without destroying any virions did not release infectious virions. It was concluded that lysogenic bacteria did not contain virions while growing and dividing. Only occasionally would all the B. megaterium cells in a droplet – “say four to eight” (Lwoff 1966) – lyse simultaneously and release approximately one hundred virions per cell. Lwoff named the quiescent, inherited factor that gave rise to these sudden, occasional lysis events the “prophage.” He explains his choice, saying, The term could be criticized, of course, for ‘prophage’ means more or less ‘before the meal.’ Strictly speaking, prophage would be a sort of ‘hors d’oeuvre.’ But since it is customary to speak of ‘phage’ instead of bacteriophage and since ‘probacteriophage’ is a long word, it would inevitably have wound up as prophage anyway. So, hors d’oeuvre – I mean prophage - was proposed at the outset. The world was obviously eagerly awaiting the coming of the prophage, for, despite its French origin, the Greek term was rapidly and unanimously adopted. (Lwoff 1966)

Having established the predominantly quiescent nature of lysogeny and naming the quiescent factor that leads to lysis in a lysogen the “prophage,” in 1949 Lwoff formally hypothesized that an environmental factor induced the prophage to lyse the cell. To determine the environmental factor, growth curves for lysogenic cultures of B. megaterium were tracked by optical density while the cultures were subjected to different environmental conditions. Without altering typical growth conditions it was apparent that towards the end of the exponential phase, 15% of the bacterial population produced phage, presumably due to the way in which bacterial metabolism altered the medium. This observation supported Lwoff’s hypothesis and drove his effort to determine precisely which conditions induced lysis. The search produced numerous negative results as many conditions did not induce lysis. Finally, in one seemingly irrational experiment, 60 min after irradiating the culture under a UV lamp for a few seconds, the culture stopped growing normally and showed signs of lysis. Plating the lysate on an indicator strain showed that the culture had indeed produced a large concentration of phages (Lwoff 1966). This was not the only condition found to induce the prophage to cause lysis, as thioglycolic acid and other reducing agents in combination with copper were also found to be inducers. Ultimately it was determined that the reducing agents were oxidized by copper, producing hydrogen peroxide, which in turn induced the prophage to lyse the B. megaterium culture (Lwoff 1966). The paradigm of prophage induction

Bacteriophage Use in Molecular Biology and Biotechnology

475

established by Lwoff and his colleagues gave rise to detailed inquiry into the molecular mechanisms of induction by many other groups, which would have far reaching consequences extending to the present day.

The Discovery of Phage l and the Beginning of Prophage Genome Mapping Simultaneously with Lwoff’s work at the Pasteur Institute, Esther and Joshua Lederberg at the University of Wisconsin were establishing the premier model of lysogeny, phage λ in Escherichia coli K-12. Esther Lederberg discovered λ in 1950 while crossing a sensitive, non-lysogenic mutant stock (either W-435 or W-518) of E. coli K-12 with an E. coli K-12 lysogen, wherein turbid plaques were observed (Lederberg and Lederberg 1953). The Lederbergs found that by crossing lysogenic auxotrophic marker strains (strains that are unable to grow without certain nutrients added to their growth media) of E. coli K-12 with sensitive E. coli K-12 auxotrophs, λ could be linked to the Gal4 locus on the E. coli chromosome. They note, however, that “This work was initiated in the expectation that λ would behave as an extranuclear factor, and might indeed provide a favorable model system for studies of cytoplasmic heredity” (Lederberg and Lederberg 1953). That λ was linked to the Gal4 chromosomal locus in crosses – which includes the gal genes responsible for galactose metabolism and can therefore be used as a genetic marker – suggested otherwise, but there still remained the possibility that “a segregating nuclear factor which is concerned with the maintenance of the pro-λ,” would allow for λ to exist free of the chromosome in the cytoplasm (Lederberg and Lederberg 1953). Only in 1960 were Calef and Licciardello able to furnish the extensive linkage mapping experiments that convincingly showed that λ was colinear with the E. coli K-12 chromosome and neither formed a branching structure off of the chromosome nor persisted apart from the chromosome (Calef and Licciardello 1960). This data was critical for Allan Campbell’s formulation of the “Campbell model” of provirus integration into the bacterial chromosome, first proposed in a 1963 paper on “Episomes” (Campbell 1963). Campbell’s model would in turn open inquiry into the molecular details of provirus integration into the bacterial chromosome (discussed in the section on “The Discovery of Site-Specific Recombination and Its Applications”).

The Discovery of Phage P1 Used in Bacterial Transduction In 1951, at approximately the same time as Lwoff and the Lederbergs were working on their respective questions regarding lysogeny, Giuseppe Bertani had stumbled on three prophages of the E. coli “Li” strain (Bertani 1951). Bertani had been asked by Joshua Lederberg to not work on the E. coli λ lysogen before he and Esther had the chance to publish their findings on λ. Instead, Joshua Lederberg sent the lysogenic E. coli “Li” strain and the phage-sensitive Shigella dysenteriae “Sh” indicator strain.

476

N. Brown and C. Cox

Bertani did his work with the Shigella strain over his then-labmate Jim Watson’s protestations, as Bertani recalls that he, “remember[s] Jim declaring at the top of his voice that he would not want to be in a lab where one used routinely the presumably pathogenic Shigella” (Bertani 2004). When phages liberated from E. coli “Li” were plated on Shigella dysenteriae strain “Sh,” Bertani observed three different plaque morphologies. He isolated phage from each type of plaque and named the isolates P1, P2, and P3 (Bertani 1951). Phage P1 would prove to be a unique phage with the useful ability to transduce large segments (approximately 100 kbp) of the bacterial chromosome. Further, phage P2 played an important role in the discovery of restriction-modification systems (discussed later in this chapter).

The Discovery of Site-Specific Recombination and Its Applications In hindsight it is difficult to appreciate the circumstances under which site-specific integration was discovered. In the early 1960s, it was still debated whether or not phage and bacterial chromosomes were circular, linear, or branched (Campbell 1993). The two leading models for lysogeny were that the phage chromosome became colinear with the bacterial chromosome and that the phage chromosome would attach to the bacterial chromosome as a branch. These models were depicted in Allan Campbell’s seminal review of episomes (genetic elements that can exist as part of a chromosome or separately from it), shown in Fig. 4 and described below.

The Campbell Model of Phage Integration Campbell, thanks to discussions with Frank Stahl about the circularity of chromosomes and genetic marker linkage data from Calef and Licciardello, proposed his “Campbell Model” of episome integration (Campbell 1993). Campbell noticed that Fig. 4 Two competing models of lysogeny from 1961. Model 1 shows the phage chromosome integrated colinearly into the bacterial chromosome. Model 2 shows the phage chromosome integrated as a branch of the bacterial chromosome. (Figure is redrawn from Campbell (1963))

Bacteriophage Use in Molecular Biology and Biotechnology

477

Fig. 5 The Campbell model of episome integration. In this model, the λ chromosome is hypothesized to be circular. It integrates colinearly with the E. coli K-12 chromosome at a region of homology “ABCD” shared between the λ and E. coli K-12 genome. The “ABCD” region, which came to be known as the attP site, sits between the h and cl markers on the λ chromosome, explaining how the apparent order of the h, cl, and mi markers is different in genetic linkage analysis of the vegetative λ chromosome and the λ lysogen. (Figure is redrawn from Campbell (1963))

according to Calef and Licciardello, when vegetative phage were crossed, the order of three markers on the chromosome were “h-cl-mi,” but the order of those markers in λ lysogens that were crossed was “try-h-mi-cl-gal,” wherein the bolded “try” and “gal” markers sit on the bacterial chromosome (Campbell 1993). It appears that the order of the mi and cl markers is transposed between the vegetative phage and lysogen unless one assumes that the phage chromosome is circular. In that case, as Campbell proposed, the phage chromosome integrates with the bacterial chromosome at a point “ABCD” that sits between h and cl, linearizing the circular phage chromosome and explaining the apparent transposition of mi and cl in genetic linkage studies. Figure 5 is a redrawn figure from the original figure depicting the Campbell Model in “Episomes.”

The Determination of Attachment Sites on Host and Phage DNA Molecules The Campbell model opened the way to discovering the proteins, DNA sequences, and molecular mechanisms involved in integrating the λ chromosome into the E. coli K-12 chromosome. In 1967, Gingery and Echols, and J. Zissler independently, determined that an integrase gene product (Int) was necessary for catalyzing integration (Gingery and Echols 1967). Gingery and Echols determined the location of int to be just to the right of the attP site on the λ chromosome by testing a library of λsus mutants (given to them by Campbell) in a colorimetric abortive lysogeny test. In 1970, Guarneros and Echols determined that an excision gene product (Xis) was necessary in addition to Int for catalyzing the excision of the chromosome from the bacterial chromosome when the lysogen was induced to enter lytic mode (Guarneros and Echols 1970). They used the same strategy, which again relied upon testing the

478

N. Brown and C. Cox

Fig. 6 The current model of λ site-specific recombination with the E. coli K-12 chromosome. The attP site is noted as POP0, where P and P0 are arm sequences of attP and O is the 15-bp core sequence shared with all att sites. attB is shown as BOB0 , where B and B0 are the arm sequences of attB. The recombined attL and attR sites are shown as BOP0 and POB0 , respectively. (Figure is redrawn from Weisberg and Landy (1983))

λsus mutant library, but instead used a prophage-curing test that depended on the cI857 temperature-sensitive repressor allele. Only much later in 1978 was a bacterial protein, Integration Host Factor (IHF), found to be important for λ chromosome integration (Kikuchi and Nash 1978). In the same year, Landy and Ross showed that the attP DNA sequence that lies at the site of integration in the λ chromosome and the attB DNA sequence at the site of integration in the E. coli K-12 chromosome have a 15-nucleotide-pair core sequence of homology (“GCTTTTTTATACTAA”). This core is also shared with the new junction sites, attL and attR, formed between the prophage chromosome and the host cell chromosome in the lysogen. These 15 nucleotide pairs are the “O” core of the att sites (Fig. 6). The full attP site is 240 bp and interacts with IHF and Int during integration, and IHF, Int, and Xis during excision (Weisberg and Landy 1983). The attB site is only approximately 25 bp, including the core region, and only binds to Int at the boundaries of the core sequence region during site-specific recombination with the λ chromosome. Its relatively passive role in recombination led to its designation as a recipient site, whereas the larger and more involved attP site is considered the donor site (Weisberg and Landy 1983). The λ integration/excision reactions are conservative in the sense that λ is never replicated during recombination. It was later discovered that replicative (rather than conservative) site-specific recombination also occurred, such as in phage mu and some insertion elements and transposons.

Bacteriophage Use in Molecular Biology and Biotechnology

479

The Development of Site-Specific Integrating Plasmids Atlung et al. in Denmark pioneered a major application of conservative site-specific recombination in 1991, when they developed the first integrative plasmid that used a phage integrase (Atlung et al. 1991). Although they were the first to develop sitespecific-integrative plasmids for general use, Koob and Szybalski were the first to integrate a circular DNA molecule into the attP site in E. coli K-12 in 1990 (Koob et al. 1988). Before this innovation, plasmids were integrated into the bacterial chromosome by ligating approximately 500–1000 bp fragments from a bacterial chromosome into plasmids unable to replicate in bacterial recipients. Plasmids are recombined at some frequency into bacterial chromosomes via the bacterial RecA pathway. Atlung et al. showed that one plasmid carrying a λ attP site and another plasmid carrying the λ int gene could facilitate the integration of the plasmid carrying the attP site into the attB site on the E. coli K-12 chromosome. This increased the frequency of plasmid recombination into the E. coli chromosome 100-fold compared to RecA-driven recombination, as well as made it easier to construct plasmids for integration into the chromosome. Since their invention, many site-specific-recombination-driven integrative plasmids have been built for different bacterial species (Lee et al. 1991; Auvray et al. 1997). Further, certain phage integrases have been shown to integrate plasmids bearing attP sites from diverse temperate phages into eukaryotic species at pseudo-att sites (Thomason et al. 2001; Nkrumah et al. 2006).

Discovery of the Cre/LoxP System in Phage P1 Phage P1 presented another productive avenue of research into phage site-specific recombination. Bertani’s discovery of temperate phage P1 is described in the section from this chapter on ▶ “Temperate Phages, Prophages, and Lysogeny.” In 1981, Nat Sternberg and Daniel Hamilton at the National Cancer Institute in Frederick, Maryland, USA discovered that a 6.5 kbp fragment of the P1 chromosome, when cloned into a λ vector, can cause recombination of the vector at two locations called loxP sites, which was observed by reassortment of λ genetic markers on the vector (Sternberg et al. 1981a). Sternberg showed in a separate study in 1981 that the same fragment of the P1 chromosome can cause integration of the λ vector into a specific site of the E. coli K-12 chromosome called loxB, though at lower efficiency than recombination between two loxP sites (Sternberg et al. 1981a). This recombination event was independent of the bacterial RecA pathway, but required two DNA sequences present on the P1 genome fragment called loxP (locus of crossing over (x), Pl), and the Cre protein encoded on the P1 genome fragment. P1 exists as an autonomous plasmid during infection of the bacterial host, so the Cre/loxP system doesn’t primarily exist to facilitate P1 integration into the bacterial chromosome. Instead Cre and the loxP sites are important in the P1 lifecycle for cyclization of the linear P1 chromosome upon infection and the resolution of circularized P1 dimer

480

N. Brown and C. Cox

molecules prior to host cell division (Sternberg et al. 1981b). Dimer resolution by Cre/loxP ensures even distribution of P1 molecules between the daughter cells, rather than allowing one daughter cell to inherit a dimer P1 molecular, diluting the population of P1-infected cells. The loxP sites were determined to be 63 bp DNA sequences that contain a loxP-loxP crossover point between two inverted repeats (Hoess et al. 1982). Further, the loxB site was shown to also contain two inverted repeats, though had a different nucleotide sequence from loxP. These findings showed that the Cre/loxP system is altogether different in mechanism and function from the Int/Xis/IHF/attP system. Later it was found that by converting the inverted repeats to two 34-bp direct repeats, called lox sites, any DNA sequence between the direct repeats would be deleted after Cre-mediated recombination.

Applications of the Cre/Lox System in Genetically Engineered Mice Practical implications of the Cre/lox system were developed most heavily in the field of eukaryotic – especially mouse – genetics. In 1987, Brian Sauer showed that the Cre/lox system worked in yeast Sauer (1987). Two years later, Sauer and Henderson at the E.I. du Pont de Nemours company showed that the Cre/lox system could function on mouse chromosomes (Sauer and Henderson 1989). This discovery enabled mouse geneticists to design DNA constructs including lox sites that could be inserted into mouse chromosomes and give conditional control over excision events therein. The Cre/lox system has since been used in mice and other eukaryotes to perform elaborate experiments that depend on deleting genes at a certain stage of the experiment, a certain stage of organismal development, or in a particular tissue type (Xiao and Weaver 1997; Postic et al. 1999). Especially, it enabled otherwise embryonic-lethal gene deletions to be studied in mice by putting Cre expression under the control of promoters that do not express Cre until later stages of mouse development (Betz et al. 1996). One of the most spectacular uses of the Cre/lox system is a study published by Livet et al. in 2007 (Livet et al. 2007). The aim of this study was to build a map of neuron connections in the mouse brain. To do this, the authors built a genetic construct with four fluorescent-protein-coding sequences punctuated by orthogonal lox-sites that are unable to recombine with the other sites (called loxP, loxN, and lox2272). The genetic construct was introduced to an embryonic mouse genome from which a stable line of mutant mice was established. When Cre is expressed in these mice as the brain develops, only one of the orthogonal lox sites undergoes recombination in each newly formed neuron, turning on or off expression of some of the fluorescent proteins and resulting in the expression of random combinations of fluorescent protein expression in each neuron. The resulting images are reminiscent of David Goodsell’s highly stylized art depicting topics from molecular biology. Consequently, the neurons in the mouse brain each fluoresce with a distinct color. This approach may one day enable the unambiguous identification and tracking of neurons as they develop in the mouse brain.

Bacteriophage Use in Molecular Biology and Biotechnology

481

Discovery of Bacterial DNA Restriction and Modification Systems The discovery of bacterial DNA restriction and modification systems began with the observation that the bacterial host could modify the properties of phages released from the host. Until the early 1950s, it was thought that, “one of virology’s most generally valid rules is that the properties of virus particles are unaffected by the host in which they grow” (Luria and Human 1952). Luria and Human showed otherwise in a 1952 paper (Luria and Human 1952). Starting with 100 independent mutant lines of E. coli strain B that had evolved resistance to phage T4, called “B/4” mutants, they found 25 B/4 mutant strains, which they divided into two classes, B/40 and B/400, that seemed to modify the host range properties of T2 and T6 phages. Though T2 and T6 could normally infect the parent E. coli B strain and its mutants, when T2 and T6 infected these B/40 and B/400 strains, the phage progeny, called phage T*, could not infect E. coli strain B or any of its B/4 mutants except for some old cells from some of the B/4 mutants. T* phages were modified by their B/40 and B/400 hosts in some way, restricting their host range, though it was unclear how.

Discovery of the Restriction/Modification Phenomenon The following year, Bertani and Weigle published a collaborative study between their separate labs that showed that both temperate phages λ and P2 also undergo host-mediated changes during infection of different bacterial hosts (Bertani and Weigle 1953). When P2 infects Shigella dysenteriae strain “Sh,” approximately 0.01% of the phage progeny, called P2B, are able to infect E. coli B, whereas 99.99% of the progeny cannot (Fig. 7). All of these P2B phage progeny, after infecting E. coli strain B, can then infect both strains S. dysenteriae strain “Sh” and E. coli strain B. The same phenomenon occurs when approximately 0.01% of λ progeny from an infected culture of E. coli strain C are able to infect E. coli strain S. These phage progeny are called λC. If λC is plated on E. coli strain S, then the resulting λ progeny can plate with 100% efficiency on both E. coli strains S and C. The high frequency of the phenomenon observed in the population of phage progeny excluded random mutation of the phage genome as an explanation, leaving no apparent mechanism. However, the chance observation of the phenomenon of host-dependent modification of the phage in two separate phages by two separate labs underscored the generality and importance of the phenomenon.

Discovery of the Role of Methylation in Restriction/Modification In 1962, Arber and Dussoix discovered that the host specificity of a phage is due, in part, to a host factor that recognizes the phage DNA molecule after it enters the cell and partially degrades the molecule (Arber and Dussoix 1962; Dussoix and Arber 1962). As observed by Lederberg in 1957 (Lederberg 1957), Arber and Dussoix saw that P32-labeled λ DNA molecules within a restrictive bacterial host cell become

482

N. Brown and C. Cox

Fig. 7 Host modification of phage progeny as understood in 1953. The host range of phage λ is shown to be modified by plating on E. coli strains S and C. In the same way, the host range of phage P2 is shown to be modified by plating on S. dysenteriae strain “Sh” or E. coli strain B. (Figure is drawn from Bertani and Weigle (1953))

partly acid-soluble, an indication of partial DNA degradation. Additionally, by coinfecting with λ phage mutants adapted the host at the same time as nonadapted λ phage that are restricted by the host, they were able to rescue 10–20% of the restricted λ genome from degradation. This strongly supported the notion that the phage DNA molecule enters the cell and is then partially degraded within that cell by some host factor. They noted that this degradation phenomenon had been observed with other types of foreign DNA entering bacterial cells, such as plasmid DNA. In 1963, Gold and Hurwitz described site-specific methylation of nucleic acid molecules (Gold and Hurwitz 1963). In particular, they noted that the λ DNA molecule is methylated less after passage through E. coli strain K-12 than after passage through E. coli strain C, coincident with a change in host range (Gold and Hurwitz 1963). Methylation was hypothesized to be part of the host modification of phage DNA, but the full nature of bacterial modification/restriction systems was still unclear. Methylation is now known to involve the addition of methyl groups to different positions on adenine and cytosine nucleobases in bacterial DNA molecules.

The First Identification of a Specific Restriction Site in Bacterial DNA A major breakthrough in understanding bacterial restriction/modification systems came in 1970 when Kelly and Smith published the first DNA recognition site of a

Bacteriophage Use in Molecular Biology and Biotechnology

483

restriction enzyme responsible for degrading foreign DNA in bacteria (Kelly and Smith 1970). The enzyme, endonuclease R from Haemophilus influenzae strain Rd., does not degrade host DNA, but degrades foreign DNA at a specific site. When phage T7 DNA was used as an in vitro substrate for endonuclease R, the DNA molecule was cleaved into pieces approximately 1 kbp long and the 50 termini of the fragments showed the degradation site to be 50 -GTY|RAC-30 , where “Y” is a pyrimidine nucleobase, “R” is a purine nucleobase, and the double-stranded cleavage site is the pipe character “|” in the middle. Fortunately, endonuclease R was from the Type II family of restriction endonucleases, which meant that its cleavage and recognition site were the same, unlike restriction endonucleases from other families that Arber and others had studied. Further, the DNA sequence specificity of methylase IIa isolated from Haemophilus influenzae strain Rd. was shown to correspond to the DNA sequence specificity of endonuclease R (Roy and Smith 1973). In 1975, Roszczyk and Goodgal found that methyltransferases could block restriction endonuclease activity by methylating the restriction sites of foreign DNA molecules (Roszczyk and Goodgal 1975). They treated phage T7 DNA in vitro with methylase IIa and showed that it partially protected the DNA from degradation by endonuclease R. Bacterial restriction (degradation of phage and other foreign DNA molecules) and modification (methylation) could now be seen as part of a single system for defense against parasitic foreign DNA molecules. Unfortunately, “endonuclease R” was not pure and contained two different restriction enzymes, later named HindII and HindIII (Loenen et al. 2014). In 1974, Landy et al. determined the exact recognition and cleavage site of pure HindIII to be 50 -A|AGCTT-30 , which is a palindrome like many Type II restriction endonuclease recognition sites (Landy et al. 1974). Like many restriction endonucleases, it produces a staggered cut, which results in cohesive or “sticky” ends: 4-bp single-stranded DNA (ssDNA) ends that can anneal to complementary ssDNA ends. These “sticky ends” proved useful in molecular cloning, one of the most well-known and important techniques in molecular biology.

The Development of Molecular Cloning Molecular cloning is the excision of a DNA molecule and insertion of that molecule into some replicon (such as a plasmid or a bacterial chromosome) so that the DNA molecule can be multiplied for many downstream applications. Restriction endonucleases are often used for excision of the DNA molecule and its insertion into a bacterial replicon. A restriction endonuclease, usually from the Type II family, can create ssDNA ends on the DNA molecule of interest that will be complementary to the ssDNA ends made in a bacterial replicon by the same restriction endonuclease. When the DNA molecule is mixed with the bacterial replicon, the complementary ssDNA ends on the two molecules will anneal to each other and the remaining gap in the DNA phosphodiester backbone between the two molecules can be sealed with DNA ligase (itself derived from E. coli bacteriophage T4), resulting in a single circular DNA molecule that can replicate after being introduced into a bacterium.

484

N. Brown and C. Cox

The process of uniting two DNA molecules in vitro was first described by Jackson et al. in 1972, who cut SV40 circular DNA (from Simian vacuolating virus 40 that grows on green monkey cells) with restriction endonuclease RI, cut the λdvgal replicon with the same enzyme, trimmed the double-stranded DNA ends of the molecules back with λ exonuclease, and joined them in vitro with a cocktail of enzymes including DNA polymerase and ligase (Jackson et al. 1972). The same year, Mertz and Davis showed that the sticky ends produced by the RI restriction endonuclease were sufficient to promote recombination between two separate molecules cut by the RI enzyme (λ exonuclease was unnecessary for cloning) (Mertz and Davis 1972). Restriction enzymes were ultimately adopted as the reagent of choice for cloning. Another notable contribution of phage biology to cloning was the development of cosmids. Phage λ replicates its chromosome through rolling circle replication, creating a long concatemer of several attached copies of the chromosome. Each copy of the chromosome is bordered by cos regions, which are approximately 200 bp regions on the λ chromosome that are cut by the λ terminase holoenzyme at the cosN site to form two complementary 12-bp ssDNA overhangs, similar to many restriction sites (Feiss et al. 1983). Each of the chromosomal fragments that are cut from the concatemer is then packaged into the λ capsid by the terminase. Cosmids take advantage of this λ biology. In 1978 Collins and Hohn showed that by including the cos ends of the λ chromosome on cloning plasmids (hence the name “cosmids”), the plasmids could be packaged in vitro in λ capsids (Collins and Hohn 1978). Cosmids allowed for tens of kilobases of recombinant DNA to be packaged into λ capsids, which would then efficiently deliver the large recombinant DNA payload into an E. coli cell, where it could be recombined into the chromosome or maintained as a separate replicon.

The Development of Restriction Mapping Restriction mapping is an approach to identifying physical locations on a DNA molecule by digesting the molecule with restriction enzymes and measuring the sizes of the fragments that result. Each DNA molecule will have a distinct pattern of fragment sizes. It was broadly used in biology to replace genetic linkage maps that were based on scoring biochemical or other phenotypic traits in breeding or bacterial crosses. There are many more restriction sites in a DNA molecule than there are phenotypic markers, so restriction mapping increased the resolution of genomic maps, as well as made it easier to build a map. It was used in several contexts for several reasons, giving rise to restriction-fragment-length-polymorphism (RFLP) mapping of whole genomes (Botstein et al. 1980), RFLP-based phylogenetics (Upholt 1977; Nei and Tajima 1981), DNA fingerprinting for identifying perpetrators based on DNA evidence left at a crime scene (Gill et al. 1985), and verification of trivial changes made to DNA molecules with new molecular techniques such as cloning.

Bacteriophage Use in Molecular Biology and Biotechnology

485

The Discovery of Viral Metabolic Products The early 1950s saw a flurry of research on changes in cell metabolism when Escherichia coli were infected with T-even phages. Several of these changes in E. coli metabolism were interesting because they gave vital clues to how nucleic acids and proteins encoded genetic information. The composition of nucleic acid, usually consisting of the nucleobases adenine, guanine, cytosine, and thymine/ uracil, was an important clue to understanding these processes. In 1953 GR Wyatt and Seymour Cohen discovered that phages T2, T4, and T6 did not use the typical cytosine nucleobase (Wyatt and Cohen 1953). By using chromatography to separate what had previously been thought to be cytosine and using ultraviolet spectroscopy and elemental analysis to further characterize the nucleobase, they found that there was likely an additional hydroxymethyl group attached to the 50 carbon of the “cytosine” nucleobase. The hypothetical hydroxymethylcytosine (HMC) nucleobase was synthesized and its spectroscopic properties were found to be the same as those of the HMC nucleobase isolated from T-even phage DNA, confirming their hypothesis. They further found that HMC provided the T-even phages with some resistance to deoxyribonucleases, which, they noted, may give the phages a competitive advantage during infection. These viruses also seemed to be unique among many other viruses in using HMC in their DNA. The most important implication of their work, however, was that viruses could induce the synthesis of a compound that could not be synthesized by the host cell. This provided a convenient marker that could be used to distinguish viral nucleic acid from host nucleic acid and later led to deeper studies of a phenomenon called “thymineless death” in E. coli.

The Discovery of Thymineless Death and Its Implications for Chemotherapy Research Thymineless death occurs in certain strains of E. coli that are auxotrophic for thymine. If these strains are deprived of thymine, then they will not revive if thymine is restored, whereas most nucleobase auxotrophs can be revived after temporary nucleobase deprivation. Cohen described this as “unbalanced” growth because only nucleotide metabolism was affected; the rest of cell metabolism continued unabated (Cohen and Barner 1954). However, if a thymine auxotroph, such as E. coli 15T-, was deprived of thymine and then infected with phage T2, it produced large amounts of thymine and HMC (Barner and Cohen 1954). This was found to be due to the overproduction of thymidylate synthase and hydroxymethylcytidylate synthase induced by the phage (Flaks and Cohen 1959). Importantly, this system allowed for inhibitors of these important nucleotide metabolism enzymes to be rigorously studied at the biochemical level. Some of the first cancer chemotherapy candidates were in fact inhibitors of nucleotide metabolism enzymes that could cause conditions similar to thymineless death in bacteria. Simultaneously with Cohen’s work on thymine metabolism, Charles Heidelberger and Robert Duschinsky found that

486

N. Brown and C. Cox

5-fluorouracil prevented tumor growth in a mouse model (Heidelberger et al. 1957). They asked Cohen if he could work out the mechanism of antineoplastic activity for 5-fluorouracil in his bacteria and phage model system, which he did and published in 1958 (Cohen et al. 1958). Phage-induced synthesis of thymidylate synthase and hydroxymethylcytidylate synthase thus made important mechanistic work in nucleotide metabolism possible when it was needed in the early days of cancer chemotherapy research.

The Role of Phages in Understanding Gene Structure, Expression, and Regulation In what is often considered the beginning of the field of molecular biology, Beadle and Tatum proposed their “one gene-one enzyme” model in 1941, which connected the role of genes to that of proteins and gave a broad but useful concept of the gene that largely persists today (Beadle and Tatum 1941). However, many questions remained outstanding about the physical nature of genes and how they are related to proteins. Phages played a central role in determining the physical structure of genes and their role in protein synthesis. According to Jacques Monod, a prominent pioneer in these early days of molecular biology: You know, the gene was something in the minds of people - especially of my generation which was as inaccessible, by definition, as the material of the galaxies. That experiments we were doing would involve an actual physical interaction between a compound in the cell and, actually the gene itself, was something extremely difficult to come to. (Judson 1979)

A series of experiments over the next decades would clarify the nature of the gene and its role in the cell. Phages played an essential role in many of these experiments, partly thanks to the foundational work laid by the American Phage Working Group led by Max Delbrück at Cold Spring Harbor in the USA.

Luria and Delbruck’s “Fluctuation Test” Shows That Spontaneous Mutations Occur in Bacterial Genes Delbrück, Salvador Luria, and Alfred Hershey would go on to share the Nobel Prize in Physiology or Medicine in 1969 for elucidating the “replication mechanism and genetics of viruses” using phages as their experimental model. Notable among their many discoveries that advanced the field of molecular biology was the “fluctuation test,” whereby Luria and Delbrück demonstrated the role of genetic mutation in bacterial phage resistance. In January 1943, Luria wrote to Delbrück to describe the experiment: The first written statement of it I find in a letter to Delbrück dated January 20: “I thought that a clean cut experiment would be to find out how the fluctuations in the number of

Bacteriophage Use in Molecular Biology and Biotechnology

487

[T1 phage]-resistants depend on the culture from which they come. That is: If I plate with [T1] ten samples of the same culture of B, I find numbers of resistants which fluctuate according to Poisson’s law. If I plate 10 samples of 10 different cultures of B, all containing the same amount of B, I find much larger fluctuations. If the resistants were produced on the plate, after contact with [T1], they should show the same fluctuations in both cases.” (Cairns et al. 1968)

In essence, they observed that resistant phenotypes could be detected in the absence of selective pressure rather than in response to it (Luria and Delbrück 1943). Importantly, this work supported Darwin’s theory of natural selection and showed that random mutation was involved in bacterial evolution as well as the evolution of more complex organisms. As another example, Delbrück’s “mutual exclusion principle” showed that when exposed to two different phages at the same time, a bacterial host could only be infected by one phage type and that only the infecting phage type was replicated following infection (Delbrück 1945). This observation motivated Delbrück and numerous others to further describe the phage lifecycle, and at the same time reveal new details regarding DNA transfer, gene regulation, lysogeny, and genetic recombination.

Hershey and Chase Demonstrate That DNA Is the Hereditary Material In 1952, several years after Delbrück’s initial observations on phage infection, Alfred Hershey and Martha Chase provided evidence that phages actively transferred their DNA to the host during infection, rather than transferring protein, effectively demonstrating DNA as the phage genetic material (Hershey and Chase 1952). It was generally assumed that proteins carried genetic information because DNA appeared too simple until the pioneering work of Avery, MacLeod, and McCarty. However, it required further experimental evidence to completely overturn that paradigm. Hershey and Chase conducted experiments to track the fate of phage proteins and DNA (the two primary components known to compose the phage) during bacterial infection. They accomplished this by radiolabeling T2 phage proteins and DNA with 35S and 32P, respectively. As shown in Fig. 8, in independent parallel experiments 35S- and 32P-labeled phages were used to infect E. coli. Following adsorption and infection, cultures were subjected to a shear force by blending to remove attached phages. Cultures were centrifuged to separate the bacteria. The supernatant and pellet were then separated and each tested for the presence of the radiolabel. 35S-labeled proteins were found to remain in the supernatant, while 32Plabeled DNA was observed to have been taken up by the host. This simple yet informative experiment provided strong evidence for the role of DNA as the viral genetic material and made plausible the concept of DNA as a heritable genetic molecule used throughout nature. Although Avery, MacLeod, and McCarty had already shown that DNA was the transformative agent able to convert Streptococcus pneumoniae from one phenotype to another, the Hershey-Chase experiment lent more weight to the hypothesis that DNA was the hereditary material (MacLeod and McCarty 1944).

488

N. Brown and C. Cox

Fig. 8 Outline of the Hershey and Chase experiment. Alfred Hershey and Martha Chase used phage T2, which is composed of protein and DNA, to determine whether protein or DNA is part of the hereditary material. (A) Phages were labeled with radioactive isotope: in one case, 35S was used to label protein molecules in the phage, in the other case, 32P was used to label DNA molecules in the phage. (B) One E. coli culture was infected by phage with 35S-labeled protein, and another was infected by phage with 32P-labeled DNA. (C) After infection, phage and bacteria were separated by disrupting phage-bacteria binding with a blender and removing the bacteria from phage in suspension by centrifugation. (D) Progeny phage was checked for the presence of 35S-labeled protein in the first case and 32P-labeled DNA in the second case. Only 32P-labeled DNA was detected in the progeny, not 35S-labeled protein, indicating that the hereditary material transmitted from parent to progeny phage was DNA, and not protein

Benzer Maps the Fine Structure of Genes in the Phage T4 rII Region As it became clear that DNA was the hereditary material, new questions arose as to the structure of the DNA molecule itself and how it worked. Watson and Crick’s description of the double-helix base-pairing structure of the DNA molecule in 1953 was the most important discovery in this area, but many other aspects of the physical structure of genes in relation to the DNA molecule remained outstanding after their discovery, and phages were critical to answering these questions. Phages and bacteria were so useful for answering questions about the nature of the genes encoded in DNA in part because they were so numerous and mutants of all kinds

Bacteriophage Use in Molecular Biology and Biotechnology

489

were easier to obtain compared to other organisms. A quote from Seymour Benzer in 1959 illustrates this: Mapping of a genetic structure is done by observing the recombination of its parts, and recombination involving parts of the structure that are very close together is a rare event. Observation of such rare events requires very many offspring and a selective trick for detecting the few individuals in which the event is recorded. It is for this reason that microorganisms are the material of choice for studies of genetic fine structure, and have made it feasible to extend the fineness of genetic mapping by orders of magnitude. In favorable systems, the attainable resolution reaches the level of the molecular subunits of the hereditary material and experimental testing of the linear arrangement of the finest structural details is therefore possible. (Benzer 1959)

It had been possible to place genes in order in a linear sequence within a genome using linkage mapping since the days of Thomas Hunt Morgan (Benzer 1959). Benzer wanted to know about “genetic fine structure,” that is, whether the subunits of a gene (what we now know to be nucleotides and what he referred to as the “molecular subunits of the hereditary material”) could be put in order in a linear sequence (instead of a branched structure) within the gene in an analogous fashion to linkage mapping. Benzer took advantage of the availability of the abundance of phage T4 progeny to isolate mutants in the rII region of phage T4. The rII region had previously been mapped by growing T4 on both E. coli strains B and K. T4 always grew on E. coli B, though T4 with mutations in the rII region would form plaques of a different size than those of wildtype T4. Although wildtype T4 could replicate on E. coli K, T4 with mutations in the rII region could not replicate to form plaques on that particular strain. Thus, a large number of T4 rII mutants could be isolated on E. coli B, indicated by distinct plaque sizes, and confirmed as T4 rII mutants because they would not infect E. coli K. A large stock of T4 rII mutants gave Benzer the luxury of selecting only those mutants that were stable (that would not revert to a wildtype T4 phenotype) and still have a large enough stock of mutants to conduct his experiment, making his results reliable. Despite having a large stock of mutants, he knew that he would never be able to score enough recombinant progeny from mated phage mutants to be able to calculate meaningful recombination frequencies at the genetic distances that he was interested in mapping, and thus could not use linkage mapping to define the genetic fine structure. Instead, he conducted phage matings between T4 rII mutants and simply scored them as being able to form plaques on E. coli K (scored as an “O” in Fig. 9) or not (scored as an “I” in Fig. 9). This is also known as a complementation test. If two T4 rII mutants were mated and produced progeny able to form plaques on E. coli K, then they did not share the same molecular subunits of the hereditary material and could “complement” each other to form viable progeny and plaques. If there were no progeny from T4 rII mutants mated on E. coli K, then the mutations shared some of the same molecular subunits of the hereditary material and could not complement each other. If enough mutants could be collected in the rII region, then it could be shown that the fine genetic structure of a gene could indeed be mapped to the resolution of the molecular subunit of the hereditary material (that is, to

490

N. Brown and C. Cox

Fig. 9 Mapping the fine genetic structure of the rII region of phage T4. (A) An illustration of six different mutations in a hypothetical branched fine genetic structure with a matching matrix. The matrix of complementary “I” and non-complementary “O” phage matings is arranged to maximize the number of series of unbroken series of non-complementary “O” phage matings. It shows that if the fine genetic structure is branched, then there would be broken series of non-complementary matings “O” present (such as at [4,6]). (B) An actual matrix of complementary and non-complementary phage matings of T4 rII mutants. The matrix shows that there are no unbroken series of non-complementary matings “O.” Thus, the fine genetic structure of the T4 rII region is linear, not branched. (C) Fine genetic structure of the rII region as inferred from a much larger set of rII mutant matings. Each black line indicates the physical extent of a mutation marked by non-complementary matings between pairs of mutants. The map indicates that there are two “cistrons,” or genes: “A cistron” and “B cistron” present in the rII region (these genes are presently called rIIA and rIIB). (Figure is redrawn from Benzer (1959))

nucleotide resolution). When the different T4 rII mutants scored in complementation tests were arranged in a matrix so that non-complementary mutants were nearest to each other on each axis, then it becomes apparent that all the non-complementary mutants (marked as “O”) formed unbroken series (Fig. 9B). An unbroken series of non-complementary mutants indicates that the mutants must occupy a linear space,

Bacteriophage Use in Molecular Biology and Biotechnology

491

or that genetic fine structure is indeed linear rather than branched (a branched structure would yield broken series of non-complementary mutants, as illustrated in Fig. 9A). Further, the fine genetic structure of the T4 rII region indicated that there were two distinct “cistrons” (Fig. 9C). Cistrons were defined as groups of mutants, any one of which could complement mutants of another group, but could not necessarily complement mutants within their own “cistron” group. The term “cistron” has since become synonymous with the term “gene.” Benzer had shown that the fine genetic structure was linear and for the first time mapped, with nucleotide precision in some instances, the location of a gene (later called rIIA) relative to another gene (rIIB) within what had previously simply been known as a single rII region.

The Discovery that Messenger RNA Is a Direct Product of Genes Although the “one gene-one enzyme” hypothesis had been developed in 1941, it was unclear in the early 1950s how genes were responsible for the enzymes that did the work within a cell. By the late 1950’s it had become clear that ribosomes – composed of RNA and protein – rather than genes were the sites of protein synthesis. Nonetheless, genes somehow directed protein synthesis and perhaps even encoded the proteins being synthesized. How did the information from genes become transferred to newly synthesized proteins? Several models were in vogue, including the idea that ribosomes themselves encoded the information for proteins. However, the RNA in ribosomes was not diverse enough in composition to match the diversity of genes or proteins, so some other kind of molecule was hypothesized to encode the information for proteins, perhaps another type of RNA. Several experiments done in the late 1940s and early 1950s showed that while bulk RNA metabolism occurred continuously in uninfected Escherichia coli cells, it seemed to stop in Escherichia coli infected with bacteriophages (Cohen 1948; Koch et al. 1952; Manson 1953). In 1953 Hershey showed that RNA metabolism in fact continued at a low level after bacteriophage infection (Hershey 1953). While exploring this phenomenon in 1956, Volkin and Astrachan used phage T2 to infect E. coli in 32P isotopically labeled media. RNA synthesized during phage infection was labeled with 32P from the media and when the composition of the labeled RNA was measured, it was found to differ from the composition of the RNA typically produced by Escherichia coli cells (Volkin and Astrachan 1956). Importantly, in later experiments, they noticed that the composition of the RNA produced during phage infection matched the composition of T2 genomic DNA, suggesting that the RNA played a role in T2 development (Astrachan and Volkin 1958). Finally, at Cambridge in 1960, Sydney Brenner and François Jacob hypothesized that these DNA-like-RNA molecules were unstable RNA intermediates that function as a bridge for the flow of genetic information from genes to protein synthesis. Brenner named this hypothetical class of RNA molecules, “messenger RNA,” and it was included in a formal model in, “Genetic regulatory mechanisms in the synthesis of proteins,” published by Jacob and Monod in 1961 (Jacob and Monod 1961;

492

N. Brown and C. Cox

Judson 1979). In 1961, Brenner, Jacob, and Meselson tested the messenger RNA model against two other models and determined that their model was correct; messenger RNA molecules carried genetic information from DNA residing in the host cell nucleus to the cytoplasm (Brenner et al. 1961). Further confirmation of messenger RNA was provided by Hall and Spiegelman when they used nucleic acid hybridization to show that RNA made after T2 infection was complementary to phage DNA (Hall and Spiegelman 1961). All these discoveries taken together gave clear evidence for the existence of mRNA and established its role in protein synthesis; and thus phage biology provided definitive verification of the messenger RNA link between DNA and the ribosome. With this knowledge, Francis Crick later put forward his “central dogma” of molecular biology in a famous paper published in Nature in 1970, which indicated that genetic information in the cell flows from DNA to RNA to proteins (Crick 1970).

The Determination of the Genetic Code Prior to the 1960s, the base composition of DNA had been determined to be composed of adenine, cytosine, guanine, and thymine. Proteins were known to be by-and-large composed of 20 amino acids. It was also known that there was a surprisingly constant relationship between the length of a gene and the length of its cognate protein. Facts like these suggested that there was a code that converted the information in genes into the information in proteins, but the nature of that code (the genetic code) was a mystery. In the 1950s, George Gamow, a nuclear physicist in Boulder, Colorado, had developed an interest in biochemistry, and in determining the genetic code in particular. He reasoned that the code for each amino acid in a protein could not be composed of a single nucleotide, but rather would more logically have a triplet nature, that is, three nucleotides per amino acid (Segre 2000). This was confirmed by Crick et al. in 1961 using the T4 phage rIIA and rIIB mutation assay developed by Seymour Benzer in 1959 (Benzer 1959). By using mutations and reversions to add and delete nucleotides in the phage T4 genome they found that triplet additions or deletions had minimal impact on the encoded protein, while smaller or larger combinations resulted in shifts out of the proper reading frame, which resulted in improperly translated proteins (Crick et al. 1961). They determined that there were 64 possible codons, 61 of which encoded specific amino acids, and three of which signaled the end of a given protein (stop codons). They further deduced that the triplet codons did not overlap, that the code was degenerate (i.e., some amino acids were coded for by more than one codon), and that each reading frame was translated from a specific starting point (a start codon).

The First Observation of Gene Regulation Due to a Repressor Factor (The PaJaMo Experiment) In 1959, the same year that Benzer mapped the fine genetic structure of the T4 rII region at Purdue University in the United States, a collaboration of three men at the

Bacteriophage Use in Molecular Biology and Biotechnology

493

Pasteur Institute in Paris, France, was busy developing a physical model of gene regulation. Anthony Pardee, François Jacob, and Jacques Monod conducted their famous PaJaMo experiment (named as a concatenation of the first two letters of their last names) in E. coli to show that production of the β-galactosidase enzyme from its gene could be induced by releasing control of a repressor factor encoded by another gene that would somehow otherwise inhibit production of the β-galactosidase enzyme (Pardee et al. 1959). Although the famous experiment was conducted in E. coli, the authors relied heavily upon their previous observations of what they believed to be an analogous repressor system in phage λ in order to interpret their results. In fact, according to interviews recorded by Judson in The Eighth Day of Creation: “If I had to tell the story, I would begin with that corridor [in Andre Lwoff’s attic at the Pasteur Institute], the work going on at each end - the phage business of Lwoff and Monod’s enzymes,” Jacob said. “The two things which merged.” Jacob was the agent of their merger . . . In Lwoff’s attic, the two ends of the corridor were getting closer together. Now lysogeny as well as galactosidase was being investigated in E. coli K12. Terminology had merged; thinking was on a converging course. By the early fifties, Lwoff later wrote, “Jacques Monod used to say that the induction of enzyme synthesis and of phage development are the expression of one and the same phenomenon. The statement looked paradoxical, but was, paradoxically, a remarkable intuition.” (Judson 1979)

Pardee, Jacob, and Monod drew key insights for their repressor model of β-galactosidase gene regulation from experiments that Jacob and Élie Wollman had done in 1956 with high-frequency recombination (Hfr) strains of E. coli K12 and λ lysogens (Jacob and Wollman 1956). Hfr strains had a conjugative plasmid integrated into their chromosomes, so that they could efficiently transfer their entire chromosome – but not the contents of their cytoplasm – into recipient cells lacking the plasmid via a conjugative mating bridge. Hfr strains or their recipient cells could also have a λ prophage integrated into their chromosomes. They observed that if an Hfr strain was a λ lysogen, but the recipient was not, then the recipient would be infected by the λ lysogen and produce λ phage. However, if the Hfr strain was not a λ lysogen, but the recipient strain was, then the recipient would not produce λ phage. Consequently, they reasoned that a factor responsible for preventing λ lysogens from producing λ phage was present in the cytoplasm of the λ lysogen prior to conjugation. When the λ lysogen was the Hfr strain, which transferred the λ prophage to the recipient, it failed to transfer this cytoplasmic factor to repress λ phage development, leading to λ phage development, cell lysis, and plaque formation in the recipient. Conversely, when the λ lysogen was the recipient strain in the conjugation, it already had sufficient levels of the repressor factor in its cytoplasm to prevent λ phage development. Their observations of galactosidase enzyme regulation in E. coli were similar enough to λ phage development that they believed galactosidase enzyme induction and λ phage induction were both due to a repressor encoded by its own gene and present in the cytoplasm of the cell: The most likely explanation is that, in the inducible strains, the synthesis of β-galactosidase is inhibited by a cytoplasmic repressor whose production is genetically controlled. The

494

N. Brown and C. Cox

induced synthesis would result from the release of the repression by a specific inducer. The analogy between this phenomenon and immunity of lysogenic cells is such that we can hardly escape the assumption that immunity also corresponds to the presence of a repressor in the cytoplasm of lysogenic cells. (Kutter and Sulakvelidze 2004)

Their discovery led them to ask two fruitful questions: “What is the chemical nature of the repressor? Should it be considered a primary or a secondary product of the gene?” and, “Does the repressor act at the level of the gene itself, or at the level of the cytoplasmic gene-product (enzyme-forming system)?” (Jacob and Monod 1961; Judson 1979). The answer to both questions awaited a high-suspense (for science) episode nicknamed “the race for the repressor,” and the answer to the second question came, according to Monod, as a surprise to everyone: “Everybody recognized that the regulation occurred, of course. We had proved it, and others had proved it. And that the regulation occurred on the basis of some sort of genetic determination. But everybody was assuming that the regulation was occurring somewhere lower down.” Lower down? “Further along the string of information. That it would turn out to operate right at the level of the gene itself was a – in fact it was hard for François and myself to really come to that conclusion and state it in black and white. It seemed so incredible.” (Judson 1979)

For their work on the genetic regulation of protein expression and viral lifecycles, Jacob, Monod, and Andre Lwoff shared the 1965 Nobel Prize in physiology or medicine.

The First Isolation and Characterization of a Repressor Factor Involved in Gene Regulation The “race for the repressor” lasted 7 years until two papers were published 2 months apart in 1966 that described two different protein repressors for two different genes. Although Jacob and Monod had demonstrated that regulation occurred, it still was unclear what the repressor was made of (though it was widely assumed to be a protein) and how it repressed the genes encoding an enzyme. The key challenge was to isolate enough of the repressor factor to be able to characterize it. Walter Gilbert and Benno Müller-Hill, well respected biochemists at Harvard, published the first paper describing the isolation of a repressor in October 1966. They had isolated the lac repressor of the lactose operon in E. coli by selecting for a mutant with a lac repressor that more tightly binds to the inducer molecule and tracking the repressor through a protein purification scheme with radioactive isopropyl-thio-galactoside (IPTG), which is an inducer of lac. Although they were able to isolate the lac repressor and show that it was a protein, rather than an RNA molecule as some hypothesized, they were not able to show how the repressor prevented the expression of the lactose operon. That was left to their competitor, Mark Ptashne. Ptashne was a relative newcomer to the field, having just completed his PhD. In an interview later, he said:

Bacteriophage Use in Molecular Biology and Biotechnology

495

Certainly the reason I went into molecular biology, probably one of the major reasons I went into science at all, was because it seemed to me that the repressor was the great problem . . . and it seemed to me that people who claimed to be trying to isolate the repressor and prove or disprove the theory, weren’t really serious. Weren’t really willing to take the kind of risks that were necessary . . . psychic risks. (Judson 1979)

Ptashne certainly took risks to isolate the repressor. Instead of focusing on the lac repressor, he chose to isolate the phage λ repressor and, “there was an incredible period when things would come and go. I’d have a result and then I couldn’t repeat it, and [Gilbert] would have a result and couldn’t repeat it, and so it was just absolutely hair-raising” (Judson 1979). By December 1966, 2 months after Gilbert and Müller-Hill had published their characterization of the lac repressor, Ptashne finally published the isolation and characterization of the λ repressor (Ptashne 1986). He isolated the repressor through three techniques that depended on the fact that he was trying to isolate a repressor from a phage in particular. First, he irradiated E. coli cells prior to infecting them with λ phages, which damaged E. coli nucleic acid and prevented any E. coli protein expression. Second, the E. coli host that he used was a λ lysogen that could not be induced by irradiation, which would prevent most phage protein expression (due to the presence of repressor protein in the cytoplasm) except for the repressor protein itself. Third, he infected the E. coli cells with a ratio of more than one λ phage for each cell, which maximized the number of repressor genes present in the cell for expression. These phages had mutations in the N gene, which was needed for the expression of phage genes other than the repressor. Thus, these phages could produce the λ repressor, but no other proteins. Protein expressed by the cells were labeled with radioactive amino acids supplied in the media. During purification of the repressor through chromatographic fractionation, the repressor fraction could be tracked by detecting the radioactivity. Unfortunately, Ptashne isolated a smaller concentration of λ repressor protein than Gilbert and Müller-Hill had of lac repressor and the λ repressor was not pure of other proteins. He had to demonstrate the specific presence of the λ repressor with the aid of a cleverly controlled labeling experiment. Nonetheless, the following year, Ptashne was able to use his system to show for the first time that a repressor acted by binding directly to a specific sequence in a DNA molecule. This was the critical experiment, which showed that, “the simplest model for the mechanism of action of the repressor is correct - namely, that the repressor blocks transcription from DNA to RNA by directly binding to DNA” (Ptashne 1967). Ptashne’s experiment was convincing because he had two different strains of lambda that only differed in their genomes at the region encoding the repressor gene. Their differing repressors could only repress gene expression in their native genome, but not the genome of the other lambda phage strain. By mixing the repressor protein fraction of one λ phage strain with the DNA from another, he could show that the repressor would not bind to DNA and sediment along with it in a sucrose gradient. Conversely, if he mixed the repressor protein fraction of a λ phage strain with DNA from that same strain, then the repressor would bind the DNA and sediment along with it in a sucrose gradient.

496

N. Brown and C. Cox

This showed the specificity of the repressor for a particular sequence on the DNA in the region of the λ genome that encodes for the repressor. In a sense, he had won the “race for the repressor,” because he had first shown the mechanism of gene repression. Regulation of gene expression by a protein repressor that binds to DNA is only one of a myriad of mechanisms of expression regulation that have since been discovered in many different organisms. However, phage λ provided a simple enough system to take the first important conceptual leap into the complex world of genetic regulation. Over the next several decades, Ptashne and several other lambdologists (researchers specializing in the study of phage λ) further described the λ repressor and the complex genetic regulatory circuit of which it is just a piece. The λ repressor is now known to be a homo-octomeric protein complex that binds to the operator region of the λ chromosome in two locations approximately 2.4 kb apart (Dodd et al. 2004). This protein-DNA interaction causes a loop to form in the DNA molecule, which can stop RNA polymerase from initiating transcription at two divergent promoter sites, depending on the concentration of the repressor and several other molecules involved in regulating the genetic circuit, and determine whether λ undergoes a lytic developmental pathway or remains as a prophage in the E. coli chromosome. Extensive characterization of the λ repressor and its genetic circuit has made it a paradigm of epigenetic switches. Ptashne has published a detailed book with artistic renderings of the molecular details of the “λ switch“that has become a classic in the field (Ptashne 1986).

Isolation of the r Termination Factor and Characterization of Termination/Antitermination in l Regulating gene expression is much more complicated that just initiating expression. Gene expression also must be terminated, and that also must be regulated somehow. In 1969, right on the heels of the discovery of the first repressors, Jeffrey Roberts isolated the first transcriptional terminator, ρ-factor, from E. coli and demonstrated its activity on the λ chromosome (Roberts 1969). Although not much was known about λ transcription at the time, E. coli RNA polymerase could be isolated and used to set up in vitro transcription reactions, producing messenger RNA (mRNA) from a template DNA molecule. Roberts showed that by adding isolated ρ-factor to the in vitro transcription reaction, he could change the size of the two main mRNA molecules produced. Adding ρ-factor gave rise to two mRNA molecules smaller than were produced in vitro without ρ-factor. The population of mRNA molecules produced under the effect of ρ-factor was composed of a consistent size rather than a range of different sizes, which suggested that ρ-factor acted at specific locations. Roberts identified the locations on the lambda chromosome where ρ-factor terminates each transcript with some precision and further explained interesting behavior of the λ N protein which had earlier been observed. The N protein is produced early in the λ genetic program and stimulates the transcription of genes later in the λ genetic program. Roberts noticed that genes for which the N protein stimulated

Bacteriophage Use in Molecular Biology and Biotechnology

497

transcription were distal to the locations on the lambda chromosome where ρ-factor terminates each transcript. He hypothesized that N likely acts by preventing ρ-factor from terminating each of the two major λ transcripts. The antitermination hypothesis was confirmed over time, and involves a suite of E. coli proteins called the Nus proteins. These proteins interact with λ N and E. coli RNA polymerase to cause RNA polymerase to ignore termination signals and accelerate transcription (Adhya et al. 1974).

The Construction of a Synthetic Genetic Regulatory Circuit that Exhibits Complex Behavior Drawing heavily on detailed characterization of the λ switch, in 2000 Elowitz and Leibler built the first synthetic genetic regulatory circuit that exhibited novel, complex behavior (Elowitz and Leibler 2000). They named their synthetic genetic regulatory circuit “the repressilator,” a portmanteau of “repressor” and “oscillator.” The repressilator is a high-copy plasmid designed to hold the genes for three different repressors that are made more unstable by adding a protease tag that targets them for proteolysis (named tetR-lite, lacI-lite, and cI-lite). These repressors each have strong pL promoters taken from λ to drive their expression, but the promoter for each repressor contains an operator site where one of the other repressors could bind to the DNA and inhibit expression of the downstream repressor gene from the pL promoter. This created a rock-paper-scissors effect wherein each repressor repressed one of the other repressors: a simple regulatory circuit. The repressilator plasmid was introduced into an E. coli cell alongside a reporter plasmid with a semistable green fluorescent protein (GFP) gene expressed from a pL promoter with a tetR-lite operator. The repressilator circuit in each cell of the E. coli culture was synchronized with the others by adding a pulse of IPTG, which would relieve repression by the lacI-lite repressor simultaneously across the bacterial population. The culture was then observed under an epifluorescent microscope for patterns in GFP expression. As hypothesized, GFP on the reporter plasmid was expressed in pulses due to the design and regulation of the repressilator circuit. This marked the first instance that a synthetic genetic regulatory circuit exhibited new, complex, and predictable behavior. Since 2000, there have been many more forays into engineering genetic regulatory circuits in order to both understand them and get them to perform in a new but predictable way. This new area of research falls under the umbrella of synthetic biology, itself a new area of biological research broadly devoted to engineering biological systems. Phages play a pivotal role in this new field, because their relative simplicity compared to other organisms and the abundance of scientific literature describing their genetic regulation makes them useful as tools and model systems of complex gene regulation. Five years later, in addition to publishing a well-known manifesto describing the field of synthetic biology (Endy 2005), Endy and his colleagues exploited E. coli phage T7 to show that he could rearrange the organization of genes in the T7 genome and explore the effects on the new T7 phage he created.

498

N. Brown and C. Cox

Refactoring Phages to Learn About Complex Genetic Regulatory Circuits at the Organismal Level In 2005, Chan, Kosuri, and Endy published their unprecedented effort to refactor a genome (Chan et al. 2005). “Refactoring” is a term borrowed from software engineering, and it means, “to improve the internal structure of an existing system for future use, while simultaneously maintaining external system function” (Chan et al. 2005). In more concrete terms, Chan et al. intended to rewrite large parts of the T7 genome without preventing the modified T7 from propagating on E. coli. In particular, they decided to alter a common feature of phage genomes: overlapping genes. Phage genomes are small and must adhere to size restrictions due to the limited space in the capsid for packaging genomic DNA. Genes often overlap in phage genomes, presumably to save space, but this may also have effects on genetic regulation. Chan et al. carefully, systematically disentangled genes in roughly one quarter of the T7 genome (11,515 of 39,937 bp), making sure that none of them overlapped. Genetic changes were made on synthetic fragments of DNA in vitro and using E. coli cloning vectors, then the DNA fragments were introduced to E. coli together by transfection, giving rise to refactored phage progeny. They named their refactored phage “T7.1,” borrowing from software versioning conventions. Sequencing the T7.1 phage genome revealed a handful of unintended mutations introduced during refactoring, but those mutations appeared to have a minimal impact on phage growth. When they cultured phage T7.1 on E. coli, they observed nearly identical lysis times as those of wildtype T7 phage at 30 °C. Phage T7.1 formed plaques on E. coli plates incubated at 37 °C as did phage T7, but the T7.1 plaques were much smaller. Although the T7.1 growth phenotype was not identical to that of T7, it was completely viable and offers a new, simplified resource for studying complex genetic regulatory networks in a whole phage, wherein the effects of gene overlap have been removed for approximately one quarter of the genome. Perhaps more importantly, they demonstrated that genome-wide refactoring was not necessarily as prohibitively difficult as previously thought. Since this first effort, many genomes – including several phage genomes – and gene regulatory circuits have been refactored in a similar fashion in an effort to more fully understand gene interactions and phenotypes with complex genetic origins (Ghosh et al. 2012; Jaschke et al. 2012; Temme et al. 2012).

Conclusion There is a tired refrain often heard in molecular biology circles that we already know everything about phage biology and that phages have already told us all the secrets of biology that they hold. A member of Delbrück’s American Phage Working Group, Gunther Stent, proclaimed that the broader field of molecular biology had already delivered its greatest miracle before 1968: characterizing the physical nature of the gene. He laid out his argument in the article, “That Was the Molecular Biology That

Bacteriophage Use in Molecular Biology and Biotechnology

499

Was” (Stent 1968) and expanded on that view in his book, “The Coming of the Golden Age: A View of the End of Progress.” However, if one traces the chain of discoveries made in molecular biology as we have for the molecular biology of phages, then one sees a trend that predicts many more important discoveries to be made. Stent was partly right in predicting the rise of neuroscience as a frontier of discovery and the importance of molecular biology to that new field, as evidenced by the use of the Cre-lox system for tracing neuronal development in the brainbow mouse. However, there are other frontiers of discovery just as momentous as neuroscience that are sprouting from molecular and phage biology or experiencing a renaissance thanks to them. Some of these frontiers have even yielded important discoveries about “the Molecular Biology That Was.”

Systems Biology Systems biology, the modeling of complex molecular interactions within or between cells, tissues, and organisms in larger frameworks such as ecosystems or symbioses, is heavily dependent on molecular biology. The field has contributed fundamental insights into molecular details of Crick’s central dogma that have only been observable with large-scale data generation from new tools such as tandem mass spectrometers and high-throughput DNA sequencers. For example, proteomic and transcriptomic measurements in human cell lines have clearly shown that the phenotype of the cell does not correlate with the transcriptome as well as with the proteome (Battle et al. 2015). This directly addresses Crick’s central dogma, indicating that while transcriptional regulation is critical to the flow of information from gene to protein, changes in the cell phenotype after development is complete are largely dictated by post-transcriptional regulation. Phage biology is also important to systems biology. Phage λ and the regulatory switch governing its lytic/lysogenic lifestyle change have been well enough studied over the past several decades that λ serves as a genetic model coupled with quantitative transcriptional data sufficient for evaluating mathematical models of transcription, including even models of cancer (Ackers et al. 1982; Arkin et al. 1998; Wang et al. 2013).

Evolution New insights into molecular mechanisms of evolution made possible by a deluge of DNA sequencing data are enabling refinements and expansions of Darwin’s theory. An example of this is the new field of evolutionary developmental biology, which one might partly trace to studies of the developmental program of induced prophage λ (Oppenheim et al. 2005). Another example of this is the Escherichia coli LongTerm Evolution Experiment and the discipline of experimental evolution which it vindicated (Lenski 2017). A number of evolutionary experiments use phages as model organisms to understand contingency, speciation, and other fundamental

500

N. Brown and C. Cox

evolutionary questions (Meyer et al. 2012, 2016). Yet a third very recent example of a renaissance in evolutionary biology thanks to molecular biology is threedimensional protein structure prediction from large alignments of homologous amino acid sequences with direct coupling analysis (Weigt et al. 2009). In combination with a recent breakthrough in mapping the reticulate relationships of all DNA virus proteins (Iranzo et al. 2016) and recent developments in cryoelectronmicroscopic virion structure determination (Fokine et al. 2005), it may one day be possible to trace the evolutionary lineage of viruses and address longstanding questions of viral origins (Haldane 1980; Koonin 2009). Phages will necessarily play an essential role.

Ecology Phages on Earth are famously estimated to outnumber, at 1031 (Hendrix et al. 1999), the stars in the universe, with a popular estimate at 1024 (Howell 2014). Their ecology is important in biogeochemical nutrient cycling, disease, and bacterial evolution (see ▶ “Bacteriophage Ecology” chapter). Phage ecology has expanded tremendously with the application of high-throughput sequencing to environmental DNA samples, called metagenomics. Given that only an estimated few percent of bacteria can be cultured (Staley and Konopka 1985), only a few percent of phages can be grown as plaques on bacterial cultures or as prophages within lysogens. Metagenomics freed researchers from this limitation by allowing them to sequence the genomes of organisms that cannot be cultured and infer their characteristics from the genomes. Breitbart et al. pioneered this method in 2002 for the first metavirome, which is a metagenome of viral DNA from the environment (Breitbart et al. 2002). The result since then has been spectacular, with the vast majority of viral (phage) genome and protein diversity being novel when compared to the DNA sequence databases. Given the unfortunate nickname “viral dark matter,” these genomes and proteins are a new frontier in molecular biology that invites discovery of untold novelties (Reyes et al. 2012). These frontiers in biology share complexity exemplified by reticulate relationships, feedback signaling, extreme diversity, and multiple interacting levels of abstraction. Phages will continue to yield important discoveries in molecular biology and related fields thanks to their relative simplicity among biological entities, their unmatched abundance and ubiquity, and their evolutionary antiquity. Each of these attributes make phages ideal tools in our effort to break down complex biological phenomena into comprehensible pieces. As we struggle to understand complexities like emergent phenomena in systems biology, track the role of contingency in evolution, characterize the diversity of microorganisms in ecology, and build complex regulatory networks in synthetic biology, the humble phage will continue to give us important footholds of understanding like they did during the search for the gene in the “Golden Age of Molecular Biology.”

Bacteriophage Use in Molecular Biology and Biotechnology

501

Cross-References ▶ Detection of Bacteriophages: Statistical Aspects of Plaque Assay

References Abshire T, Brown J, Ezzell J (2005) Production and validation of the use of gamma phage for identification of Bacillus anthracis. J Clin Microbiol 43:4780–4788 Ackers GK, Johnson AD, Shea MA (1982) Quantitative model for gene regulation by lambda phage repressor. Proc Natl Acad Sci 79:1129–1133 Adhya S, Gottesman M, De Crombrugghe B (1974) Release of polarity in Escherichia coli by gene N of phage λ: termination and antitermination of transcription. Proc Natl Acad Sci 71:2534–2538 Arber W, Dussoix D (1962) Host specificity of DNA produced by Escherichia coli: I. Host controlled modification of bacteriophage λ. J Mol Biol 5:18–36 Arkin A, Ross J, McAdams HH (1998) Stochastic kinetic analysis of developmental pathway bifurcation in phage λ-infected Escherichia coli cells. Genetics 149:1633–1648 Astrachan L, Volkin E (1958) Properties of ribonucleic acid turnover in T2-infected Escherichia coli. Biochim Biophys Acta 29:536–544 Atlung T, Nielsen A, Rasmussen LJ, Nellemann LJ, Holm F (1991) A versatile method for integration of genes and gene fusions into the λ attachment site of Escherichia coli. Gene 107:11–17 Auvray F, Coddeville M, Ritzenthaler P, Dupont L (1997) Plasmid integration in a wide range of bacteria mediated by the integrase of Lactobacillus delbrueckii bacteriophage mv4. J Bacteriol 179:1837–1845 Bail O (1922) Elementarbakteriophagen des Shigabacillus. Wien klin Wochenschr: 743 Barner HD, Cohen SS (1954) The induction of thymine synthesis by T2 infection of a thymine requiring mutant of Escherichia coli. J Bacteriol 68:80 Battle A, Khan Z, Wang SH, Mitrano A, Ford MJ, Pritchard JK, Gilad Y (2015) Impact of regulatory variation from RNA to protein. Science 347:664–667 Beadle GW, Tatum EL (1941) Genetic control of biochemical reactions in Neurospora. Proc Natl Acad Sci 27:499–506 Benzer S (1959) On the topology of the genetic fine structure. Proc Natl Acad Sci 45:1607–1620 Bertani G (1951) Studies on lysogenesis I. The mode of phage liberation by lysogenic Escherichia coli. J Bacteriol 62:293 Bertani G (2004) Lysogeny at mid-twentieth century: P1, P2, and other experimental systems. J Bacteriol 186:595–600 Bertani G, Weigle J (1953) Host controlled variation in bacterial viruses. J Bacteriol 65:113 Betz UA, Voßhenrich CA, Rajewsky K, Müller W (1996) Bypass of lethality with mosaic mice generated by Cre–loxP-mediated recombination. Curr Biol 6:1307–1316 Botstein D, White RL, Skolnick M, Davis RW (1980) Construction of a genetic linkage map in man using restriction fragment length polymorphisms. Am J Hum Genet 32:314 Breitbart M, Salamon P, Andresen B, Mahaffy JM, Segall AM, Mead D, Azam F, Rohwer F (2002) Genomic analysis of uncultured marine viral communities. Proc Natl Acad Sci 99:14250–14255 Brenner S, Jacob F, Meselson M (1961) An unstable intermediate carrying information from genes to ribosomes for protein synthesis. Nature 190:576–581 Cairns J, Stent GS, Watson JD (1968) Phage and the origins of molecular biology. J Hist Biol 1(1):155–161

502

N. Brown and C. Cox

Calef E, Licciardello G (1960) Recombination experiments on prophage host relationships. Virology 12:81–103 Campbell AM (1963) Episomes. Adv Genet 11:101–145 Campbell AM (1993) Thirty years ago in genetics: prophage insertion into bacterial chromosomes. Genetics 133:433 Chan LY, Kosuri S, Endy D (2005) Refactoring bacteriophage T7. Mol Syst Biol 1:1–10 Cherry W, Davis BR, Edwards PR, Hogan R, others (1954) A simple procedure for the identification of the genus Salmonella by means of a specific bacteriophage. J Lab Clin Med 44:51–55 Cohen SS (1948) The synthesis of bacterial viruses II. The origin of the phosphorus found in the desoxyribonucleic acids of the T2 and T4 bacteriophages. J Biol Chem 174:295–303 Cohen SS, Barner HD (1954) Studies on unbalanced growth in Escherichia coli. Proc Natl Acad Sci 40:885–893 Cohen SS, Flaks JG, Barner HD, Loeb MR, Lichtenstein J (1958) The mode of action of 5-fluorouracil and its derivatives. Proc Natl Acad Sci 44:1004–1012 Collins J, Hohn B (1978) Cosmids: a type of plasmid gene-cloning vector that is packageable in vitro in bacteriophage lambda heads. Proc Natl Acad Sci 75:4242–4246 Cox CR (2012) 10 Bacteriophage-based methods of bacterial detection and identification. In: Hyman P, Abedon ST (Eds) Bacteriophages in health and disease, vol 24. CABI, Oxfordshire, UK, p 134 Cox CR, Rees JC, Voorhees KJ (2012) Modeling bacteriophage amplification as a predictive tool for optimized MALDI-TOF MS-based bacterial detection. J Mass Spectrom 47:1435–1441 Crick F (1970) Central dogma of molecular biology. Nature 227:561–563 Crick F, Barnett L, Brenner S, Watts-Tobin RJ (1961) General nature of the genetic code for proteins. Macmillan Journals, London d’Herelle F (1917) Sur un microbe invisible antagoniste des bacilles dysentériques. CR Acad Sci Paris 165:373–375 d’Herelle F (1931) Bacterial mutations. Yale J Biol Med 4:55 Delbrück M (1945) Interference between bacterial viruses: III. The mutual exclusion effect and the depressor effect. J Bacteriol 50(2):151 Dodd IB, Shearwin KE, Perkins AJ, Burr T, Hochschild A, Egan JB (2004) Cooperativity in longrange gene regulation by the λ CI repressor. Genes Dev 18:344–354 Dussoix D, Arber W (1962) Host specificity of DNA produced by Escherichia coli: II. Control over acceptance of DNA from infecting phage λ. J Mol Biol 5:37–49 Ellis EL, Delbrück M (1939) The growth of bacteriophage. J Gen Physiol 22:365–384 Elowitz MB, Leibler S (2000) A synthetic oscillatory network of transcriptional regulators. Nature 403:335 Endy D (2005) Foundations for engineering biology. Nature 438:449 Feiss M, Widner W, Miller G, Johnson G, Christiansen S (1983) Structure of the bacteriophage lambda cohesive end site: location of the sites of terminase binding (cosB) and nicking (cosN). Gene 24:207–218 Felix A (1956) Phage typing of Salmonella typhimurium: its place in epidemiological and epizootiological investigations. Microbiology 14:208–222 Fisk RT (1942) Studies on staphylococci: I. occurrence of bacteriophage carriers among strains of Staphylococcus aureus. J Infect Dis 71:153–160 Flaks JG, Cohen SS (1959) Virus-induced acquisition of metabolic function I. Enzymatic formation of 5-hydroxymethyldeoxycytidylate. J Biol Chem 234:1501–1506 Fokine A, Leiman PG, Shneider MM, Ahvazi B, Boeshans KM, Steven AC, Black LW, Mesyanzhinov VV, Rossmann MG (2005) Structural and functional similarities between the capsid proteins of bacteriophages T4 and HK97 point to a common ancestry. Proc Natl Acad Sci U S A 102:7163–7168 Ghosh D, Kohli AG, Moser F, Endy D, Belcher AM (2012) Refactored M13 bacteriophage as a platform for tumor cell imaging and drug delivery. ACS Synth Biol 1:576–582

Bacteriophage Use in Molecular Biology and Biotechnology

503

Gildmeister E, Herzberg K (1924) Zur theorie der bakteriophagen (d’Herelle Lysine). 6. Mitteilung über das d’Herellesche phanomen. Zentr Bakteriol Parasitenk I Abt Orig 93:402–420 Gill P, Jeffreys AJ, Werrett DJ (1985) Forensic application of DNA ‘fingerprints’. Nature 318:577–579 Gingery R, Echols H (1967) Mutants of bacteriophage lambda unable to integrate into the host chromosome. Proc Natl Acad Sci 58:1507–1514 Gold M, Hurwitz J (1963) The enzymatic methylation of the nucleic acids. Cold Spring Harb Symp Quant Biol 28:149–156 Guarneros G, Echols H (1970) New mutants of bacteriophage λ with a specific defect in excision from the host chromosome. J Mol Biol 47:565–574 Haldane JBS (1980) The origin of life. In: Goldsmith D (Ed) The quest for extraterrestrial life. University Science Books, Mill Valley, CA, p 28. Hall BD, Spiegelman S (1961) Sequence complementarity of T2-DNA and T2-specific RNA. Proc Natl Acad Sci 47:137–146 Heidelberger C, Chaudhuri N, Danneberg P, Mooren D, Griesbach L, Duschinsky R, Schnitzer R, Pleven E, Scheiner J (1957) Fluorinated pyrimidines, a new class of tumour-inhibitory compounds. Nature 179:663–666 Hendrix RW, Smith MC, Burns RN, Ford ME, Hatfull GF (1999) Evolutionary relationships among diverse bacteriophages and prophages: all the world’sa phage. Proc Natl Acad Sci 96:2192–2197 Hershey AD (1953) Nucleic acid economy in bacteria infected with bacteriophage T2. J Gen Physiol 37:1–23 Hershey AD, Chase M (1952) Independent functions of viral protein and nucleic acid in growth of bacteriophage. J Gen Physiol 36:39–56 Hillier K (2006) Babies and bacteria: phage typing, bacteriologists, and the birth of infection control. Bull Hist Med 80:733–761 Hoess RH, Ziese M, Sternberg N (1982) P1 site-specific recombination: nucleotide sequence of the recombining sites. Proc Natl Acad Sci 79:3398–3402 Holland R, Wilkes J, Rafii F, Sutherland J, Persons C, Voorhees K, Lay J (1996) Rapid identification of intact whole bacteria based on spectral patterns using matrix-assisted laser desorption/ ionization with time-of-flight mass spectrometry. Rapid Commun Mass Spectrom 10:1227–1232 Howell ES (2014) How many stars are in the universe? Space.com, May 31 Iranzo J, Krupovic M, Koonin EV (2016) The double-stranded DNA virosphere as a modular hierarchical network of gene sharing. MBio 7:e00978–e00916 Jackson DA, Symons RH, Berg P (1972) Biochemical method for inserting new genetic information into DNA of Simian Virus 40: circular SV40 DNA molecules containing lambda phage genes and the galactose operon of Escherichia coli. Proc Natl Acad Sci 69:2904–2909 Jacob F, Monod J (1961) Genetic regulatory mechanisms in the synthesis of proteins. J Mol Biol 3:318–356 Jacob F, Wollman E (1956) Sur les processus de conjugaison et de recombinaison chez Escherichia coli. 1. Linduction par conjugaison ou induction zygotique. Ann Inst Pasteur (Paris) 91:486–510 Jaschke PR, Lieberman EK, Rodriguez J, Sierra A, Endy D (2012) A fully decompressed synthetic bacteriophage øX174 genome assembled and archived in yeast. Virology 434:278–284 Judson HF (1979) The eighth day of creation. Touchstone Books, New York, p 550 Kelly TJ, Smith HO (1970) A restriction enzyme from Hemophilus influenzae: II. Base sequence of the recognition site. J Mol Biol 51:393–409 Kikuchi Y, Nash HA (1978) The bacteriophage lambda int gene product. A filter assay for genetic recombination, purification of int, and specific binding to DNA. J Biol Chem 253:7149–7157 Koch AL, Putnam FW, Evans E Jr (1952) The purine metabolism of Escherichia coli. J Biol Chem 197:105–112 Koob M, Grimes E, Szybalski W (1988) Conferring operator specificity on restriction endonucleases. Science 241:1084–1087

504

N. Brown and C. Cox

Koonin EV (2009) On the origin of cells and viruses. Ann N Y Acad Sci 1178:47–64 Kutter E, Sulakvelidze A (2004) Bacteriophages: biology and applications. CRC Press, New York Landy A, Ruedisueli E, Robinson L, Foeller C, Ross W (1974) Digestion of deoxyribonucleic acids from bacteriophage T7, λ, and ϖ80h with site-specific nucleases from Hemophilus influenzae strain Rc and strain Rd. Biochemistry 13:2134–2142 Lay JO (2001) MALDI-TOF mass spectrometry of bacteria. Mass Spectrom Rev 20:172–194 Lederberg S (1957) Suppression of the multiplication of heterologous bacteriophages in lysogenic bacteria. Virology 3:496–513 Lederberg J, Lederberg EM (1952) Replica plating and indirect selection of bacterial mutants. J Bacteriol 63:399 Lederberg EM, Lederberg J (1953) Genetic studies of lysogenicity in Escherichia coli. Genetics 38:51 Lee MH, Pascopella L, Jacobs WR, Hatfull GF (1991) Site-specific integration of mycobacteriophage L5: integration-proficient vectors for Mycobacterium smegmatis, Mycobacterium tuberculosis, and bacille Calmette-Guerin. Proc Natl Acad Sci 88:3111–3115 Lenski RE (2017) What is adaptation by natural selection? Perspectives of an experimental microbiologist. PLoS Genet 13:e1006668 Livet J, Weissman TA, Kang H, Draft RW, Lu J, Bennis RA, Sanes JR, Lichtman JW (2007) Transgenic strategies for combinatorial expression of fluorescent proteins in the nervous system. Nature 450:56–62 Loenen WA, Dryden DT, Raleigh EA, Wilson GG, Murray NE (2014) Highlights of the DNA cutters: a short history of the restriction enzymes. Nucleic Acids Res 42:3–19 Luria SE, Delbrück M (1943) Mutations of bacteria from virus sensitivity to virus resistance. Genetics 28:491 Luria SE, Human ML (1952) A nonhereditary, host-induced variation of bacterial viruses. J Bacteriol 64:557 Lwoff A (1953) Lysogeny. Bacteriol Rev 17:269 Lwoff A (1966) The prophage and I. In: Phage and the origins of molecular biology. Cold Spring Harbor Laboratory Press, New York, pp 88–99 MacLeod AO, McCarty M (1944) Studies of the chemical nature of the substance inducing transformation of pneumococcal types. Induction of transformation by a deoxyribonucleic acid fraction isolated from pneumococcus type III. J Exp Med 79:137–158 Madonna AJ, Cuyk SV, Voorhees KJ (2003) Detection of Escherichia coli using immunomagnetic separation and bacteriophage amplification coupled with matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Rapid Commun Mass Spectrom 17:257–263 Madonna AJ, Voorhees KJ, Rees JC (2007) Method for detection of low concentrations of a target bacterium that uses phages to infect target bacterial cells. U.S. Patent US7166425B2 Manson LA (1953) The metabolism of ribonucleic acid in normal and bacteriophage infected Escherichia coli. J Bacteriol 66:703 Mertz JE, Davis RW (1972) Cleavage of DNA by R1 restriction endonuclease generates cohesive ends. Proc Natl Acad Sci 69:3370–3374 Meyer JR, Dobias DT, Weitz JS, Barrick JE, Quick RT, Lenski RE (2012) Repeatability and contingency in the evolution of a key innovation in phage lambda. Science 335:428–432 Meyer JR, Dobias DT, Medina SJ, Servilio L, Gupta A, Lenski RE (2016) Ecological speciation of bacteriophage lambda in allopatry and sympatry. Science 354(6317):1301–1304. https://doi.org/ 10.1126/science.aai8446 Nei M, Tajima F (1981) DNA polymorphism detectable by restriction endonucleases. Genetics 97:145–163 Nicolle P, Le Minor L, Buttiaux R, Ducrest P (1952) Phage typing of Escherichia coli isolated from cases of infantile gastroenteritis. II. Relative frequency of types in different areas and the epidemiological value of the method. Bull Acad Natl Med 136:483–485

Bacteriophage Use in Molecular Biology and Biotechnology

505

Nkrumah LJ, Muhle RA, Moura PA, Ghosh P, Hatfull GF, Jacobs WR, Fidock DA (2006) Efficient site-specific integration in Plasmodium falciparum chromosomes mediated by mycobacteriophage Bxb1 integrase. Nat Methods 3:615–621 Oppenheim AB, Kobiler O, Stavans J, Court DL, Adhya S (2005) Switches in bacteriophage lambda development. Annu Rev Genet 39:409–429 Pardee AB, Jacob F, Monod J (1959) The genetic control and cytoplasmic expression of “inducibility” in the synthesis of β-galactosidase by E. coli. J Mol Biol 1:165–178 Pfankuch E, Kausche G (1940) Isolierung und, übermikroskopische Abbildung eines Bakteriophagen. Naturwissenschaften 28:46–46 Pleceas P, Brandis H (1974) Rapid group and species identification of enterococci by means of tests with pooled phages. J Med Microbiol 7:529–534 Postic C, Shiota M, Niswender KD, Jetton TL, Chen Y, Moates JM, Shelton KD, Lindner J, Cherrington AD, Magnuson MA (1999) Dual roles for glucokinase in glucose homeostasis as determined by liver and pancreatic β cell-specific gene knock-outs using Cre recombinase. J Biol Chem 274:305–315 Ptashne M (1986) A genetic switch: gene control and phage lambda. Cell Press and Blackwell Scientific Publications, Cambridge, MA Ptashne M (1967) Specific binding of the lambda phage repressor to lambda DNA. Nature 214:232–234 Rees JC, Voorhees KJ (2005) Simultaneous detection of two bacterial pathogens using bacteriophage amplification coupled with matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Rapid Commun Mass Spectrom 19:2757–2761 Reyes A, Semenkovich NP, Whiteson K, Rohwer F, Gordon JI (2012) Going viral: next generation sequencing applied to human gut phage populations. Nat Rev Microbiol 10:607 Roberts JW (1969) Termination factor for RNA synthesis. Nature 224:1168–1174 Roszczyk E, Goodgal S (1975) Methylase activities from Haemophilus influenzae that protect Haemophilus parainfluenzae transforming deoxyribonucleic acid from inactivation by Haemophilus influenzae endonuclease R. J Bacteriol 123:287–293 Roy PH, Smith HO (1973) DNA methylases of Hemophilus influenzae Rd: II. Partial recognition site base sequences. J Mol Biol 81:445–459 Ruska H (1940) Die Sichtbarmachung der bakteriophagen lyse im übermikroskop. Naturwissenschaften 28:45–46 Sauer B (1987) Functional expression of the cre-lox site-specific recombination system in the yeast Saccharomyces cerevisiae. Mol Cell Biol 7:2087–2096 Sauer B, Henderson N (1989) Cre-stimulated recombination at loxP-containing DNA sequences placed into the mammalian genome. Nucleic Acids Res 17:147–161 Schlesinger M (1936) The Feulgen reaction of the bacteriophage substance. Nature 138:508 Segre G (2000) The big bang and the genetic code. Nature 404:437–437 Seng P, Drancourt M, Gouriet F, La Scola B, Fournier P-E, Rolain JM, Raoult D (2009) Ongoing revolution in bacteriology: routine identification of bacteria by matrix-assisted laser desorption ionization time-of-flight mass spectrometry. Clin Infect Dis 49:543–551 Shcheglova MK, Neidbailik IN (1968) [Experience in phage typing of Listeria]. Veterinariia 45:102–103 Staley JT, Konopka A (1985) Measurement of in situ activities of nonphotosynthetic microorganisms in aquatic and terrestrial habitats. Annu Rev Microbiol 39:321–346 Stent GS (1968) That was the molecular biology that was. Science 160:390–395 Sternberg N, Hamilton D, Hoess R (1981a) Bacteriophage P1 site-specific recombination: II. Recombination between loxP and the bacterial chromosome. J Mol Biol 150:487–507 Sternberg N, Hamilton D, Austin S, Yarmolinsky M, Hoess R (1981b) Site-specific recombination and its role in the life cycle of bacteriophage P1. Cold Spring Harb Symp Quant Biol 45:297–309

506

N. Brown and C. Cox

Stewart G, Jassim S, Denyer SP, Newby P, Linley K, Dhir V (1998) The specific and sensitive detection of bacterial pathogens within 4 h using bacteriophage amplification. J Appl Microbiol 84:777–783 Temme K, Zhao D, Voigt CA (2012) Refactoring the nitrogen fixation gene cluster from Klebsiella oxytoca. Proc Natl Acad Sci 109:7085–7090 Thal E, Nordberg B (1968) On the diagnostic of Bacillus anthracis with bacteriophages. Berl Munch Tierarztl Wochenschr 81:11 Thomason L, Calendar R, Ow D (2001) Gene insertion and replacement in Schizosaccharomyces pombe mediated by the Streptomyces bacteriophage fC31 site-specific recombination system. Mol Gen Genomics 265:1031–1038 Upholt WB (1977) Estimation of DNA sequence divergence from comparison of restriction endonuclease digests. Nucleic Acids Res 4:1257–1266 Volkin E, Astrachan L (1956) Phosphorus incorporation in Escherichia coli ribonucleic acid after infection with bacteriophage T2. Virology 2:149–161 Wallmark G, Laurell G (1951) Phage typing of Staphylococcus aureus some bacteriological and clinical observations. Acta Pathol Microbiol Scand 30:109–114 Wang G, Zhu X, Hood L, Ao P (2013) From phage lambda to human cancer: endogenous molecular-cellular network hypothesis. Quant Biol 1:32–49 Weigt M, White RA, Szurmant H, Hoch JA, Hwa T (2009) Identification of direct residue contacts in protein – protein interaction by message passing. Proc Natl Acad Sci 106:67–72 Weisberg RA, Landy A (1983) Site-specific recombination in phage lambda. Cold Spring Harb Monogr Arch 13:211–250 Wyatt G, Cohen SS (1953) The bases of the nucleic acids of some bacterial and animal viruses: the occurrence of 5-hydroxymethylcytosine. Biochem J 55:774 Xiao Y, Weaver DT (1997) Conditional gene targeted deletion by Cre recombinase demonstrates the requirement for the double-strand break repair Mre11 protein in murine embryonic stem cells. Nucleic Acids Res 25:2985–2991

Detection of Bacteriophages: Phage Plaques Stephen T. Abedon

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plaques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Initiating Plaques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bacterial Lawns, Plaques, and Spots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plaque-Forming Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Too Many or Too Few Plaques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plaques Versus Spots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plaque Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phage Clumped Dispersion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bacterial Clumped Dispersion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plaque Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plaque-Based Phage Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Efficiency of Plating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Efficiency of Center of Infection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mixed-Indicator Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

508 509 509 515 516 517 519 522 522 526 528 531 531 532 534 536 536

Abstract

Plaques are spatially constrained populations of bacteriophages that become visible to the eye as they locally deplete numbers of susceptible bacterial hosts. Plaques develop within what are known as “lawns” of bacteria, as grown either on or in solid or semi-solid media, media which typically is agar-based. These plaques, by definition, are initiated from an approximation of a point source, that is, usually from a single phage virion or instead from a phage-infected bacterium, what often collectively can be described as plaque-forming units or PFUs. These point sources then spread spherically to form circular “holes” of S. T. Abedon (*) Department of Microbiology, The Ohio State University, Mansfield, OH, USA e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_16

507

508

S. T. Abedon

reduced turbidity, i.e., less cloudiness in the bacterial lawn. Phage plaques are important for at least four reasons. First, the process of their growth can vary in interesting ways, with differences in outcomes that are dependent on differences in phage and bacterial types along with differences in plaquing conditions. Second, phage plaques are the most readily accessible and common circumstance in which spatially structured phage growth is observed in the laboratory. As such, plaques can serve as first-approximation models for phage population growth within naturally occurring spatially structured bacterial populations. Third, phage plaques are a common means by which phage activity can be macroscopically observed for the sake of phage isolation, phage clonal purification, and phage enumeration. Four, phage plaques may be employed to biologically characterize phages such as in terms of their efficiency of plating or host range. Provided here is an overview of the biology of phage plaques and their formation.

Introduction The concept of phage plaques dates back to d’Hérelle’s original phage publication (d’Hérelle 1917, 2011): “one obtains after incubation a layer of dysentery bacilli with a certain number of holes (‘circles’) of about 1 mm in diameter without culture.” These holes are clearings found within otherwise turbid bacterial lawns, that is, regions in which bacteria are either lacking or are relatively lacking in density compared to the surrounding lawn. Plaques typically are a consequence of a combination of phage propagation, virion diffusion, and lysis of bacterial hosts (Abedon and Yin 2008, 2009; Abedon 2017c). In association with phage propagation and virion diffusion, plaques can also result from partial suppression of bacterial growth, that is, even without outright bacteriolytic or bactericidal effects. This latter mechanism is observed with phages that release progeny phage virions from infected bacteria without causing the cell to lyse, as is seen with filamentous phages such as M13 (Salivar et al. 1964) (chapter ▶ “Phage Infection and Lysis”). In any case, plaques are what one observes when growing bacteriophages in or on bacteriacontaining solid media. Though a seemingly straightforward and certainly ubiquitous component of the phage laboratory experience, in fact phage plaques are not as simple as many students of phages might imagine and certainly are not as straightforward as nonspatially structured phage growth within well-mixed broth cultures. In this chapter I consider the complexity that is the phage plaque along with the utility of plaques for at least first-approximation characterization of certain phage properties, particularly host range, efficiency of plating, and viability. In a separate chapter, the enumeration of plaque-forming units (PFUs) is considered particularly in terms of the basic statistical principles constraining the quality of the data that can be obtained from this laboratory procedure (chapter ▶ “Detection of Bacteriophages: Statistical Aspects of Plaque Assay”). A number of reviews of the biology of phage plaques and modeling of their formation have been published relatively recently, and the

Detection of Bacteriophages: Phage Plaques

509

reader is encouraged to seek these out for additional discussion: Abedon and Yin (2008, 2009), Krone and Abedon (2008), and Abedon (2011a). See also the discussion of why phage propagation within plaques may serve as models for phage propagation in association with bacterial microcolonies as presented towards the end of Abedon and Thomas-Abedon (2010). See Table 1 for a glossary of important plaquing terminology and concepts.

Plaques There exist a number of technical issues associated with plaque formation. At its simplest the process involves mixing phages with bacteria in association with some kind of solid or semi-solid bacterial growth media, which is then followed by incubation and plaque growth. This section mostly considers just the first, pregrowth steps of this process, which nevertheless can be surprisingly complex. Key to understanding this process is that, following plating, the steps of plaque formation involve a repeated (i) free phage adsorption to phage-uninfected bacteria, (ii) phage infection of those bacteria, (iii) some degree of inhibition of bacterial growth usually as associated with phage-induced bacterial lysis, (iv) virion release from phageinfected bacteria, and then (v) virion diffusion (repeat). This sequence thus can be described as a reaction- (i.e., infection-) diffusion process. Plaque initiation is simply a process of establishing the center of what ultimately will grow, through phage infection and virion diffusion, into a visible plaque.

Initiating Plaques When bacterial cultures are used to support plaque formation, they are collectively described as an indicator strain or indicator bacteria. Plaques themselves in most cases are initiated by first mixing phages with indicator bacteria. The phages that are added to indicator bacteria may be either isolated virions (free phages) or previously phage-infected bacteria. Given this variation, a phage or phage infection that may go on to initiate a plaque can be described as an infective center (see Table 1). This designation is because as either phages or as phage infections, infective centers are capable of giving rise to the infection of surrounding indicator bacteria and, as initiators of plaques, infective centers are located at what will be the center of the to-be-formed clearing. More typically in the modern literature one will instead see the phrase plaque-forming unit or PFU. Indicator bacteria support plaque formation. Initially the indicator bacteria are in a phage-uninfected state. If phages have been preadsorbed to different bacteria before mixing with the indicator bacteria, then the resulting infections are considered to have begun prior to the mixing of phages with indicator bacteria, i.e., prior to the actual plating step. In addition, indicator bacteria usually will be supplied in far greater numbers than phages, so in any case the majority of indicator bacteria at the start of plaque formation should still be found in a phage-uninfected state. These

510

S. T. Abedon

Table 1 Glossary of terminology Term Abortive infection

Bacterial “clump”

Clumped dispersion

Colony-forming unit (CFU) Confluent lysis

Double-agar technique Efficiency of center of infection (ECOI)

Efficiency of plating (EOP)

Definition Phage infection of a bacterium that results in inactivation of both the infected bacterium and the infecting phage. A less precise operational definition is a phage infection that results in bacterial death and results in a low efficiency of plating. Bacteria which in some manner have become physically associated, as in “clumped dispersion.” Bacterial “clumps” include cellular arrangements and microcolonies but also bacteria which have aggregated together, i.e., come to be clumped together especially by some active process (rather than forming as clumps as they grow, or instead existing as bacteria which by chance happen to be found in the same location). Bacteria within maturing lawns typically are found as clumps. From ecology, a nonrandom as well as noneven distribution of organisms within an ecosystem. Bacteria that display a clumped dispersion exist either as cellular arrangements, microcolonies, or aggregations. For phages, however, clumped dispersions are concentrations of phages within a given area, e.g., as within a plaque, but do not necessarily represent an actual physical clumping of virions (although some phages can form aggregates). The clumped dispersion of both phages and bacteria are likely important determinants of the biology of plaque formation. Single bacterium, or clump consisting of more than one bacterium, that upon plating forms into a single colony, or into a microcolony when plated at higher densities. Clearing of a bacterial lawn caused by formation of extremely large numbers of closely associated plaques (an extreme form of TNTC). Confluent lysis implies that the resulting plaques have merged. Confluent lysis can be mimicked, that is, lawn clearance without prior plaque formation, by the occurrence of an inhibition of bacterial replication that results instead in what should be described as zones of inhibition. This could occur, for example, given application of so many phages to a bacterial lawn, i.e., as in the course of spot testing, that a majority of bacteria are reached and killed early on and thus are unable to support plaque formation. Equivalent to soft-agar overlay. Likelihood of plaque formation given preadsorption prior to plaquing, and generally with use of indicator bacteria which are highly permissive to plaque growth. Failure to form a plaque during ECOI determinations generally is equivalent to platedphage inviability, though less commonly also can be a consequence of excessively long phage latent periods during the first cycle of infection. Likelihood of plaque formation either absolutely (i.e., as a function of number of virions plated) or instead relatively, as in comparison with plating under different conditions or using different indicator bacteria. EOP determination requires fewer steps than ECOI determination, but EOP determinations also supply less simply interpreted information. In particular, failure to form a plaque during EOP determinations is not necessarily equivalent to phage inviability. (continued)

Detection of Bacteriophages: Phage Plaques

511

Table 1 (continued) Term Indicator (bacteria or strain) Infective center

Isolated plaque

Halo

Lawn

Lysis

Microcolony

Definition Bacteria used to initiate a bacterial lawn that is used to support plaquing. Contrast with bacteria employed for phage preadsorption prior to plaquing. Individual virion or, instead, individual virus-infected bacterium which upon plating can form a plaque. An infective center serves as the physical center of what ultimately will form into a plaque, and ideally an infective center represents just a single virion or instead a cell that has been infected by just a single virion. This is versus, e.g., a single cell infected by more than one virus, a clump of virions, or more than one attached cell collectively infected by more than one virion. A plaque that visually is not touching another plaque. A wellisolated plaque has a distance of at least multiple millimeters between itself and other plaques. Beware, however, that phages can spread further from the center of a plaque than the visually observed clearing. Isolated plaques are desirable during enumeration, contrasting overlapping plaques which are difficult to distinguish. Well-isolated plaques are especially desirable towards phage isolation into pure culture. Region surrounding a plaque’s clearing that is less turbid than a bacterial lawn, but more turbid than the plaque proper. Halos are caused by hydrolytic enzymes, i.e., extracellular polymeric substance depolymerases (Pires et al. 2016). As such, halos can continue to grow even during storage of plates in a refrigerator since the halos are caused by simple enzymatic action, versus the much more complex phage infections. These regions also are larger than the plaque itself because enzymes, as smaller particles, tend to be able to diffuse faster through semi-solid media than can phage virions. Dense, macroscopically somewhat homogeneous, two-dimensional bacterial culture, e.g., as made by spreading a high concentration bacterial culture on the surface of agar in a Petri dish. Bacterial lawns at more microscopic levels instead can be somewhat heterogeneous, often consisting of discrete microcolonies. Destruction of individual cells such that their cytoplasmic contents leak out through their now-damaged cell envelopes. The turbidity associated with an individual bacterial cell generally drops substantially following its lysis, and lysis is the means by which lytic bacteriophages both terminate their infections and release otherwise intracellular phage progeny to the extracellular environment. It is important to recognize that though lysed bacteria are dead, phage infections of bacteria can result in bacterial death without, in all cases, those bacteria necessarily lysing (particularly, see abortive infections). Microcolonies, as distinct from cellular arrangements, consist of cells which have failed to spatially separate due to inhibitions on daughter-cell movement following bacterial replication. Movement-inhibiting substances can include agar in solid or semisolid growth media or instead extracellular polymeric substances in bacterial biofilms. Microcolonies differ from colonies in that microcolonies are much smaller, i.e., they are very small colonies of generally clonally related cells. (continued)

512

S. T. Abedon

Table 1 (continued) Term Permissive conditions Permissive host or strain Plate lysate

Plating Plaque

Plaque harvesting Plaque purification

Plaque-forming unit (PFU) Plaquing

Pour plate

Pre-adsorption during plaquing Pre-adsorption prior to plaquing

Definition Chemical and physical parameters within which plaquing of a given phage strain can occur with reasonable efficiency. Bacterium upon which phage plaquing of a given phage strain can occur with reasonably efficiency. Growth of phage stocks as multiple, confluent plaques rather than using well-mixed broth. Plate lysates are a variation on the softagar overlay technique and, postincubation, resulting phages must be separated from the agar. Placement of microorganisms on or in solid or semi-solid media in a plate/Petri dish. Generally circular area of clearing of a bacterial lawn that is both initiated and caused by, ideally, a single phage particle or phageinfected bacterium. Plaques tend to be defined as entities which are visual to the naked eye. Removal of a plaque from a lawn so as to suspend its virion contents into fluid such a broth or buffer. Serial plating and harvesting of well-isolated, single plaques, generally through at least three rounds, as a means of obtaining a virus pure culture. Entity that when mixed with indicator bacteria and plated forms a single plaque. Ideally equivalent to an infective center. Process of generating one or more plaques, i.e., typically beginning with mixing of phages with indicator bacteria, whether in the course of preadsorption or instead during application to solid or semi-solid media, and which is followed by a period of incubation which continues typically through the point of the indicator bacteria entering into stationary phase. Means of generating solid or semi-solid environments for bacterial lawn formation within Petri dishes, involving mixing of microorganisms with molten forms of this media prior to their being poured into the Petri dish. Mixing of virions with indicator bacteria present at relatively high densities, which is then followed by some degree of incubation (e.g., 1 min or more) prior to plating. Mixing of virions with relatively high densities of bacteria so as to assure adsorption prior to the mixing of phages with indicator bacteria. In this way, the initial phage infection will be to a bacterial strain which can be different from that used as indicator bacteria. This preadsorption is indicated as occurring prior to plaquing because of the requirement for subsequent mixing of infective centers with indicator bacteria in order to initiate plaque formation. To avoid artifacts, virions should be removed following this preadsorption step so that they are not present at the point of mixture with indicator bacteria. See section “Efficiency of Center of Infection.” (continued)

Detection of Bacteriophages: Phage Plaques

513

Table 1 (continued) Term Reaction-diffusion

Semi-solid media

Soft-agar overlay

Solid media

Spreading (spread plate)

Spatial structure

Spot

Definition Borrowed from physical descriptions of phenomena in which particle movement occurs due to a repeated combinations of generating new particles (“reaction”) and subsequent movement of those particles (“diffusion”). Plaque formation is a reactiondiffusion process in which phage infections of individual bacteria provide the reaction, whereas thus-produced virions then diffuse. Reaction-diffusion processes are observed within environments in which substantial environmental mixing or fluid flow does not occur, i.e., environments possessing substantial spatial structure and in which particle movement therefore is dominated by diffusion. Solid media containing smaller amounts of hardening agent (e.g., agar) such that diffusion especially of larger particles (e.g., phage virions) can more easily occur. Two-step methods towards generating a pour plate where a layer of sterile solid media is prepoured and allowed to solidify. A second layer of semi-solid media, containing the to-be-plated microorganisms, is then poured over the first layer and allowed to solidify. Generally growth media that contains a gel-like hardener, usually agar, which prevents the media from flowing or mixing, with movement with solid media therefore tending to be limited to diffusion. Means of applying microorganisms evenly to the surface of solid media as found within Petri dishes. Spreading is one means of generating bacterial lawns. Prior to incubation, a spread plate typically will not possess perceptible turbidity and the lawn is seen to form in the course of incubation. Property of environments in which movement of constituents is inhibited, though not necessarily fully inhibited. Generally such environments lack substantial gross mixing, i.e., gross mixing as instead can be seen with stirred or shaken broth, and will lack in fluid flow generally, at least as associated with the spatial structure. With microorganisms, diffusion instead dominates as a means of movement. Solid and semi-solid media impose spatial structure on their resulting growth environments. The formation of plaques is a consequence of virus population growth within spatially structured environments. Zone of inhibition resulting from the application of substances to an immature bacterial lawn, via spotting, that are capable of inhibiting the replication of bacteria including inhibiting the replication of bacteria which have not yet formed into microcolonies. Spots are visualized as large circles of confluent clearing of bacterial lawns, though this clearing technically is not necessarily equivalent to confluent lysis. (continued)

514

S. T. Abedon

Table 1 (continued) Term Spotting

Too few to count (TFTC)

Too numerous to count (TNTC)

Viability

Zone of inhibition

Zone of lysis

Definition Application of fluids to the surface of what typically are immature bacterial lawns especially so as to supply bacteria-replication inhibiting substances to bacteria, such as the application of bacteriophages. Depending on concentrations of applied phages, spotting can give rise to spots (given the application of more phages) or instead to one or more individual plaques (application of fewer phages). Insufficient numbers of plaques per Petri dish to allow for statistically precise enumeration. For example, less than 50 plaques per plate can define TFTC. Typically, given consistently TFTC plaque numbers for a given plated dilution, then one will employ a lower dilution so as to obtain more plaques per plate for enumeration. Sufficient numbers of plaques per Petri dish such that plaque counts substantially decline in number due to plaque overlap. For example, TNTC might be set to over 500 plaques per plate, or a smaller number for very large plaques, or a greater number for smaller and highly distinct plaques. Generally TNTC is agreed upon for a given set of phages, bacteria, and conditions rather than determined on a plate-by-plate basis. In microbiology, viability generally is equated with an ability to produce offspring. A lack of plaque formation, however, is not sufficient to indicate a lack of phage viability (i.e., phage inviability) since plaques can fail to form as a consequence of either low infection fecundity (low numbers of virions produced per infection, i.e., low burst size), due to excessively long infections (extended latent periods), or due to excessively slow rates of virion diffusion. Compare ECOI with EOP with regard to phage viability determination. From antibiotic testing, these are regions in or on solid or semisolid media in which bacteria are visibly absent due to inhibition of bacterial growth. The zones of lawn clearing seen with spot formation are zones of inhibition. With spot testing, this inhibition thus can be due to phage infection, but also can potentially be due to bacteriocin activity. Description, often inadequate or false, of the clearing observed following spotting and subsequent lawn incubation. The idea of a zone of inhibition is broader and thereby generally more likely to be a correct description of zones of lawn clearing as associated with spots.

phage-bacteria combinations then need to be incubated under sufficiently permissive growth conditions such that the indicator bacteria will grow into a bacterial lawn, ideally a somewhat turbid lawn. For subsequent plaque-based enumeration, PFUs will be counted only when plaque numbers are neither Too Few To Count (TFTC) nor Too Numerous To Count (TNTC) (chapter ▶ “Detection of Bacteriophages: Statistical Aspects of Plaque Assay”). The goal in any case, whether for phage

Detection of Bacteriophages: Phage Plaques

515

enumeration or phage isolation, is to consider only what can be described as “isolated plaques,” that is, plaques which are visually separate from other plaques. Alternatively, though not considered here, is the growth of phage stocks on plates, so-called plate lysates, in which plaques are intentionally not isolated from one another but instead are plated in sufficient numbers to give rise to a confluent lysis.

Bacterial Lawns, Plaques, and Spots A bacterial lawn is a thin layer of homogeneous bacterial culture. Lawns are created when these cultures are applied to and then grown either within semi-solid media or instead on the surface of solid media. Lawns can be made by either spreading bacteria over the surface of solid media or instead by mixing the cells with molten solid or semi-solid media and then pouring that agar over the surface of solid media as a thin layer – usually of what is variously described as semi-solid, soft, top, or “sloppy” agar – and onto a thicker layer of fully-gelled solid media (hard or bottom agar). The procedure is known as a soft-agar overlay (Sanders 2012), or the double agar overlay plaque assay (Kropinski et al. 2009). Alternatively, in some circumstances the lower layer is skipped and the semi-solid media may be poured directly into the plate (Rizvi and Mora 1963; Yin 1991; Mazzocco et al. 2009a), though this latter approach requires thicker agar layers which can interfere with the observation and therefore with the enumeration especially of smaller plaques. To initiate plaque formation, generally phages are mixed with lawn-forming bacteria either prior to their application to a plate or, in the case of spread plates, this can be during their application to the plate. Plaques not only can be initiated by mixing PFUs with indicator bacteria, they also can be observed in association with spotting (e.g., spot or drop titering), which involves phages suspended in a liquid that is applied to the bacterial lawn after lawn initiation (typically within minutes of lawn initiation). Formation of isolated plaques in the course of spotting, however, will typically be seen only given sufficiently small numbers of phages present per applied drop of phage-containing media or buffer (Carlson and Miller 1994; Mazzocco et al. 2009b). It is important to note, however, that the plaque-associated lysis observed in this manner is not equivalent to spots as considered elsewhere in this chapter, which instead can represent the confluent merger of multiple plaques, that is, where the density of phages applied per drop is too high to result in the formation of isolated plaques. Alternatively, and potentially more commonly, spots can form in the course of killing of most bacteria prior to plaque formation, i.e., such that a majority of bacteria become adsorbed by added phages rather than following subsequent lysis of the thus-produced phage-infected bacteria, what some authors refer to, for the most part incorrectly, as a lysis from without (Abedon 2011b). More generally, confluent spots are simply another name for what in the antibiotic literature is described as a zone of inhibition. See Fig. 1 for summary of plaque versus spot formation. Spots are considered in further detail below.

516

S. T. Abedon

Spread Plate Technique

Plaques Forming Within Lawn

note bacteria as rods

note circular clearings

Overlay Technique

Initiation of Spots Upon Lawn

bacteria and overlay are mixed prior to pouring

infective centers (phages) are applied after initiation of lawn

Pour Plate Technique

Plaques Form as Spheres

only very large plaques will form clearings that completely span the lawn

the clearing to the left has failed to completely span the lawn

Fig. 1 Variations on the initiation of lawns as well as illustration of the difference between plaques and spots. To form a lawn, a bacterial culture must be associated as a confluent layer with solid media. This can be on top of that media (spread plate), embedded within a thin layer on top of a thicker layer (overlay), or found completely through the media (pour plate). Plaques form within the lawns (shown, upper right, within an overlay as white circles). Spots are initiated on top of immature lawns and typically will involve the application of far higher phage densities than is the case for the growth of isolated plaques (center, right; note that with actual spot formation the applied droplets are much larger relative to lawn bacteria than as illustrated). For a plaque to become visible, it must be large enough in diameter to be seen horizontally but also, ideally, will be large enough in height to span the lawn vertically. In any case, the plaque itself forms as a three-dimensional sphere that is constrained in its growth by an absence of bacteria found both above and below the lawn. The apparent two-dimensionality of a plaque is a consequence of these vertical constraints rather than representing an inherent property of phage population growth within spatially structured environments (bottom, right). In any case, what we call a plaque is a localized reduction of bacterial lawnassociated turbidity rather than a direct visualization of the presence of phages

Plaque-Forming Units As noted, we can describe both free phages and phage-infected bacteria as plaqueforming units (PFUs). The simple as well as desirable case, under most circumstances, is when a plaque is formed from a single phage. Thus, a single virion, a single cell infected by a single virion, or even a “clump” of cells (e.g., a cellular arrangement) infected by a single virion all represent essentially a single PFU. Alternatively, there are a number of situations in which PFUs consist of more than one phage, and this can be a concern. Indeed, PFUs generated by more than one phage can be problematic in situations other than plaque formation, such as when claims are made of larger phage burst sizes with larger multiplicities of infection, but when infected bacteria exist as “clumps” (e.g., staphylococci) rather than as isolated

Detection of Bacteriophages: Phage Plaques

517

bacteria, in which case it is not so much higher multiplicities of infection which are being observed but instead multiple bacteria that are being infected per colonyforming unit. In any case, examples of where more than one phage may be counted as a PFU can include the following: physically associated (tangled-together) virions, an isolated bacterium that has been infected by more than one phage, or instead more than one phage-infected bacterium that happens to be physically associated at the point of plaque initiation (e.g., multiple phage-infected bacteria that have become aggregated together or instead which make up a single cellular arrangement). Rather than simply artifactual circumstances, it is important instead to keep in mind that the concept of PFU and the concept of individual phage are not always identical, just as the concept of CFU (colony-forming unit) and that of individual bacterial cell is not always equivalent (e.g., plating nondisassociated staphylococci). In addition, a single plaque can form from two or more free phages or two or more phage-infected bacteria that by chance happen to have been plated fairly close to one another (this too is distinct from the clearing seen with a spot, which instead represents a larger area of lawn clearing than that associated with a single plaque). Regardless, only entities that are physically cohesive prior to the point of their plating can legitimately be described as a single PFU, if not necessarily also an individual phage. On the other hand, if two PFUs ultimately appear to have formed into only a single plaque due to chance proximity but not actual, physical attachment prior to plating, then they are still, technically, two PFUs, even if they are observed to be a single plaque. In other words, even taking into account only their initiation, and ignoring spots, there exists a complex diversity of phenomena from which one can, starting with phage and indicator bacteria, generate a plaque. See Fig. 2 for summary.

Too Many or Too Few Plaques The number of phages plated per Petri dish is important for a variety reasons. These include a need to have sufficient plaque numbers for enumeration, not so many plaques that it is difficult to distinguish one plaque from another, and in addition a sufficiently turbid bacterial lawn needs to be able to form to allow the visualization of the plaque. The issue of too few phages is referred to a Too Few To Count (TFTC) and is a statistical issue of importance during plaque-based enumeration (chapter ▶ “Detection of Bacteriophages: Statistical Aspects of Plaque Assay”). This is also a matter of efficiency when plaquing for phage isolation since the more phages which form per plate then the more potential phage isolates that will be present. Numbers of plaques found on a plate can be controlled via a combination of dilution prior to plating and manipulation of the phage-containing volumes that are plated. To minimize the initiation of plaques by more than one phage, or to avoid having plaques form which are so close together that they are indistinguishable, typically one seeks to limit the number of PFUs plated. This practice is important in terms of plaque-based enumeration since the addition of too many phages will result in a plate with plaques that are Too Numerous to Count (TNTC) (chapter ▶ “Detection of

518

S. T. Abedon

C

B

D

E

A

F

Fig. 2 Six examples of infective centers/plaque-forming units. It is important to keep in mind that none of these are equivalent to plaques but instead represent the first stages as well as the centers of what, following incubation, will become plaques. (A) a phage-infected bacterium, (B) a free phage, (C) a bacterium that has adsorbed more than one phage, (D) a cellular arrangement containing a single bacterium that has been infected by a single phage, (E) a larger cellular arrangement also containing a single bacterium that has been infected by a single phage, and (F) a cellular arrangement containing multiple bacteria with more than one of these bacteria infected by separate phages. Additional variations are possible including phage-infected bacteria that have become clumped together (aggregated) rather than existing strictly as cellular arrangements, or virions that have become clumped together. In all cases, note that the infective center exists as a single, physically cohesive, phage-containing entity prior to its plating and that a single infected center/ PFU, regardless of how many virions or phage-infected bacteria they contain at the point of plating, will give rise to only a single plaque

Bacteriophages: Statistical Aspects of Plaque Assay”). This practice is important in terms of plaque-based phage isolation as well because it is best to avoid having plaques which are overlapping at the point of phage isolation. Indeed, during phage isolation plaques are often subject to multiple rounds of “purification,” which is to say harvesting a plaque and then re-plaquing associated phages so that there is high certainty that a given plaque has been initiated by only a single phage clone (i.e., pure phage isolate). Keeping plaque numbers per plate relatively low is useful for plaque purification (e.g., no more than 200 per plate). Keeping numbers of plaques per plate low is relevant as well during the study of plaque sizes or morphologies, since the goal in all of these cases are well-separated plaques. Confluent lysis, that is, clearing of all or most of a bacterial lawn, will occur if far too many PFUs are plated and thus represents an extreme form of TNTC. Alternatively, lawn clearing can occur when bactericidal phages are plated which nevertheless cannot plaque with high efficiency, that is, which display a low efficiency of plating – phages that can kill bacteria but do not reproduce, in other words, can still clear bacterial lawns if enough of these phages are supplied. In this case,

Detection of Bacteriophages: Phage Plaques

519

even if PFUs do not exceed TNTC, one can still observe what at least by appearances appears to be a confluent lysis. Here the problem therefore is not too high PFUs but instead too high ratios of bactericidal virions to indicator bacteria; for discussion of ratios of phages to bacteria, often described, not necessarily accurately, as multiplicities of infection (MOI), see Abedon (2016b). Note that bactericidal virions that cannot form plaques can still be enumerated via a technique described as killing titer determination (Carlson 2005). In any case, if numbers of bactericidal but not plaquing virions are not too excessive, then this problem can be addressed by a combination of plating fewer PFUs (and thereby fewer virions overall) as well as by supplying greater numbers of indicator bacteria. A related issue occurs in the course of what is known as multiplicity reactivation (Hyman 1993) in which DNA-damaged virions cannot form plaques upon singly infecting bacteria but can form plaques upon multiply infecting bacteria. In this case excessive ratios of virions to bacteria can result in more plaques than anticipated, or even confluent lysis. Note that to appreciate likelihoods of multiple infection of individual bacteria by multiple virions, it is crucial to understand Poisson distributions (Dulbecco 1949; Stent 1963; Carlson 2005; Abedon 2011a, 2016b; Abedon and Katsaounis 2017) (chapter ▶ “Detection of Bacteriophages: Statistical Aspects of Plaque Assay”).

Plaques Versus Spots Lawns are initiated with fewer bacteria than will be present at the point of cessation of lawn growth. Growth cessation occurs at the point of entrance of lawn bacteria into stationary phase. As such, a lawn in terms of its formation has both a temporal beginning and, at least in terms of growth, a temporal end. Phages typically are applied to lawns before, at, or relatively soon after the initiation of lawn growth. There are two general means by which phages can be applied to lawns. These are either by mixing with bacteria prior to the point of lawn initiation or by application to lawns after lawn initiation. Whether poured or spread, the mixing of bacteria and phages prior to lawn initiation is typical of plaque initiation. This process usually results in relatively isolated zones of bacterial clearance centered on plated PFUs, with plaques forming over the course phage population growth, lawn maturation, and localized lawn-growth delay or instead lawn lysis stemming from phage action. It is also possible to generate phage plaques by applying phages to already initiated lawns (Carlson and Miller 1994; Mazzocco et al. 2009b). This is done with the application of small volumes of liquid on top of already applied bacteria and has the utility of allowing for multiple, individual phage enumeration processes (spots/drops) per individual petri dish. Alternatively, one can apply so many phages in this manner that confluent lysis centered on the applied drop occurs. In this case, the collective clearing is not described as a plaque but instead as a spot. Applying even more phages can result in killing of a majority of bacteria prior even to the initiation of plaque formation by applied phages, which also results in the formation of a spot.

520

S. T. Abedon

More precisely, a plaque is a localized region of reduced lawn turbidity that has been initiated by approximately a single PFU. A spot, by contrast, is a localized region of reduced lawn turbidity that has been initiated by multiple PFUs or, more generally, by multiple bacterial-growth inhibiting entities. Drop enough phages onto a growing, phage-susceptible bacterial lawn, and you will generate a spot. Drop enough bacteriocin or phage-produced endolysin (chapter ▶ “Enzybiotics: Endolysins and Bacteriocins”), however, and you similarly can generate a clearing. Indeed, drop enough antibiotic onto a growing, sensitive lawn, and this also will succeed in generating a clearing. Normally for most combinations of phages, indicator bacteria, and conditions, a spot – no matter its cause – will display a larger diameter than an individual plaque, usually of roughly the same size and shape as the liquid aliquot spotted onto the lawn. An important distinction between plaques and spots is that while plaques can form only in conjunction with phage population growth, a spot can form either with or without such growth. Specifically, spots require only the killing or otherwise growth inhibition of bacteria, i.e., as equivalent to a zone of inhibition as mediated by an antibiotic. In spot formation it is important as well to keep in mind that at the point of spot initiation the concentration of lawn bacteria is typically somewhat lower than the concentration of those bacteria that would be present at the point of cessation of lawn growth. This means that numbers of lawn-applied phages typically must be fairly high if localized bacterial clearance is to be achieved in the absence of associated phage replication, but not as high as would be required were these nonreplicating phages applied at the end of lawn formation versus at its beginning. Complicating things further, spots also can spread beyond their point of application. Such additional clearing may be associated with phage population growth, as though a spot were a giant plaque, though one that has been initiated by multiple rather than individual PFUs. Given the potential for spots to produce clearings by killing off bacterial populations relatively early in lawn growth, then in principle at least some clearing outside of the area of spot initiation could occur due to virion diffusion and subsequent bacterial adsorption even without bacterial infections producing phage progeny. This additional clearing can be significant, however, only if sufficiently large numbers of bacteria-growth inhibiting substances, e.g., phage virions, have been supplied to initiate spots, and clearing will be substantial to the point of being noticeable only if the applied virions, or other substances, are capable of diffusing somewhat beyond their initial point of application, as may be readily achieved, of course, if those phages are capable of productively infecting lawn bacteria (i.e., as seen when spotting using phages that also are capable of forming plaques). Thus, (1) spot formation can occur as a consequence of more than one mechanism. (2) The process of spot formation can be relatively complex. (3) A spot is not a plaque, though in the course of spotting it is possible to observe discrete plaques if sufficiently low numbers of PFUs are provided to initiate a spot (Carlson and Miller 1994; Mazzocco et al. 2009b). (4) Entities other than phages that can be present in phage lysates can also produce spots (e.g., bacteriocins). (5) The potential for spot formation can vary with plating conditions (e.g., temperature, media composition).

Detection of Bacteriophages: Phage Plaques

521

No Plaque

Fig. 3 How to interpret presence or absence of spots in combination with presence or absence of plaques. Generally plaque formation should require greater phage “performance” than spot formation. Thus, an absence of a spot is more meaningful than the absence of a plaque, whereas the presence of a plaque is more meaningful than the presence of a spot when determining phage host range. ‘No Spot’ should be interpreted as ‘No Spot and No Plaque’

Plaque Formation

(6) Spots are produced via the inhibition of bacterial growth, often by phages simply infecting and thereby killing bacteria. Consequently, with regard to this last point, there is no logical justification for assuming that spot formation is a consequence of lysis from without, a peculiar as well as potentially rare form of phage induced lysis that is associated with high multiplicities of phage adsorption (Abedon 2011b). (7) Spots can form even when employing phages that otherwise (Mazzocco et al. 2009b) cannot plaque with relatively high efficiencies of plating, just so long as those phages nevertheless can kill bacteria with high enough efficiency following adsorption. In short, anything that can kill or inhibit the growth of bacteria can form a spot on a bacterial lawn. Spot formation by a lytic-phage lysate therefore can at best be viewed as probably a consequence of bactericidal phage adsorption. This relative lack of precision as to the mechanism of spot formation can be relevant especially to issues of phage host range determination (Hyman and Abedon 2010; Mirzaei and Nilsson 2015), so spots always should be viewed at best as facile first-steps towards the characterization of phage interactions with specific host bacteria, and never as definitive descriptions of such interactions, that is, never as any more than a draft description of a phage’s productive host range as particularly there is a great likelihood of false positives. Alternatively, in terms of phage host range, absence of a spot or plaques may be viewed as a good indication that a bacterium is outside of a phage’s productive host range, at least under the conditions being tested, and if enough virions have been used then outside of a phage’s bactericidal host range as well, while presence of a plaque serves as a good indication that a bacterium is found within a phage’s productive host range (see Fig. 3). The issues of spots and their relevance to determinations of phage host range were recently explicitly tested by Mirzaei and Nilsson (2015), who compared phage host range determinations using spot testing versus efficiency of plating (EOP). In their words, from their abstract: “The analyses of the differences between the two methods show that spot tests often overestimate both the overall virulence and the host range Spot Formation

No Spot

Interpretation

Interpretation

Bacterial host used as indicator is found within a phage’s productive host range

Error: Lack of spot formation should not occur given plaque formation

Interpretation

Interpretation

Possibility of spotformation false positive; best to explore further in broth culture

Bacterial host used as indicator likely is not found within a phage’s bactericidal host range

522

S. T. Abedon

and that the results are not correlated to the results of EOP assays.” Thus, spot testing at best should be viewed as a preliminary step in phage host range determination, though spot formation failure given initiation using otherwise large numbers of phage virions suggests low bactericidal activity of a specific phage against a strain of lawn bacteria. Spot formation itself, however, conversely cannot be conclusively equated with phage bactericidal activity against a specific bacterial host strain or potential for propagation on that bacterium. Furthermore, spots should never be described as plaques and vice versa. Lastly, spots as zones of inhibition of bacterial growth are not necessarily zones of bacterial lysis, and except in the case of application of phage endolysins (chapter ▶ “Enzybiotics: Endolysins and Bacteriocins”), should not be considered without evidence to have formed as a consequence of lysis from without (Abedon 2011b).

Plaque Formation The concept of a phage plaque is straightforward: They are a localized, at least relative absence of lawn bacteria that can form during culturing of phages either on or in solid or semi-solid media. Under greater “magnification,” however, plaquing is far more complicated. Generally the plaque-formation process begins with agarassociated random dispersions of infective centers along with separate random dispersions of colony-forming units (CFUs), as both can be considered to be randomly distributed spatially at their point of plating. With growth of both the bacterial lawn and associated plaques, however, both phages and bacteria instead will typically take on clumped dispersions. This clumping for the bacteria is represented by their formation into microcolonies. Clumped dispersions are highly relevant to the biology of plaque formation, as consider below, though as an aside, note that a “clumped dispersion” does not mean that so-dispersed entities have necessarily “clumped” together, but rather clumping is simply a tendency for entities to be found closer together than would have been expected by random. In biology the most common means by which a clumped dispersion is generated is by offspring not traveling too far from each other or their parents following their formation, i.e., “the apple doesn’t fall far from the tree,” which is the mechanism considered here.

Phage Clumped Dispersion The clumping of phages into individual plaques – that is, the failure of newly produced phage virions during plaque formation to substantially separate from one another – has at least four consequences: plaque visibility, plaque variation across their diameters, secondary adsorption/infection, and also a potential for researchers to gain information from the appearance of plaques. The first of these issues is that plaques can become macroscopic, which is crucial to their use as a means of phage enumeration (chapter ▶ “Detection of Bacteriophages: Statistical Aspects of Plaque Assay”), isolation (Millard 2009) and for phage characterization based on plaque

Detection of Bacteriophages: Phage Plaques

523

appearance (Abedon and Yin 2008). Plaques, that is, represent localized aggregations of bacterial lysis. The second issue is that phage dynamics and therefore plaque dynamics will vary as a function of relative position within a plaque. Specifically, plaques are three-dimensional entities that display variations in properties as a function of location within a plaque, location relative especially to a plaque’s periphery. For simplicity, note that plaques are typically conceptualized as two dimensional (and even as one dimensional), and this abstraction can be acceptable so long as one realizes that these are simplifications rather than strict representations of plaque geometry. Plaques in particular can be differentiated into at least three zones. Going from outside inward (see Fig. 4), these are: (3) locations within a plaque possessing a combination of reduced bacterial densities, phage virions, phage-infected bacteria, and phage-uninfected bacteria; (2) locations within a plaque where intact but phage infected bacteria remain but not uninfected phage-sensitive bacteria; and (1) locations within a plaque where intact bacteria – except for phage-resistant bacteria – no longer remain. There also is potentially a fourth and particularly poorly appreciated zone that is found outside of the visible periphery of a plaque but which likely consists of at least some phage-infected, or at least phage-adsorbed bacteria, though this region is not yet visibly cleared via phage action, especially when viewed at macroscopic scales. The latter may or may not be included as a plaque’s leading edge since technically, as noted, it is not part of a plaque’s clearing. Thus, going from outside inward, a typical plaque will be surrounded by (4) a region where bacteria are still physically intact but at least some may have become phage infected, (3) a region where new phage infections are being initiated and older phage infections are ending; (2) a region where new phage infections are no longer being initiated but nonetheless phage-infected bacteria continue to persist, and (1) a region in which neither phage infected nor phage sensitive bacteria persist. As plaques become larger, the portion consisting of region 1 in particular will become larger while regions 2, 3, and 4 can consist of a wave of phage diffusion, bacterial infection, and bacterial lysis. In terms of plaque turbidity, we expect – at least as the default situation and particularly ignoring phage-resistant bacteria as well as plaques formed by temperate phages – that there will be greater turbidity near to a plaque’s periphery with decreasing turbidity (i.e., cloudiness) in the course of heading visually towards a plaque’s center. To the extent that more peripheral cloudiness consists of phage-infected or even phage-uninfected bacteria (Abedon and Yin 2008), then this turbidity should be viewed as coinciding with a persistence of region 3, or even of region 2 in the case of delays in the lysis of individual bacterial infections. Generally, however, regions of persisting cloudiness will not be present in region 1 unless, as noted, the turbidity is associated with phage-resistant bacterial mutants or lysogens. See Krone and Abedon (2008) for further consideration of how individual plaques may be parsed into regions of virion movement and bacterial lysis going from the center of a plaque outward to the bacterial lawn. There these regions are described as (1) a zone of clearing, (2 and 3) a zone of reduced turbidity, and (4) a zone of infection. In addition, Krone and Abedon consider a plaque’s periphery, there a plaque’s “leading edge,” as a narrow band surrounding

524

S. T. Abedon

5. Lawn

5 4

3

5. Lawn

2

3. Intermingling of infected and uninfected bacteria

1

2. Infected, phagesensitive bacteria only (distinct region exists only if infections are relatively long lasting, e.g., lysis inhibition)

Plaque

5. Lawn

4. Leading edge

5. Lawn

1. Phage-resistant bacteria only

Fig. 4 Regions of a forming plaque. The plaque is initiated at what will become its center (point “0,” the presence of which can be assumed, though is not shown). Prior to initiation at point “0” there is found an infective center. Following a plaque’s initiation, one or more lawn bacteria become phage infected. Phage virions released from these infections in part diffuse outward, colliding with and then infecting other bacteria. This generates what can be described as a wave, with uninfected bacterial lawn (5) found at the front of that wave. An absence of phage-sensitive bacteria (1) is found at the back of the wave, and this absence of phage-sensitive bacteria spreads over time away from a plaque’s center. Within the wave, increasing numbers of infected and then lysed bacteria are found (4–2). The wavefront, or leading edge (4), can be viewed as the point of phage virion entrance into the uninfected bacterial lawn. Note that these regions not only will change in position and dimensions over time, but in most cases also will change in terms of bacterial densities since the bacterial lawn supporting the outward migration of a plaque’s periphery will be growing in terms of bacterial numbers over the course of plaque formation. Not indicated would be the formation of plaque halos, which can occur due to the diffusion of hydrolytic enzymes past a plaque’s leading edge, where reduced cloudiness may be observed if these enzymes – generally extracellular polymeric substance depolymerases of phage origin – can, for example, degrade bacterial glycocalyx (e.g., extracellular capsule material). Halos typically have the property of continuing to grow even following plate/plaque refrigeration as they are not a direct consequence of phage infection but instead a consequence of partial chemical decomposition of uninfected bacteria. Furthermore, the halos are observed outside of the plaque proper due to a combination of extracellular polymeric substance clearance being less obvious within the plaque and faster inherent rates of diffusion by enzymes than by virions due to the former’s smaller size (Sutherland et al. 2004)

the regions containing not-yet phage-infected bacteria within which free phages may have invaded but phage adsorption to bacteria has not yet occurred. These regions in both of these models in turn are surrounded by otherwise undisturbed bacterial lawn. See Fig. 4 for summary.

Detection of Bacteriophages: Phage Plaques

525

The third issue stemming from phages displaying clumped dispersions in the course of plaque formation is that of phage secondary adsorption. Secondary adsorption is defined from the perspective of already phage-infected bacteria, where the primary adsorption is mediated by the first virion which infects a bacterium, while secondary adsorption refers to additional phages which then adsorb this already phage-infected bacterium (Abedon 2015). Such secondary adsorption is presumably highly likely during plaque formation because virion densities are expected to be higher within plaques than they would be were these virions instead randomly distributed across environments, the latter, i.e., is as seen during phage population growth within well-mixed broth cultures. These secondary adsorptions have the effect, for example, of inducing lysis inhibition during plaque formation by T-even phages (Abedon 1994, 2011a, 2012) and also of inducing lysogenic infections for temperate phages (Abedon 2017d). The latter can be observed particularly in the center of plaques, presumably because phage-resistant bacteria there – resistant due to superinfection immunity displayed by newly formed bacterial lysogens (chapter ▶ “Temperate Phages, Prophages, and Lysogeny”) – are able to grow with less crowding by other lawn bacteria for longer periods prior to our observation of them, and hence these colonies or microcolonies can grow larger and become more visible. Note that some authors describe the resulting turbid-centered plaque as having a “bullseye” morphology, though other authors use this term instead to describe plaques which are clearer towards their center (zone 1, Fig. 4) as representing a “bullseye.” The fourth consideration is that different phage types often can be distinguished in terms of their plaque morphologies, whether in terms of plaque size (below) or instead due to more subtle characteristics (Abedon and Yin 2008). A related issue is that during experimental evolution studies it is possible to compete phages within plaques (Yin 1993, 1994). Note that “competition” is an ecological or evolutionary concept in which different genotypes “compete” either indirectly for resources (exploitative competition) or instead directly in the course of antagonistic interactions (e.g., one individual directly attempts to kill or injure another). Thus, “to compete phages within plaques” is a description of the interaction of different phage genotypes within a single plaque, in this case different as resulting from phage mutation, that have come to display differential reproductive success. Indeed, given phage mutation to more within-plaque competitive genotypes, plaques can display regions in their peripheries which bulge outward, e.g., as giving rise to a “sectored” plaque morphology (Abedon 1994). In addition to within-plaque competition, it is also possible for phage populations that make up individual plaques to compete between those plaques, with withinplaque and between-plaque phage fitness not necessarily resulting from identical properties (Abedon and Culler 2007b). Specifically, we have an expectation that shorter phage latent periods will result in faster rates of growth in plaque size – as more generally will anything that increases rates of plaque-diameter growth (as discussed further below) – thereby being a desirable characteristic given within-plaque competition. For between-plaque competition, on the other hand, larger burst sizes can be particularly desirable, and this is especially so if more

526

S. T. Abedon

phages can be produced per infected bacterium without negatively impacting other phage properties such as lengthening the phage latent period. More generally, anything which increases the numbers of phages produced per individual plaque should result in greater between-plaque competitiveness.

Bacterial Clumped Dispersion The fact that bacteria within bacterial lawns display clumped distributions is conceptually complicating to the mechanics of plaque formation. Unless bacteria are especially motile during lawn development, then individual bacteria found at the point of lawn inoculation will be expected to form, in the course of their binary fission, into microcolonies (Abedon 2011a, 2012, 2017a, b). These microcolonies are equivalent to the formation by bacteria of colonies as seen when those bacteria are plated at lower densities, except that microcolonies are smaller than colonies. Indeed, microcolonies are individually microscopic, essentially by definition. The smooth texture of a typical bacterial lawn as generated for plaque growth, in other words, is not smooth if observed under magnification but instead will tend to consist of closely associated microcolonies, each of which has formed from an individual indicator bacterium colony-forming unit. In addition, prior to maturation of these lawns, these microcolonies will be smaller and more separated than will be the case as they grow from individual colony-forming units to the maximum sizes they can attain within lawns. The complication on plaque formation that stems from bacteria displaying this clumped dispersion, as microcolonies within a bacterial lawn, is not only that these “clumps” become larger as lawn development progresses, but that bacterial physiology as well as phage access to bacteria will tend to differ across individual microcolonies, i.e., microcolony outer regions, at least following the biofilm model, should be more accessible to phages as well as better suited physiologically to supporting phage infections (Abedon 2016a, 2017b). To the extent that the existence of bacteria as microcolonies affects phage dynamics during plaque formation, then that impact may vary over the course of plaque formation, that is, vary over the course of bacterial clumps growing larger and physiologically aging over time. Inner regions of microcolonies in particular will likely display physiologies that are closer to those of stationary phase (Dennehy et al. 2007) than will the outer regions of microcolonies, and these inner regions will be expected to grow in size within the bacterial lawn as microcolonies mature. What is the impact of bacteria existing as microcolonies during plaque formation? As has been suggested elsewhere (Abedon and Yin 2008; Abedon 2011a), this clumping may impact rates of plaque formation overall, may give rise to plaques displaying a relatively constant rate of plaque-size increase, may explicitly represent what in toto must be removed to result in the substantial clearing of lawn bacteria that is required for plaque visual presence, and also probably is what gives rise to

Detection of Bacteriophages: Phage Plaques

527

gradations in plaque turbidity. For the latter, regions of greater plaque turbidity may be associated with greater numbers of intact microcolonies or, instead, with similar numbers of microcolonies but with those that remain larger in size due, for example, to phages lysing microcolony-center bacteria more slowly than they may be able to lyse those bacteria which instead are found on microcolony peripheries (Abedon 2016a). In particular, it is important to keep in mind that the leading edge of developing plaques changes in position over time. Therefore, with time, the microcolonies reached by a plaque’s leading edge will be larger, implying that the dynamics of phage-bacterial interaction will not necessarily have followed identical paths across a plaque’s breadth. The visibility of plaques against a bacterial lawn may be a function in many cases of the degree to which the microcolonies that make up a bacterial lawn have been depleted either quantitatively or qualitatively over time (Abedon and Yin 2008): Closer to a plaque’s center, microcolonies will have been smaller at the point of initial phage encounter. As a consequence, these microcolonies may be more easily cleared and thereby no longer able to contribute to plaque turbidity. Closer to a plaque’s periphery, microcolonies will have been larger at the point of initial phage encounter, simply because they will have been growing for longer prior to that encounter. Therefore, these microcolonies may be less easily cleared or less easily substantially reduced in size. Phages that are adept at more completely lysing the bacteria associated with larger microcolonies therefore should display less turbidity near plaque peripheries and might also, all else held constant, produce plaques that are at least moderately larger. The process of microcolony infection and depletion is crucial to plaque formation and therefore deserves additional discussion. In models of plaque formation (Krone and Abedon 2008), particularly what are known as reaction-diffusion models, it is typical to assume that bacteria display a random dispersion within lawns, that is, with individual cells spatially isolated from one another. Indeed, it is common to assume in models that bacterial densities within these lawns do not change over time. If densities do change, however, then conditions during plaque formation, particularly at a plaque’s leading edge, also must change. As a consequence, it becomes surprising that rates of plaque-size increase tend to remain constant over time, that is, until what appears to be an abrupt cessation of plaquesize increase as lawns enter into stationary phase. How to reconcile these two issues is an open question – how conditions changing over the course of plaque formation, on the one hand, and apparent constancy of dynamics of plaque-size increase, on the other, might coincide. Perhaps constancy of rates of growth in plaque diameter may be a consequence simply of a balancing of numerous factors, some of which should increase rates of plaque-size growth as lawns mature, while others at the same time should decrease those rates (Abedon 2011a). That explanation, however, does not represent a rigorously worked out answer to this question, so it certainly could benefit from further experimental as well as theoretical consideration.

528

S. T. Abedon

Plaque Size

Plaque Diameter

Plaque formation can in most cases be viewed as a race between phage-induced bacterial lysis, on the one hand, and bacterial lawn maturation on the other. With maturation the lawn reaches stationary phase and this is typically refractory to continued phage replication as well as to phage-induced bacterial lysis; a notable exception however is phage T7 which forms plaques which can continue to expand as a consequence of phage replication despite the bacteria entering into stationary phase (Yin 1991). As plaque formation involves alternating phage infection, bacterial lysis, and virion diffusion – that is, reaction-diffusion – the more extensively that phages can continue these processes, particularly spreading outward and otherwise lysing lawn bacteria, then the larger plaques should become. Thus, plaque size is typically a function of a combination of the speed of plaque-size increase during

Time of plaque formation, which typically occurs at a more or less constant rate

Time prior to initial phage adsorption with lawn bacterium

Cessation of plaque growth as occurs when lawn reaches a phage-infection refractory stationary phase

Time

Fig. 5 Formation of a plaque in terms of its diameter. Plaque formation does not begin until the initiating infective center gives rise to a bacterial infection within the lawn. This initial infection can occur more rapidly when a plaque is initiated with a phage-infected bacterium rather than with a free phage because it takes time for phage adsorption to occur when starting plaques with free phages. The initiation of plaques with phage-infected bacteria also can help to limit variability in plaque size. Following virion release from the initial phage infection, plaques then will increase in diameter at what has tended experimentally to be a more or less constant rate. This increase will then cease at the point where the bacterial lawn is no longer able to support continued phage infection or lysis, though soluble enzymes, such as phage-produced bacterial capsule digesting enzymes, may continue to diffuse as well as function (even at refrigerated temperatures), generating what are known as plaque halos (Sutherland et al. 2004). A plaque’s diameter at the point of lawn maturation to a phage-induced-lysis refractory state is a function of a combination of when the plaque-initiating infection first occurs relative to lawn initiation, initial bacterial density, initial bacterial physiological state, how fast the plaque increases in diameter while still growing, and how long that growth can continue, with the latter usually terminated at the point where the bacterial lawn enters into stationary phase. Note that generally we are referring to visible lawn clearing when referring to plaque size, but nonetheless that it is possible for both free phages and phage-infected bacteria to be found outside of the region of observable clearing

Detection of Bacteriophages: Phage Plaques

529

plaque formation, which as noted can occur more or less at a constant rate, and the duration of plaque growth. See Fig. 5. The duration of plaque growth is affected by both phage and bacterial properties. The earlier that the initial phage adsorption occurs during lawn maturation, then the longer that plaque growth will occur. The result, all else held constant, is a larger plaque. Note, though, that these issues do not necessarily speak to the morphology of the resulting plaque. The greater a phage’s potential to penetrate into lawn microcolonies and then lyse adsorbed bacteria, especially late during plaque development, then presumably the greater the declines in lawn turbidity within a plaque that may be observed, especially nearer a plaque’s periphery. The lower the turbidity near a plaque’s periphery then the clearer the overall plaque and, potentially, the larger the overall lawn clearing that defines the size of a plaque. Additional phage-related issues are the length of the phage latent period and the rate of virion diffusion (Abedon and Culler 2007a; Gallet et al. 2011). That is, longer latent periods should result in slower plaque growth due to less time that virions spend diffusing, whereas faster as well as longer durations of virion diffusion should result in faster plaque spread. Note that physical and chemical aspects of the environment can impact these phage properties (Abedon and Yin 2009). Shorter phage latent periods, for example, may be seen with richer media (Hadas et al. 1997). So too, as a result, may bacterial lawns mature earlier due to faster bacterial population growth, thereby offsetting potentially faster plaque growth rates associated with shorter phage latent periods. As a further complication, and as noted below, bacteria in richer media likely will grow to higher densities, increasing the duration of plaque formation since it then can take longer for bacterial populations to reach higher overall densities. Conversely, however, greater bacterial population growth may be associated with larger microcolonies and these larger microcolonies in turn may be more resistant, particularly within their interiors, to phage-induced lysis. Greater agar densities or greater agar dryness, in turn, can result in slower virion diffusion and therefore slower rates of plaque growth. Greater durations of plaque growth can result when bacterial lawns are initiated with fewer bacteria. In some cases it also may be possible to treat bacteria in such a way that they remain phage sensitive but nonetheless display a delay in the initiation of lawn growth, e.g., such as for certain phages via UV treatment (Stent 1963). Greater nutrient availability as well as buffering of bacterial waste products, the latter as can modify media pH, should increase the duration of lawn growth, which can be achieved both via the use of media that possesses greater nutrient densities and by employing thicker bottom agar in soft-agar overlays, which can serve as reservoirs of nutrients as well as sinks for wastes. If the bottom agar is too thick, however, then it may visually obscure plaques. Though starting with fewer bacteria can result in longer lawn growth, and therefore larger plaques, initiating lawns with fewer bacteria can result in excessively grainy lawns. Initiating lawns with fewer bacteria also can delay phage adsorption to uninfected microcolonies. Since on average phage diffusion before encountering a bacterium will be longer when starting with fewer bacteria, this can have the effect of delaying the initiation of plaque formation. Delays prior to the initial adsorption

530

S. T. Abedon

event also can result in variation in the timing of initiation of plaque formation and thereby variation in resulting plaque sizes. Preadsorbing phages to bacteria thus has the effect of more or less synchronizing the initiation of plaque formation across a phage population. Since more than one virion should be released from a preadsorbed bacterium per burst, subsequent rounds of phage adsorption, as propagated by what are now highly localized phage populations rather than individual virions, should be more consistent across different plaques. Note that one can also slow lawn growth by employing less-rich media. This also can lengthen phage latent period, which should slow rates of growth in plaque diameters, but may have little effect on virion diffusion rates. Lawns however will tend to stop growing earlier in terms of numbers of bacteria given reduced nutrient availability, thereby resulting in less potential for this approach to result in increased plaque size. See especially Abedon and Yin (2009) for further discussion of these issues, as well as Gallet et al. (2009, 2011), Roychoudhury et al. (2014), plus Box 1 for summary. Box 1 Phenomena That Can Potentially Result in Larger Plaques

• Shorter phage latent periods due to – Phage genetic predisposition – Higher nutrient quality or density – Greater phage physiological compatibility with bacterial host • Faster virion diffusion due to – Lower agar densities – Lower agar molecule length – Lower agar purity – Higher agar moisture content (i.e., less dried out agar) – Virion physical properties • Longer virion diffusion prior to bacterial adsorption due to – Lower bacterial densities (this though can lead to grainy lawns as well as delayed adsorption) – Lower phage adsorption affinity to bacteria (perhaps of lower importance, however, given bacteria existence within microcolonies since there virion failure to attach to one encountered bacterium should be followed with relatively high likelihood by virion encounter with adjacent bacteria) • Earlier plaque-initiating phage adsorption to host bacterium due to – Chance earlier adsorption – Preadsorption (earlier adsorption that is not due to chance) – Higher initial lawn densities – Faster phage virion diffusion (speculative) • Longer lawn growth due to – Greater nutrient availability or buffering capacity of bottom agar – Greater nutrient density or quality in top agar (continued)

Detection of Bacteriophages: Phage Plaques

531

Box 1 Phenomena That Can Potentially Result in Larger Plaques (continued)

– Bacterial physiological characteristics – Particularly relevant if phage population growth rates are not equivalently affected – Can be relevant especially if virion diffusion rates are less affected • Phage ability to productively infect and lyse physiologically older bacteria – Particularly to the extent that stationary phase lawns are permissive to plaque growth – Ability to lyse older bacteria delays termination of plaque growth – For example, as can be seen with phage T7 • Larger Phage Burst Size – Has an impact but not as large as equivalent changes in latent period or virion diffusion – Impact nonetheless is large given initially small phage bursts sizes, e.g., the jump from a burst size of 10–20 should have a larger impact on plaque size than a jump from 20 to 40

Plaque-Based Phage Characterization A number of plaque-based techniques exist that allow relatively easy determination of certain phage characteristics. In many cases, however, these determinations are based upon what may be unexpectedly complex biology and/or which should be viewed as only preliminary determinations, preliminary particularly if rigorous characterization of phage properties is ultimately desired. Nonetheless, these techniques are fairly powerful and also somewhat widespread in their use, especially efficiency of plating (EOP) determinations (Adams 1959). Efficiency of center of infection (ECOI) assays provide related but nonetheless different information pertaining to phage viability (Sing and Klaenhammer 1990), and mixed- or dualindicator assays may be employed as a first step towards the characterization of certain aspects of phage host range (Adams 1959; Stent 1963), particularly when that host range has mutated or when seeking to bias phage isolation towards those phages possessing specific host-range characteristics.

Efficiency of Plating An important though surprisingly biologically complex means of phage characterization is the determination of efficiency of plating (EOP) (Adams 1959). EOP can take on two forms, absolute versus relative. Absolute EOP compares plaque number with virion particle number, the latter as determined, for example, via electron microscopy (chapter ▶ “Detection of Bacteriophages: Electron Microscopy and

532

S. T. Abedon

Visualization”). With absolute EOP determinations, EOPs of less than 1.0 can be due either to virion inviability or less than perfect plaque forming ability under a given set of conditions that are present despite an ability of an infecting virion to produce new phages (e.g., low efficiency of plating such as due to production of small burst sizes). Far more common, however, is the comparison of plaque numbers using different bacterial hosts, a.k.a., different indicator bacteria, or, less commonly, different plating conditions, which are described instead as relative EOPs. In all cases, however, a specific dilution of a given phage stock will give rise to different numbers of plaques depending on differences in host or other plating conditions. There are two general basic mechanisms that can give rise to a reduced relative EOP. Rightly or wrongly, however, one explanation tends to be preferentially assumed versus the other. The preferred explanation considers particularly the viability of phage infections, which is assumed to be lower given a lower EOP. Thus, for example, if half of infections are nonviable then an EOP of 0.5 may result, and this often appears to be assumed to be a consequence of only half as many plaques being initiated. The alternative explanation is that sufficient “plaques” may form to result in an “EOP” of 1.0 – that is, “plaques” as areas of reduced numbers of bacteria such as due to phage-induced bacterial lysis and “EOP” as relative numbers of these “plaques” – but these “plaques” may not become large enough to become visible (hence the quotes). This can be because plaque initiation is delayed, because plaque growth is slow such as due to long phage latent periods or small phage burst sizes, instead because lawn growth is particularly fast, or indeed as due to some combination of these. In any case, here it is not that phages are not producing progeny at the beginning of plaque formation, that is, the phages are not inviable, but instead that phages are not producing enough progeny fast enough, along with subsequent bacterial infection and lysis, to ultimately give rise to visible plaques. Reduced EOP thus should not automatically be assumed to be equivalent to reduced phage viability, as low EOPs alternatively can be due to what may be described as a reduced infection “vigor.” This latter term was coined explicitly to describe the potential for bacterial abortive infection systems to give rise to low EOPs – often a defining phenotype associated with these mechanisms – without necessarily preventing infecting phages from producing virion progeny (Hyman and Abedon 2010). To distinguish among these two explanations, reduced phage viability (complete loss of infection “vigor”) versus less than fully reduced infection vigor, it is simplest to make an effort to determine what can be described as efficiency of center of infection (ECOI), rather than relying solely on EOP determinations. See Abedon and Yin (2009) for discussion of alternative approaches.

Efficiency of Center of Infection Reduced EOP due to reduced phage viability, contrasting reduced infection “vigor,” can be distinguished using a modification of EOP determination called efficiency of center of infection or ECOI (Sing and Klaenhammer 1990). The

Detection of Bacteriophages: Phage Plaques

533

resulting distinction is useful in terms of phage phenotypic characterization, e.g., in comparing phage mutants to wild type, and in characterizing what are known as phage abortive infection systems. The latter are bacterial mechanisms that result in a combination of bacterial death and reduction in phage infection vigor (Hyman and Abedon 2010; Labrie et al. 2010). With standard EOP determinations these two possibilities may not be distinguishable (previous section). Alternatively, with ECOI determination in most cases they should be readily distinguished. With ECOI assays, one employs preadsorption such that plaques are initiated using phage-infected bacteria rather than with free phages. These phage-infected centers – centers of infection – are separated from any residual free phages, e.g. by washing (repeated pelleting and resuspension of phage-infected bacteria while discarding free phages with the supernatant), and then they are plated along with indicator bacteria and/or under conditions that are known to be highly permissive to plaque formation by the phage in question. Thus, the plated phages during ECOI determinations are exposed to two different types of bacteria. The first is that to which the phages are preadsorbed (which could be described as the selective strain) and these bacteria may or may not be productively infected and thus may or may not lyse and release progeny phages. The second bacterial type functions solely as the indicator bacteria and serves exclusively as a means of detecting the presence of released virions from the first infection. Ideally the phage being tested will display a high EOP on this second, indicator strain. Therefore, only the first infection provides selective infection conditions, contrasting standard EOP determinations where all infections, first plus all subsequent infections during plaque formation, provide selective infection conditions. Avoidance of directly plating free virions is a crucial aspect of successful execution of the ECOI technique. It also makes ECOI determinations technically more difficult than EOP determinations, since any free phages that are not successfully removed prior to plating should be able to readily form plaques given properly chosen indicator bacteria, that is, inadvertently plated free phages will give rise to ECOI false positives. As a means of controlling for this issue, it is possible to plate following washing for remaining free phages only, e.g., such as by exposing infective centers to chloroform prior to plating. With ECOI determinations, if the initial infection is not viable (phage virions are not produced) then no plaque will form. Alternatively, if the initial infection displays reduced vigor, then released phages (ideally) will nevertheless be able to form plaques on the permissive bacteria making up what will become the bacterial lawn. Inviability of plaque-initiating bacterial infections thus results in a reduced EOP as well as a reduced ECOI, whereas viable phages with only reduced vigor during infection of preadsorbed, plaque-initiating bacteria should result in a reduced EOP but not in a reduced ECOI. An exception, however, is if the initial infection is viable but nonetheless extremely long in duration prior to lysis, e.g., as may be seen given lysogenic rather than lytic cycles. These ideas are summarized in Fig. 6. If one is interested in unambiguously ascertaining the impact of phage genetics, bacterial hosts, or plaquing conditions on phage viability, then it is crucial to employ ECOI

S. T. Abedon

Low EOP

High EOP

534

High ECOI

Low ECOI

Well-Behaved Plaquing

Possible Experimental Error

High phage viability and high infection vigor, e.g., as seen with an effectively plaquing wild-type phage

This can occur for example given poor indicator bacteria choice or accidentally poor plating conditions

Low Infection Vigor

Low Phage Viability

Plated phages display high infection viability but nonetheless poor plaque-forming ability (EOP) on that same host strain

Initial infection during plaque formation on host strain of interest has a relatively low likelihood of successful virion production

Fig. 6 Conclusions reached given EOP versus ECOI determination. Note that full understanding can be reached only by employing additional approaches to phage phenotypic characterization such as burst size and latent period determination, e.g., via single-step growth experiments (Hyman and Abedon 2009). That is, ECOI alone can detect low phage viability but cannot detect low infection vigor whereas EOP alone cannot distinguish between low phage viability and low infection vigor. Used together, however, it is possible to both detect and distinguish low phage viability and low infection vigor

rather than standard EOP determinations. Another way of stating this is that EOP determinations in many or most instances should be viewed as only preliminary characterizations of phage properties. Alternatively, rather than employing ECOI determinations, one can do infection-viability experiments in liquid culture (Abedon and Yin 2009).

Mixed-Indicator Technique The mixed- or dual-indicator technique (Adams 1959; Stent 1963) is a means of rapidly ascertaining phage host range breadth rather than overall phage viability or infection vigor. This is done especially when isolating phage host range mutants such that thousands of presumptive phage mutants may be screened simultaneously for expansion of host range (i.e., now able to lyse two different bacterial strains but previously only one) or contraction of host range (i.e., now able to lyse only one of previously two different bacterial strains). The mixed-indicator technique also allows a rapid initial screening of host range breadth in the course of phage isolation. The utility of this technique is that it is far easier to perform than to instead determine

Detection of Bacteriophages: Phage Plaques

535

No Plaques, Indicator 1

Plaques, Indicator 1

host range in a more step-wise manner, and the potential for false positives with the mixed-indicator technique is lower than is the case with spot testing. Here two strains of indicator bacteria are employed on a single plate. Assuming that a phage gives rise to clear plaques on a single indicator, if a phage is able to infect and lyse both bacterial strains, then the resulting plaques will be clear, or at least not highly turbid. Because this clearing occurs in the course of plaque formation, it is indicative of lysis of both strains, since during plaque formation phage virions spread into the maturing bacterial lawn. This contrasts with spot formation which is indicative only that something is present in a phage lysate that is capable of inhibiting bacterial replication especially early during lawn formation (above). Proof that this dual-indicator lysis is due to phage action, as well as indication of phage ability to replicate using both bacteria as hosts, requires further characterization, such as plaque formation on each host individually and/or broth-culture characterization. Key, though, is that these subsequent steps need to be done only with presumptive positive outcomes rather than for thousands of possible positives. If the plaques formed using dual indicator are turbid, in particular given somewhat less turbid plaques when plating parental phage strains using single-indicator lawns, then the phage probably is no longer able to infect and lyse both hosts. An absence of plaque formation altogether, meanwhile, is of course suggestive of an Plaques, Indicator 2

No Plaques, Indicator 2

Clear Plaque

Turbid Plaque

Both indicators efficiently lysed (though you can’t tell whether both are productively infected)

Only one indicator is efficiently lysed (but you must do more experiments to figure out which one)

Turbid Plaque

No Plaque

Only one indicator is efficiently lysed (but you must do more experiments to figure out which one)

Neither indicator is efficiently lysed (either both don’t support plaquing or it’s more complicated)

Fig. 7 Interpretation of mixed-indicator results. This technique can serve a good first step towards identifying a phage’s plaquing host range and can be particularly useful if that host range is being experimentally modified. Specifically, change from turbid to clear plaques, such as observation of a clear plaque among many turbid ones, is suggestive of a phage mutant that displays a broader host range than its parent (i.e., newly acquired ability to lyse the second, “test,” bacterial strain) whereas change from clear to turbid (such as one turbid plaque among many clear plaques) is suggestive that a phage mutant displays a host range that is narrower than that of its parent

536

S. T. Abedon

inability to plaque on either host strain. Determining which of the two hosts is permissive to plaque formation requires further analysis (see Fig. 7 for summary). The mixed-indicator technique thus, ideally, will represent the start of phage hostrange characterization rather than necessarily an end point.

Conclusions Phage plaques are more complicated entities than many may be aware, and also are not identical to phage spots. Plaques do not simply happen but instead are the product of relatively complex as well as imperfectly understood phage-bacterial population dynamics. They are certainly an indication of phage viability, though an absence of plaque formation is not necessarily equivalent to phage inviability (and spot formation, it should be added, is not necessarily an indication of phage viability). To a first approximation, one can distinguish between these possibilities by using the relatively common technique of efficiency of plating determination, though to distinguish between inviability and merely low phage performance it is important as well to employ, for example, efficiency of center of infection determinations. In addition to phage isolation, purification, and characterization, phage plaques are also routinely employed in phage enumeration, as considered in a different chapter of this volume (chapter ▶ “Detection of Bacteriophages: Statistical Aspects of Plaque Assay”).

References Abedon ST (1994) Lysis and the interaction between free phages and infected cells. In: Karam JD, Kutter E, Carlson K, Guttman B (eds) The molecular biology of bacteriophage T4. ASM Press, Washington, DC, pp 397–405 Abedon ST (2011a) Bacteriophages and biofilms: ecology, phage therapy, Plaques. Nova Science Publishers/Hauppauge, New York Abedon ST (2011b) Lysis from without. Bacteriophage 1:46–49 Abedon ST (2012) Thinking about microcolonies as phage targets. Bacteriophage 2:200–204 Abedon ST (2015) Bacteriophage secondary infection. Virol Sin 30:3–10 Abedon ST (2016a) Bacteriophage exploitation of bacterial biofilms: phage preference for less mature targets? FEMS Microbiol Lett 363:fnv246 Abedon ST (2016b) Phage therapy dosing: the problem(s) with multiplicity of infection (MOI). Bacteriophage 6:e1220348 Abedon ST (2017a) Active bacteriophage biocontrol and therapy on sub-millimeter scales towards removal of unwanted bacteria from foods and microbiomes. AIMS Microbiol 3:649–688 Abedon ST (2017b) Phage “delay” towards enhancing bacterial escape from biofilms: a more comprehensive way of viewing resistance to bacteriophages. AIMS Microbiol 3:186–226 Abedon ST (2017c) Plaques. In Reference module in life sciences. Elsevier. https://doi.org/10.1016/ B978-0-12-809633-8.06915-6 Abedon ST (2017d) Commentary: communication between viruses guides lysis-lysogeny decisions. Front Microbiol 8:983 Abedon ST, Culler RR (2007a) Bacteriophage evolution given spatial constraint. J Theor Biol 248:111–119 Abedon ST, Culler RR (2007b) Optimizing bacteriophage plaque fecundity. J Theor Biol 249:582–592

Detection of Bacteriophages: Phage Plaques

537

Abedon ST, Katsaounis TI (2017) Basic phage mathematics. Methods Mol Biol 1681:3–30. https:// www.ncbi.nlm.nih.gov/pubmed/29134583 Abedon ST, Thomas-Abedon C (2010) Phage therapy pharmacology. Curr Pharm Biotechnol 11:28–47 Abedon ST, Yin J (2008) Impact of spatial structure on phage population growth. In: Abedon ST (ed) Bacteriophage ecology. Cambridge University Press, Cambridge, UK, pp 94–113 Abedon ST, Yin J (2009) Bacteriophage plaques: theory and analysis. Methods Mol Biol 501:161–174 Adams MH (1959) Bacteriophages. InterScience, New York Carlson K (2005) Working with bacteriophages: common techniques and methodological approaches. In: Kutter E, Sulakvelidze A (eds) Bacteriophages: biology and application. CRC Press, Boca Raton, pp 437–494 Carlson K, Miller ES (1994) Enumerating phage: the plaque assay. In: Karam JD (ed) Molecular biology of bacteriophage T4. ASM Press, Washington, DC, pp 427–429 d’Hérelle F (1917) Sur un microbe invisible antagoniste des bacilles dysentériques. C R Acad Sci Ser D 165:373–375 d’Hérelle F (2011) On an invisible microbe antagonistic to dysentery bacilli. Note by M. F. d’Herelle, presented by M. Roux. Comptes Rendus Academiedes Sciences 1917; 165:373–5. Bacteriophage 1:3–5 Dennehy JJ, Abedon ST, Turner PE (2007) Host density impacts relative fitness of bacteriophage Φ6 genotypes in structured habitats. Evolution 61:2516–2527 Dulbecco R (1949) Appendix: on the reliability of the Poisson distribution as a distribution of the number of phage particles infecting individual bacteria in a population. Genetics 34:122–125 Gallet R, Shao Y, Wang I-N (2009) High adsorption rate is detrimental to bacteriophage fitness in a biofilm-like environment. BMC Evol Biol 9:241 Gallet R, Kannoly S, Wang IN (2011) Effects of bacteriophage traits on plaque formation. BMC Microbiol 11:181 Hadas H, Einav M, Fishov I, Zaritsky A (1997) Bacteriophage T4 development depends on the physiology of its host Escherichia coli. Microbiology 143:179–185 Hyman P (1993) The genetics of the Luria-Latarjet effect in bacteriophage T4: evidence for the involvement of multiple DNA repair pathways. Genet Res 62:1–9 Hyman P, Abedon ST (2009) Practical methods for determining phage growth parameters. Methods Mol Biol 501:175–202 Hyman P, Abedon ST (2010) Bacteriophage host range and bacterial resistance. Adv Appl Microbiol 70:217–248 Krone SM, Abedon ST (2008) Modeling phage plaque growth. In: Abedon ST (ed) Bacteriophage ecology. Cambridge University Press, Cambridge, UK, pp 415–438 Kropinski AM, Mazzocco A, Waddell TE, Lingohr E, Johnson RP (2009) Enumeration of bacteriophages by double agar overlay plaque assay. Methods Mol Biol 501:69–76 Labrie SJ, Samson JE, Moineau S (2010) Bacteriophage resistance mechanisms. Nat Rev Microbiol 8:317–327 Mazzocco A, Waddell TE, Lingohr E, Johnson RP (2009a) Enumeration of bacteriophages by the direct plating plaque assay. Methods Mol Biol 501:77–80 Mazzocco A, Waddell TE, Lingohr E, Johnson RP (2009b) Enumeration of bacteriophages using the small drop plaque assay system. Methods Mol Biol 501:81–85 Millard AD (2009) Isolation of cyanophages from aquatic environments. Methods Mol Biol 501:33–42 Mirzaei MK, Nilsson AS (2015) Isolation of phages for phage therapy: a comparison of spot tests and efficiency of plating analyses for determination of host range and efficacy. PLoS One 10: e0118557 Pires DP, Oliveira H, Melo LD, Sillankorva S, Azeredo J (2016) Bacteriophage-encoded depolymerases: their diversity and biotechnological applications. Appl Microbiol Biotechnol 100:2141–2151 Rizvi S, Mora PT (1963) Bacteriophage plaque-count assay and confluent lysis on plates without bottom agar layer. Nature 200:1324–1325

538

S. T. Abedon

Roychoudhury P, Shrestha N, Wiss VR, Krone SM (2014) Fitness benefits of low infectivity in a spatially structured population of bacteriophages. Proc Biol Sci 281:20132563 Salivar WO, Tzagoloff H, Pratt D (1964) Some physical-chemical and biological properties of the rod-shaped coliphage M13. Virology 24:359–371 Sanders ER (2012) Aseptic laboratory techniques: plating methods. J Vis Exp 11:e3064 Sing WD, Klaenhammer TR (1990) Characteristics of phage abortion conferred in lactococci by the conjugal plasmid pTR2030. J Gen Microbiol 136:1807–1815 Stent GS (1963) Molecular biology of bacterial viruses. WH Freeman, San Francisco Sutherland IW, Hughes KA, Skillman LC, Tait K (2004) The interaction of phage and biofilms. FEMS Microbiol Lett 232:1–6 Yin J (1991) A quantifiable phenotype of viral propagation. Biochem Biophys Res Com 174:1009–1014 Yin J (1993) Evolution of bacteriophage T7 in a growing plaque. J Bacteriol 175:1272–1277 Yin J (1994) Spatially resolved evolution of viruses. Ann N Y Acad Sci 745:399–408

Detection of Bacteriophages: Statistical Aspects of Plaque Assay Stephen T. Abedon and Tena I. Katsaounis

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . TFTC and TNTC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spot Counts Versus Plate Counts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Number of Repeats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Utility of Trimmed Means . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Number of Dilution Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Utility of Larger Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . When Comparisons Matter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

540 541 547 553 554 555 556 557 559 559 560 560

Abstract

Approaches to enumerating bacteriophages, as with microorganisms generally, tend to be limited by the small size of the subjects, which prevents unaided visualization of individuals. For phages, visualization is further hampered by our inability in most cases to view individual viruses even employing standard brightfield microscopy, which in any case would supply total rather than viable counts. A standard solution in microbiology towards obtaining viable counts involves plating, typically using Petri dishes to hold the plating medium and plated organisms. For cellular microorganisms we thus enumerate not individual viable cells but instead macroscopic clumps of cells as derived from the growth of individuals, that is, what we refer to as colonies. For phages, as well as viruses S. T. Abedon (*) Department of Microbiology, The Ohio State University, Mansfield, OH, USA e-mail: [email protected] T. I. Katsaounis Department of Mathematics, The Ohio State University, Mansfield, OH, USA e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_17

539

540

S. T. Abedon and T. I. Katsaounis

generally, such clumps not only cannot form on their own accord when employing solely abiotic growth media but also would tend to be invisible to the naked eye even given such concentrating. Instead, towards phage enumeration as well as viability determination, along with characterization of other properties, often what one counts are phage plaques. These too are areas of high organism concentration, though here as generated by localized phage replication rather than bacterial population growth. Plaques are not directly visible due to local phage presence, but instead as a consequence of a full or partial local absence of cells, that is, reduced cell numbers due to this localized phage propagation. Here we consider especially statistically derived best practices of plaque-based phage enumeration.

Introduction Plaquing may be employed as a means of phage isolation, of establishing pureculture phage stocks (see ▶ “Isolation of Bacteriophages” chapter), of characterizing phage properties such as in terms of phage host range, and of distinguishing among different phages such as in terms of plaque size or plaque morphology (see ▶ “Detection of Bacteriophages: Phage Plaques” chapter). Phage plaques are localized regions of host-bacteria depletion as can form given phage propagation within spatially constrained environments (typically, agar lawns of bacteria). Such plaques naturally tend to grow as spheres. Within a bacterial lawn as plated in a Petri dish, however, these spheres typically are truncated on their tops and bottoms, at the agar or soft-agar boundaries, thereby taking on more columnar shapes. Further, as viewed from above, these columns appear as “holes” within otherwise turbid bacterial lawns and these holes – given the cross section of both spheres and, horizontally, columns – tend to be circular in shape. Given sufficient size as well as lack of turbidity, these circular holes in bacterial lawns may be readily counted. Furthermore, individual plaques ideally will have formed from individual plated phages, consisting either of individual virions or instead of individual virion-infected bacteria, that is, of plaqueforming units (PFUs), and this is rather than from clumps of phages or phageinfected bacteria. Each individual plaque thus, ideally, will correspond to each individual, plated, viable phage or phage-infected bacterium rather than to multiple individual phages or multiply phage-infected bacteria. Plaques thereby provide a “trick,” enabling visualization and thereby enumeration of phages (Carlson and Miller 1994; Kropinski et al. 2009; Mazzocco et al. 2009a, b). Here we consider not so much plaque properties, or techniques towards their formation, but instead enumeration once plaques have formed. Assuming that the plaques in question are both large and clear enough to allow relatively unambiguous identification, then these plaques can be counted towards determination of that number of virions which had been plated. Taking into account the dilutions employed prior to that plating and given that each plaque corresponds to only a single plated phage, then plaque-based enumeration can provide a titer value associated with a phage-containing environment. Particularly in terms of the basic

Detection of Bacteriophages: Statistical Aspects of Plaque Assay

541

statistical principles as can constrain the quality of data that can be obtained from this otherwise simple laboratory procedure, here we consider such enumeration. Our combined goal is to make readers aware of where error can enter into such enumerations, on the one hand, and, on the other, how to limit such error through a combination of good laboratory practice and especially effective statistical handling of data.

Overview The basic plaquing scheme typically begins with a volume of either some unknown or instead not accurately known phage concentration. This is followed by serial dilution to numbers of PFUs that will allow for both accurate and precise plaque counting, i.e., of plaques as ultimately will be found on plates (or within spots) as prepared with each of the serial dilution steps, and then plating these dilutions in combination with permissive indicator bacteria. Plaque-based phage enumeration will often then be based on plaque counts from multiple plates so as to determine a single titer value. Concepts relevant to achieving both accurate (unbiased) and precise (small random error) plaque count determinations include: how many plaques per plate should represent “Too Numerous To Count” (TNTC) (an issue of data validity); how many PFUs should define “Too Few To Count” (TFTC) (also an issue of validity, specifically of data precision); how to minimize biases so as to improve accuracy; and which number of enumeration repeats is reasonable in practice, the latter also to achieve desirable levels of precision. Specifically, for any enumeration procedure there could be conflicts between maximizing accuracy (i.e., by reducing bias), maximizing precision (i.e., by minimizing unbiased random error), minimizing costs, and otherwise obtaining data with without excessive effort. For further discussion of the concepts of accuracy versus precision, see Moore and Notz (2014). Towards obtaining reasonable precision and accuracy for titer values, but without excessively impinging on researcher time and resources, we suggest in particular the gathering of at least three independently obtained plate counts per calculated titer value. Specifically, having more than two plate-count determinations is key and three observations is the bare minimum of “more than two.” Furthermore, we suggest that titer calculations should be based on trimmed means rather than on an averaging all obtained plate counts, with the concept of a trimmed mean most familiarly seen with the use of medians (see Table 1 for definition of trimmed mean and DeGroot (1975), for further discussion). The resulting titer value specifically would then be presented without indication of variability, e.g., plate count repeats of 60, 71, and 92 together could be presented simply as 71 (as based, in this case, on n = 3). Use of medians will tend to improve estimation validity via the elimination of outlier values in the course of median generation, but at the cost of some precision. Further justification is provided below for this approach along with, in terms of TNTC and TFTC, what range of counts per plate may be deemed acceptable.

542

S. T. Abedon and T. I. Katsaounis

Table 1 Defining terms relevant to plaque-based phage enumeration Term or concept Accuracy

Coefficient of variation

Dilution series

Drop plaque method

Efficiency of Plating (EOP)

Independence

Definition Degree of lack of measurement biases, with less bias giving rise to greater accuracy; biases can be systematic, i.e., reoccurring in the same direction (including as might be seen in even given automated enumeration methods) or instead can be a consequence of operator error (as might be reduced given automatic enumeration). The former (systematic error) cannot be reduced, in terms of impact on accuracy, by increasing the total number of repeats of measurement whereas the latter (operator error) can be so reduced. The closer an estimation is to a “true” average, the greater is its accuracy. Note that accuracy can be improved without improving precision, that is, estimations can be improved without also improving one’s confidence in an estimation (i.e., mean-squared error, a measure of accuracy plus precision). A measure of the spread of data points relative to the mean. Here this is equal to the standard deviation of one’s observations divided by their mean. The coefficient of variation allows one to compare, for example, the expected spread of plate counts across different dilutions, with larger spreads associated with lower precision. Thus, lower plate counts inherently will have a lower associated precision than higher plate counts. Serial dilution scheme ending, in terms of plaque-based titering, with the plating of a given dilution. In general, to increase the accuracy of overall measurements, regardless of the number of platings of a single dilution series, the data from that series should give rise to only a single determined value. Maximally, then, each dilution series should contribute a single value of n, with use of only a single dilution series providing an n of only one no matter how many plates are made from that dilution series. Approach to phage plating involving spotting and gives rise to spot counts. The “Drop” here refers to the application of small volumes, such as 10 μl, to bacterial lawns prior to lawn incubation. Plaques are formed with the resulting spots over a relatively small portion of the overall Petri dish allowing for multiple dilution determinations, i.e., spot counts, per plate, but typically at a cost in precision to resulting titer determinations. Ratio of number plaque counts to some other means of PFU determination. With relative efficiency of plating, this is relative to plaque counts under a different set of conditions, often ones which are more permissive to plaque formation. With absolute efficiency of plating, this instead is relative to numbers of virion particles as determined visually (total counts). Determinations of efficiency of plating is an important example of where plaque counts may be compared. Efforts towards lack of carryover of operator error and/or methods of selecting observations on plaque formation between determinations. Generally, true independence is difficult to achieve but effort nevertheless should be made to make sure that levels of independence are reasonably high, e.g., such as by not counting multiple platings from a single dilution series, or even multiple platings of a single dilution, as independent. (continued)

Detection of Bacteriophages: Statistical Aspects of Plaque Assay

543

Table 1 (continued) Term or concept Mean

Mean, arithmetic Mean, geometric

Mean, trimmed (Trimmed mean)

Mean, true (“True” mean)

Median

Definition An average of more than one value, here as determined arithmetically (i.e., sum/n) but also as may be approximated using trimmed means (see below, this Table, for definition) or, as a specific form of trimmed mean, using medians. The number of data points used to calculate a mean is designated as n. Here this is used to specify that a mean is not trimmed, that is, rather than to contrast with a geometric mean. A mean calculated based on logarithms of data points (equivalently, the nth root of a product of n data points) rather than actual data points. The concept of geometric mean is not otherwise addressed in this chapter. Method of reducing biases associated with extreme values during mean determinations. At its simplest, a trimmed mean ignores extreme values but not the fact that such values exist. For example, for the values 30, 40, 50, and 600, the trimmed mean can be calculated – such as by ignoring the two extreme values (30 and 600) – as a value of 45 as equal to (40 + 50)/2. Note that this result is different from a value of 40 as equal to (30 + 40 + 50)/3 were the extreme value of 600 (and 600 alone) simply ignored (an example of a less statistically valid and indeed biased approach to generating a trimmed mean). More trimming results in greater robustness in estimations, though at a cost of precision, whereas the greater the original n, then the greater the precision of an estimation even when employing trimmed means (and also, as always, the smaller the range of values originally obtained, then the greater the precision). Typically, one sees degree of trimming reported as percentages, e.g., a 25% trimmed mean. See titering.phage.org or the Trimmean function in Excel as examples. Note that a median is a mean that has been trimmed 100%, while an arithmetic mean is a mean that has been trimmed 0%, and neither is necessarily equal to the true mean. Due to its trimming, a trimmed mean is less susceptible to the effects of extreme (low or high) counts than is an otherwise equivalent arithmetic mean. The actual value, of which sample means or medians represent only an estimation. This is the population value, e.g., as if every PFU in a phage stock were enumerated without bias. A “true” average in practice will not be determined but instead only approximated. The closer an estimation is to the “true” mean, the higher the accuracy of that determination. In general, the smaller the range of values associated with an unbiased estimation (i.e., 59, 60, and 61 vs. 40, 60, and 80), then the higher the precision of the estimation of the “true” mean. As typically determined, this is the central number of multiple values as equal either to the actual central value for an odd count of values (e.g., n = 3 or 5 or 7) or instead the average (mean) of the two otherwise most central values for an even count of values (e.g., n = 2 or 4 or 6). A median is a kind of trimmed mean (see above for definition) and indeed medians represent the “ultimate” trimmed mean in that all values except the central values are “trimmed” from the calculation. Use of medians will tend to produce more robust estimations by eliminating outlier values in the course of median generation, but at the cost of some precision. (continued)

544

S. T. Abedon and T. I. Katsaounis

Table 1 (continued) Term or concept n

Plaque-Forming Unit (PFU)

Plaque count

Plate count

Precision

Random error

Definition Number of independent repetitions of a determination, which for titer determinations generally may be equated with number of dilution series, as counted towards generation of a mean. For example, as based on three dilution series and platings, then n would equal three, though as based on three dilution series and six platings then n would also be equal to three. Entity that gives rise to a single plaque upon plating, e.g., a single virion particle or infected cell, though various forms of clumping (in which case more than one phage together could form a single plaque) and/or inactivation of virion particles can result in numbers of PFUs underestimating virus total counts. Equivalency between numbers of PFUs and numbers of plaques can also fail for reasons of overcrowding (inability to distinguish overlapping plaques), poor plaque-forming ability (failure of plaques to grow to a size which is large enough to be visualized), or mistaking artifacts for plaques (i.e., bubbles which have remained in poured agar can mimic plaques in terms of their counting). Number of plaques determined per some area of agar surface, i.e., the surface of single Petri dish, a randomly chosen subsection of a plate, or instead the area of a single spot. Plaque counts thus can be increased by increasing the total area counted either by increasing the fraction of the area counted (i.e., by counting the entire plate rather than a random subsection of a plate) or by increasing prior to plating the total area plated (i.e., by using larger Petri dishes). Because of their small area, spots will provide somewhat lower plaque counts than when plating using the whole surface area of Petri dishes. Plaque counts typically are used to infer numbers of plaque-forming units found within plated volumes. Number of plaques found as plated onto a Petri dish, where dish and plate are synonymous terms. A plate count thus is simply a plaque count as determined on a per-plate basis and often these terms are used synonymously. The plaque count associated with a spot by contrast is not a plate count. A measure of the trueness of a measurement, that is, how similar multiple measures of the same quantity are to each other. Note that one can have high precision while still having low accuracy, which would imply good data collection (i.e., resulting in a small range of values and/or high number of data points contributing those values) of otherwise bad (i.e., biased) data. As impacting precision, this is a randomly occurring procedural aberration that results in a deviation of results from an otherwise anticipated outcome. These deviations may or may not be the “fault” of an operator (e.g., lab technician) and may or may not impact the accuracy (degree of bias) of measurements, but nonetheless (as noted) do affect measurement precision. Random error also can be a function of n, with smaller values of n potentially leading to greater overall error than larger values of n (i.e., see random sampling error). Automated plaque counting methods may achieve reductions in random error, though could also introduce systematic errors, e.g., lower likelihoods of miscounting plaques (operator error) but in combination with consistent failures to distinguish one “plaque” into two or consistent missing of smaller or more turbid plaques (systematic error). (continued)

Detection of Bacteriophages: Statistical Aspects of Plaque Assay

545

Table 1 (continued) Term or concept Random sampling error

Serial dilution

Spot

Spot count

Systematic error (bias)

Titer

Definition Loss of precision that results from determining statistics based on only a subset of a larger population. Increasing losses in precision are seen the smaller the size of the subset. That is, lower sampling error generally is seen with greater n and vice versa. Compare, for example, plate counts with plaque counts based on spotting (lower precision seen with spotting) or plaque counts exceeding TFTC with ones not exceeding TFTC (lower precision seen with failures to exceed TFTC). Note that in addition to reductions in precision due to smaller sample sizes, too lower precision can be seen when handling smaller volumes of fluid, e.g., 10 μl as may be seen with spotting rather than 100 μl as may be employed instead to obtain plate counts. Means of reducing the concentration of an entity via multiple, consecutive dilution steps, e.g., tenfold and then another tenfold towards a total dilution of “100” fold. Note that a serial dilution need not involve multiple dilution vessels since phage plating of less than unit volume onto a plate counts as a dilution as well, e.g., the plating of 0.1 ml (100 μl) given a standard titer unit of virions, phages, or PFUs per ml is equivalent to a tenfold dilution. Plating of plaque-forming units as found within a small droplet such as 10 μl, which is dropped onto an already initiated bacterial lawn. Note that spots inherently support the growth of lower plaque counts than platings over larger areas so resulting plaque counts consequently will tend to less readily exceed TFTC, thereby resulting in inherently lower validity of plaque counts which are based on spotting. In addition, difficulties in handling the inherently lower volumes employed to initiate spots that may be used when obtaining plate counts also could contribute to lower validity with spot versus plate counts. Plaque count associated with a spot. Due to the inherently smaller area of spots, spot counts will tend to possess lower validity than plate counts. Specifically, due solely to the smaller area over which plaques form within spots, it is more difficult to exceed TFTC without also exceeding TNTC with spot-based plaque count determinations than generally is the case with plate count determinations. Spot counts therefore are considered to produce less precise titer determinations than plate counts and therefore, when greater validity is required such as prior to phage use during experiments, or towards generation of experimental data, spot counts should be viewed as preliminary data to be used as a guide toward subsequent plate count determination. As related to accuracy, these are errors associated with procedures that give rise especially to inaccuracy of measurement. Systematic errors cannot be reduced by reducing random sampling error, that is, by increasing n. In other words, an inherent inaccuracy will persist given systematic error no matter how great the precision of a determination, i.e., as may be achieved by increasing the total number of measurements. Nevertheless, the less biased a measurement, the more accurate the measurement, and vice versa. Number of phages, e.g., as PFUs, present per unit volume, such as per ml, but also as may be inferred as the product of the number of PFUs and dilution of the plated volume. Note that while plaque counts may be employed towards titer determination, a plaque count is not itself a (continued)

546

S. T. Abedon and T. I. Katsaounis

Table 1 (continued) Term or concept

Too Few to Count (TFTC)

Too Numerous to Count (TNTC)

Total count

Validity

Definition titer but instead is a representation of the number of PFUs present in the plated volume. Insufficient numbers of plaques or colonies are found on plates to achieve desired levels of statistical precision. Common examples of TFTC are less than 30, 40, or 50 plaques per plate. TFTC is driven by concerns with random sampling error. Note that plating over small surface areas such as towards spot counts can make it difficult to exceed TFTC without also exceeding TNTC, resulting in an inherent imprecision to spot counts versus plate counts. Spot counts as a consequence generally are considered to represent low-precision determinations of phage titers. Sufficiently large numbers of plaques or colonies per plate such that excessive plaque overlapping exists and individual plaques or colonies cannot be distinguished with sufficient likelihood in the course of counting. Common examples of TNTC are greater than 300, 400, or 500. Unlike TFTC, TNTC can be a function of plaque or colony properties with smaller, sharper-bordered entities often easier to distinguish given equivalent proximity of their points of initiation, resulting in potentially larger TNTC cutoffs. Employing excessively high TNTC cutoffs will result in excessive systematic error, i.e., biases, and therefore in inaccuracies towards undercounting of numbers of plated PFUs. Total number of potentially enumerated entities generally as determined visually (e.g., microscopically) but which does not take entity viability into account. Absolute efficiency of plating takes into account disparities between total counts and viable counts such as equivalent to plaque counts. That is, with absolute efficiencies of plating of less than one, determined titers will be lower than determined virion total counts. A measure of the degree to which a measurement conforms to an actual value.

Note that we explicitly are not assuming any probabilistic model for the number of counts per plate, e.g., Poisson distribution versus normal distribution. Therefore, this chapter is not a discussion of how, statistically, to compare plate counts as obtained under different conditions. In any case, as we will note, such statistical comparison generally should involve somewhat more than three independently obtained plate counts per calculated titer value. Furthermore, though it is not possible to count phages without error, nevertheless the larger the number of plaques counted per plate (plate count), ignoring for the moment issues of interference with plaque counting due to crowding (see “TNTC”), then the higher the expected precision of a plaque count, i.e., the smaller the coefficient of variation, where the coefficient of variation is the ratio of standard deviation to mean, and which thus shows the extent of variability in relation to the mean of the population. In the case of Poissonally distributed plate counts, the coefficient of variation is defined as the square root of a plate count divided by the actual plate count. For example, a plate

Detection of Bacteriophages: Statistical Aspects of Plaque Assay

547

count of 16 will yield a coefficient of variation of 0.25, of 25 will yield a coefficient of variation of 0.2, and a plate count of 100 will yield a coefficient of variation of 0.1, or half that for 25, and a plate count of 225 would yield a coefficient of variation of 0.067, or roughly one quarter of that seen with a plate count of 16, and so on. See Table 1 for definitions of terms relevant to such plaque-based enumeration. Summary of much of the discussion can be found in Figs. 1 and 2. See Abedon and Katsaounis (2018) for additional discussion.

TFTC and TNTC TFTC stands for Too Few To Count and TNTC stands for Too Numerous To Count. These refer to numbers of colonies found on a plate or, as is our interest here, numbers of plaques. High plaque counts are problematic due to plaque overlap on plates, which can lead to plaque undercounts: multiple plaques instead being counted as single, individual plaques. Note that this is a problem of validity (that is, is a “single” plaque really only a single plaque?) and represents a measurement bias in counts towards undercounting (fewer plaques counted than the number of PFUs originally plated). The result is incorrect titer determination. It is also a bias (another measurement bias) that increases in magnitude the larger a plate’s plaque count or, more precisely, the higher the density of plaques per unit area of lawn (by contrast, unbiased plating error, that is, unbiased estimation of the coefficient of variation, is for statistical reasons lower given higher plate counts). Problems of validity, giving rise to systematic error (bias), can be seen with TFTC as well, particularly if a proportion of counts are due to artifacts, such as inadvertently counting bubbles within agar as plaques, thereby resulting in overcounts. Low plaque counts per plate are particularly problematic, however, because the amount of error associated with lower numbers of observations is inherently greater than the amount of error associated with larger numbers (re: sampling error). TNTC. The problem of undercounting due to plaque overlap is expected to vary in magnitude as a function of plaque size, where smaller plaques may be more readily discerned from each other than larger plaques, though smaller plaques also can be more difficult to discern from bubbles or from other irregularities in lawns, particularly in comparison with very small, highly turbid plaques. It is possible to use larger Petri dishes to combat the problem of plaque undercounting due to plaques overlapping since with greater surface area the average distance between plaques, for a given number of plaques present per plate, will be greater (though the number of bubbles can be expected to increase in number as well). Hence, the cut off for TNTC can be larger given larger plates or smaller plaques, while the problem of bubbles can be minimized as a problem through improved training, both in terms of plate pouring and plaque recognition. There are limits to how large Petri dishes can be, however, in terms of material costs, handling convenience, time (given more plaques and manual counting), and even convenience (e.g., there are limits to the size of plates that can be easily counted or fitted on counting instruments or handled using automated pouring

548

S. T. Abedon and T. I. Katsaounis

Fig. 1 Flowchart of plaque-based bacteriophage enumeration, part 1. “Data valuable?” refers to the data found on a given a plate, i.e., an answer of “No” would mean that an experiment will remain reasonably valid even without including that plate’s counts whereas an answer of “Yes” means that not including that plate count would reduce the validity of the experiment. TFTC stands for “Too Few To Count” and could be equal to less than, e.g., 30, 40, or 50 depending on preference. TNTC stands for “Too Numerous To Count” and could be greater than 500 or alternatively greater than 600 or even 700 (or, depending on preference as well as plaque size, greater than 400, etc., but generally values of 300, 400, or 500 are used, depending on plaque size as typically observed for a given phage and plating conditions). Whether a plate is countable, however, is a function of the counter’s judgment or skills, thereby helping to define what TNTC value is reasonable to use, though what counts as TNTC should remain constant within experiments. Even if a plate is not countable but is still recognizably TNTC then it should still be considered towards a trimmed mean rather than outright discarded. Thus, for example, three plate counts of 350, 450, and TNTC as based on the same dilution should be described as possessing a median of 450 rather than a median of 400. (Note that if a plate count represents a sufficiently valuable datum but nevertheless only slightly exceeds TNTC then the number of plaques it contains instead can be counted and included in the calculation of a mean, though with the caveat that if TNTC was originally reasonably well

Detection of Bacteriophages: Statistical Aspects of Plaque Assay

549

devices). In other words, there is reasonable justification for employing standard-size Petri dishes for plaque-count determinations. Normally an upper limit of approximately 500 or even 700 manually counted plaques per 9-cm plate is a good rule for counting plates consisting of typical-sized phage plaques (Carlson and Miller 1994), with permissible TNTCs easily calculated for larger (or smaller) plates based on relative plate areas. Consider breaking this rule under circumstances where data is particularly valuable and if not counting plates containing marginally excessive numbers of plaques would result in an under estimation of phage titer. If counts for four otherwise identical plates, for example, are 400, 450, 500, and 550, with the “550” value deemed TNTC, then 475 = (400 + 450 + 500 + 550)/4 may still be a better estimation of the “true” average than 450 = (400 + 450 + 500)/3, as the latter would be obtained by simply dropping the 550 count. This issue of not dropping higher-count data is particularly relevant since we would expect that the 550 plaque-count plate would be rejected, on a validity basis, because it potentially represents an under estimation of number of PFUs plated, i.e., that we would be inadvertently missing plaques during enumeration as a consequence of overlap. Particularly, this would not be 550 representing an over estimation. Thus, the calculated mean of 475 more likely would be lower than the “true” average due to under counting on the 550-count plate, but of course a calculated mean of 450, obtained by declaring TNTC for the 550-count, is lower still. The broader lesson is that care – rather than, e.g., holding oneself to hard and fast rules – should be taken before one invokes claims of TNTC. Rather, the issue of TNTC is most applicable to determining which dilution from a dilution series may be most appropriately employed towards titer determination, that is, use of that dilution which generally supplies neither TNTC nor TFTC plate counts. This is preferable to using concerns over TNTC or TFTC to decide what plates from a single dilution should or should not be counted. Thus, it would be preferable, given a choice, to use for titer determination, for example, three plates of counts 67, 82, and 74 rather than 670, 820, and 740, the latter as obtained based on a tenfold lower dilution prior to plating. As a rule, therefore, TNTC should be employed to reject entire collections of data points rather than to reject individual data points, particularly individual data points as obtained from otherwise equivalent collections of data points generated based on the same degree of dilution prior to plating. It is worth noting that it is certainly permissible to use other than tenfold differences between plated dilutions, e.g., as twofold dilutions are commonly used ä Fig. 1 (continued) adopted, e.g., as based on plate and plaque sizes, then the resulting TNTC plaque count could represent a significant undercount of the original number of PFUs – such counting may be necessary in any case, however, to confirm that a plaque count indeed exceeds TNTC.) The term “dilution-plating” is used to describe the use of a single dilution series to generate no more than a single, individual enumeration (i.e., for a given dilution series, the number of platings of a given total dilution should be no greater than one – if you want more platings at a given dilution, then you should generate additional dilution series). This flowchart continues as Fig. 2

550

S. T. Abedon and T. I. Katsaounis

Fig. 2 Flowchart of plaque-based bacteriophage enumeration, part 2. “Level of precision required” is indicated either as “Lower” or “Higher” (less precision required vs. more, respectively). Higher precision is required when values are found in denominators such as when determining burst sizes (which equals final titer divided by initial titer), i.e., example C, or when comparisons between plate counts need to be made to determine statistical significance, i.e., example D. Number of “plates” refers to number of individual enumerations (see explanation of “dilution-plating” in previous figure legend along with the last sentence of this legend). While it is certainly permissible to employ a single enumeration when, for example, checking on the titer of a phage stock, or when enumerations are repeated over the course of multiple rounds of otherwise inexpensive experiments, in most cases at least three platings should be employed per titer determination. Going from left to right on the bottom: (a) We recommend solely employing medians as titer values since solely employing mean values can result in excessive loss of titer values, i.e., as due to outlier plaque counts, e.g., TFTC or especially TNTC (use of mixtures of medians and means is to be discouraged and, hence, “problematic” means that it could lead to loss of entire experiments whereas medians are less likely to be problematic, that is, they are more robust to plating errors). (b) Means here may be used instead of medians, but as above, only so long as it is permissible to lose data due to outlier data points (e.g., discarding experiments due to single problematic plate counts; again, mixing

Detection of Bacteriophages: Statistical Aspects of Plaque Assay

551

in immunology or using fivefold dilutions instead of tenfold dilutions might be employed. The cost, however, is greater amounts of plating as well as either more or more complex diluting. For example, for the latter, with tenfold diluting, for each dilution one could plate both 100 and 50 μl volumes. In the previous example, this would yield also plate counts of, for example, 335, 410, and 370. Note in any case the relative impossibility of observing this number of plaques as spot counts rather than plate counts. Nevertheless, if re-plating samples is permissible, then ballpark determinations may be made via spot counts which are then followed by more precisely tailored dilutions for subsequent plate count determinations. TFTC. At the other extreme from TNTC, too-low plaque counts introduce the problem of unacceptably high between-plate unbiased error (Carlson and Miller 1994), which is an issue of precision. Similar to too-high plaque counts, however, one should also be careful about avoiding enumerating individual plates just because counts are TFTC since that also introduces biases. In this case, ignoring a subset of plates plated from the same dilution which have been determined as TFTC would result in over estimations of titers. For example, (25 + 30 + 35 + 40)/4 = 32.5. If by setting TFTC as less than 30 and then ignoring the 25 count as TFTC, we are left with (30 + 35 + 40)/3 = 35, which of course is higher than 32.5. In other words, by dropping the 25 value as too small a value, we have ended up with a higher titer estimation than had we not dropped this value! It is still a valid argument, though, that the count of 25 will have more associated sampling error than the higher counts (relatively speaking, that is, in terms of its coefficient of variation), and therefore may be reasonably disregarded. Doing so, however, is better accomplished by employing a trimmed mean for the estimation,

ä Fig. 2 (continued) medians and means is not to be encouraged). (c) When making comparisons, use of means is crucial; though with sufficient data points (number of plates), it is possible to use trimmed means without using medians, e.g., basing means on the three center counts by discarding both the higher and lower counts when basing titer determinations on the counts derived from five plates; for example, counts of 50, 60, 73, 80, and 200 can provide a trimmed mean of 71 as equal to (60 + 73 + 80)/3 (whereas the median in this case would equal 73). (d) If experiments are valuable and/or differences between calculated titers small, then in making comparisons it can be useful to obtain even more data points (to improve precision) and also to trim means even further (to improve accuracy), e.g., with ten plate counts per titer determination, then even with the two higher and two lower plate counts excluded, i.e., so as to reasonably robustly exclude outlier data points, six plate counts would remain for mean determination. Again, it is important that one is consistent across titer determinations in how the data is handled (means vs. medians vs. trimmed means, the latter including in terms of degree of trimming). Means employing all available data provide higher precision but are subject to distortion by outlier data (i.e., inaccuracies), medians are less subject to distortion by outlier data but are not appropriate individually for making statistical comparisons (collections of medians, though, may be compared with collections of medians, i.e., comparisons of means of medians with means of medians), and non-median trimmed means can both allow higher precision and be more robust to distortion that is due to outlier points but require somewhat more data points (plaque counts) to use. In all cases, to avoid pseudo-replication, each dilution should result in only a single plating, that is, e.g., a single dilution series should give rise to the recording of no more than one enumeration

552

S. T. Abedon and T. I. Katsaounis

such as the median, which in this case happens to also be 32.5, i.e., (30 + 35)/2. As with counts which are too numerous, those which are too few thus should not be simply dropped, since that would introduce biases. These counts nevertheless may legitimately not have their actual values contributing to estimations by limiting calculations to median determinations. Median determinations, that is, inherently exclude all extreme low or high values from calculations of estimations while at the same time not discounting the actual existence of these same values, i.e., the precision-enhancing consequence of obtaining greater values of n still applies in median calculations as it does in mean calculations. One also can make up for too-low plaque counts by simply doing more platings per data point (Carlson and Miller 1994; see, though, our serial dilution discussion, below). Five plates with an average of 25 plaques per plate, in other words, is at least as valid (indeed, more so) for estimating phage titers as one plate with 30 plaques per plate. Note, however, that simply using more plates to combat issues of TFTC is effective only if one can avoid counting artifacts as plaques such as bubbles – a validation issue – since such artifacts tend to be more or less constant in number per plate and therefore would disproportionately inflate lower versus higher plate counts (thus introducing a bias). Note that one also can minimize bubble formation by not vortexing top agar mixtures prior to pouring but instead by relying on gentle sliding of poured plates in circles to effect mixing (Carlson and Miller 1994). Alternatively, it is possible to vortex at lower speeds and otherwise mark bubble locations prior to plate incubation. Range of TFTC to TNTC. Though minimum plaque counts of 50 per plate are preferable, it is fairly standard instead to use 30 as a minimum count. People tend to employ a range of tenfold over which they would prefer their plaque counts, however, such as from 30 to 300, resulting in TFTC essentially defining TNTC. In reality the minimum number of plaques per plate that are found to be acceptable should not dictate the maximum number. In particular, “order of magnitude,” meaning tenfold differences, is an arbitrary measure, something that we use for convenience because of our use of a base-10 counting system. Thus, consideration of a range of plaque counts from 30 to potentially as high as 700 is not at all unreasonable, and particularly so if one aims in the course of plating to achieve plaque counts that actually range from 50 to 500. As for TNTC, for TFTC the far bigger issue is to determine which dilution provides more acceptable data for titer determination, with, for example, counts of 230, 250, and 270 preferable to ones of 23, 25, and 27, the latter as would be obtained given tenfold greater dilution prior to plating (and assuming here that the higher counts are below permissible TNTC, which may not be true in the case of very large plaques but typically will be true; note though that if plaques in fact are too large to allow for reasonably high TNTCs, then consider incubating for shorter periods, at lower temperatures, or using higher percentages of agar in one’s solid or semi-solid media). TFTC thus need not be determined by TNTC, and vice versa. Instead, the lower number should be indicated by one’s willingness to disregard whole magnitudes of dilutions, i.e., versus disregarding individual plate counts, with higher TFTCs resulting in higher measurement precision but more loss of data. The higher number

Detection of Bacteriophages: Statistical Aspects of Plaque Assay

553

(TNTC), by contrast, should be determined by how readily individual plaques may be distinguished, with small sharp-bordered plaques allowing for higher TNTCs and large, diffuse plaques requiring lower TNTCs. In addition, as noted, use of larger Petri dishes should allow for higher TNTCs, which increases essentially proportionally to the agar surface area within the larger Petri dish. By contrast, use of spotting for phage titering has the effect of decreasing the agar surface area within which plaque counts are determined.

Spot Counts Versus Plate Counts Contrasting the use of larger plates in order to increase TNTC is the use of spotting. With spotting, a smaller plating area is employed to generate plaque counts than the total area of a plate, often much smaller, e.g., tenfold or greater decreases in total area. Just as increases in plating area will have the effect of proportionally increasing permissible TNTCs, so too will decreases in plating area proportionally decrease permissible TNTCs. As a result, with spotting-based plaquing, as giving rise to “spot counts,” TNTCs will inherently be much smaller than when determining plate counts. Indeed, the change should be readily calculable given determinations of spot diameters versus plate diameters, with permissible spot TNTCs equal to spot areas relative to plate areas, fraction multiplied by permissible TNTCs for the larger plates. A similar effect will be seen when employing plates which are smaller than standard or instead, e.g., when using wells of six-well plates which similarly are smaller than typical Petri dishes. This issue of TNTCs being more easily exceeded means that the range of plaques over which valid counts may be obtained (i.e., less than TNTC) will be smaller. Indirectly this can result in decreases in plaque count precision, as considered in the following paragraph. TFTC once decided upon should be identical between spot counts and plaque counts. This is because TFTC is a statistical construct rather than one which stems from plating area. This inflexibility of what plaque counts constitute TFTC creates an imprecision with spot counts in comparison to plate counts, the latter as based on larger total plaquing areas. That imprecision stems from the noted reductions in the useful range of plaque counts that can be obtained with spotting as due to inherently lower TNTC values. Thus, the range of plaque counts between TFTC and TNTC could very well easily be an order of magnitude for plate counts, or more, such as from 30 to 500, whereas for spotting this number may instead be, e.g., 10, that is, as ranging from 30 to 40. The result is that tighter dilution series may need to be employed to readily obtain sufficiently precise spot counts, e.g., a series of twofold rather than tenfold dilutions, but in doing so one would reduce the utility of employing spotting-based titering as plating greater numbers of dilutions requires the use of greater numbers of Petri dishes. As such approaches to diluting phages are rare, instead TFTC counts will tend to be routinely employed towards spot count-based titer determinations despite their TFTC status. The result is that titer determinations based on spot counts essentially are unacceptably imprecise except as approximations of true titers, i.e., what Carlson

554

S. T. Abedon and T. I. Katsaounis

and Miller (1994) describe as “semiquantitative.” Where titering precision is required, such as when calculating phage titers for use experimentally, or indeed to collect data during experiments, then spotting-based titers should therefore be used only as preliminary titers, to be followed with plate count-based titers. For protocols on this “drop plaque method” of plaque generation, see Carlson and Miller (1994), Carlson (2005), Mazzocco et al. (2009b), and Letarov and Kulikov (2018), as well as Kutter (2009).

Number of Repeats It is self-evident that one data point, such as one plate-count determination, is preferable to none. Furthermore, because one plate-count determination can always be in error, two determinations are effectively always preferable to one. Three platecount determinations, however, can be substantially preferable even to two. This is particularly so since with three plate counts one can more readily identify outlier values. The jump from three to four plate counts, though, at best provides only a proportionately smaller increase in precision. This is something to consider particularly when taking into account time or materials invested. For instance, obtaining two of what should be identical plate counts, such as 300 versus 55, is quite different from obtaining three plate counts of, for example, 55, 60, and 300, which in turn is different from four counts of 55, 60, 65, and 300, but less so. In the first instance (two plate counts) it may be impossible to tell which of the two data points is in error. In the second instance (three plate counts) there certainly is an argument that can be made that the “300” data point is an outlier. In the third instance (four plate counts) that argument would be even stronger, though not qualitatively stronger. A related issue is the question of just what more than one plate-count determination does or does not mean. In a time-course, two single plate-count determinations taken in quick succession that are similar in magnitude, assuming that this similarity is expected, should for example provide essentially equivalent information as two data points taken simultaneously. An example of this can be seen when taking data points for single-step growth experiments (Hyman and Abedon 2009), where data points taken prior to the end of the eclipse should all have the same basis and so, similarly, should all data points taken following the rise, that is, past the point of population-wide completion of lysis. Thus, for example, it would be desirable to obtain at least three plate counts during single-step growth experiments prior to the end of the eclipse and three plate counts after the end of the rise for phage burst-size determinations. Since the timing of the end of the eclipse and the end of the rise are not necessarily precisely known in the course of taking these data points, the obtaining of even greater numbers of plate counts can be desirable to achieve a minimum of three counts that can be averaged together (i.e., rather than used to generate a median), with one set taken prior to and other following the rise (see ▶ “Phage Infection and Lysis” chapter). Indeed, a theme of plate count determinations is that there always exist a significant time lag between the initiation of plating and enumeration of plates,

Detection of Bacteriophages: Statistical Aspects of Plaque Assay

555

thus providing a good reason for obtaining more data points than may be minimally necessary. As a consequence, the utility of taking three replica data points can be somewhat context dependent. In particular, if an experiment is extremely expensive or otherwise difficult to perform, then only a limited number of repeats of the entire experiment may be possible. In this case, having an ability to distinguish between outlier data points and non-outlier data points may be particularly helpful, and this is far more easily accomplished given three or more plate counts per data point versus only two, with the implicit assumption that obtaining plate counts is a comparatively inexpensive aspect of such an experiment. In contrast, if experiments are inexpensive, fast, and not difficult then there can be no reason to bother doing repeats of individual plate-count determinations. Instead, one just as well could repeat the entire experiment. For instance, if a single experiment consisted of four or more independent titer determinations, such as in the course of phage adsorption rate determination (Hyman and Abedon 2009), then n for each titer determination could very well be equal to just one, with repetition obtained instead by repeating the experiment in full. Alternatively, as considered in the previous paragraph, if experiments are expensive and/or difficult – such as is the case given phage therapy experimentation using animals – then it can pay to increase one’s confidence in individual data points by doing multiple, independent titer determinations.

Utility of Trimmed Means Elimination of outlier data points without invoking simply one’s opinion can be easily accomplished by employing the median to estimate the “true” average value rather than the mean, though this utility may be achieved only when medians are based on three or more data points. A median is the “middle” observation, or more specifically serves as a trimmed mean (with what can be described as a 50% trimming of low data values in combination with a 50% trimming of high data values). Indeed, a median provides the ultimate trimmed mean since it is based on an absolute minimum of not-trimmed data points. Medians, as well as trimmed means more generally, as a consequence are robust to outliers since what is “trimmed” in a trimmed mean explicitly are outlier values. For example, the counts of 55, 60, and 300 would yield a median estimate of the “true” average of PFUs of 60. For counts of 55, 60, 65, and 300, the median instead would be 62.5, which can be viewed as a more reliable median as it is based on four data points rather than three, and therefore is a more precise estimation of the true value. There nevertheless is not a great deal of difference, in this example, between an estimation of 60 and 62.5. This therefore provides an illustration of why going from obtaining three data points to four does not necessarily result in substantial improvement in precision, and particularly so long as the median is employed to arrive at this number rather than the mean. The comparable untrimmed means in the above examples are 138 and 120, where 120 would be expected to be closer to the true value (~60) than 138, though neither value is even close to the 60 or 62.5 arrived at based on medians rather than means.

556

S. T. Abedon and T. I. Katsaounis

Given plate count values of 300 and 55, however, the median as well as the mean would be 177.5. Since in the case of two data points the median and mean are equal, what has been changed in going from n = 3 to n = 2 essentially is loss of the utility of employing the median rather than the mean for this estimation. Note further that, with this example, 177.5 is even farther than 138 or 120 from 60 or 62.5. In very simple terms, then, this is why obtaining three data points is much preferable to two, though nevertheless obtaining two data points will tend to improve upon determinations that are based on only a single data point by providing a higher potential of at least identifying that error may be present. That is, with a plate-count data set consisting of 55, 60, and 300, it is fairly likely that 300 is erroneous. With a plate count data set consisting of only 55 and 300 it is still possible to recognize that at least one value may be in error, but without additional information it is not possible to identify which one. Obtaining four data points is not necessarily much of an improvement on three in this regard because, again in this example, we would still tend to identify 300 as erroneous, though with slightly greater certainty. In either instance, three versus four data points, in employing the trimmed mean in this example the “300” value data point would be dropped from the titer determination. In any case, for the sake of precision it is always preferable to obtain more data points. Nevertheless, limitations in resources can place some limit on the number of points typically sought. Our argument is simply that attempting to obtain three plate counts per titer determination provides a sufficiently substantial improvement in precision over two plate counts that, for valuable data, that third data point is well worth obtaining. An online calculator for titer determination using means, a trimmed mean, and median can be found at titering.phage.org.

Number of Dilution Series To limit the number of plaques that will form per plate, one employs dilutions and typically serial dilutions. Discussion of serial dilutions can be found in most introductory microbiology textbooks. Here we consider statistical concerns when employing serial dilutions. In a dilution series you are generating error every time you remove one volume and add that volume to another. The greater the number of steps in the dilution series (that is, number of volume removals and mixings), and the more difficult the individual steps (i.e., higher likelihood of generating operator error), then the greater the error that will tend to be generated overall. Ideally that error is not systematic error, e.g., such as consisting of biases towards under measuring or over measuring of volumes, so therefore to a degree errors can balance. Nonetheless, each time one performs a dilution series one is introducing at least some error into a measurement and, more importantly, per measurement one cannot easily know how much error has been introduced. Particularly as a consequence of this potential for introduction of error, it is crucial to make sure that independent determinations really are independent determinations. Thus, for example, two experiments done on the same day in the same

Detection of Bacteriophages: Statistical Aspects of Plaque Assay

557

laboratory are not necessarily as independent as two experiments done on different days or in different laboratories. So too, two plate-count determinations that are drawn from the same dilution series are not nearly as independent as two data point determinations that are drawn from different dilution series. When we make individual titer determinations, it is very obvious that, as noted, two measures are preferable to one and three to two, etc. This statement, though, does not address the need for statistical independence of determinations. Another way of saying this is that multiple data point determinations, here multiple platings, should improve the precision of a measurement, but not necessarily prevent biases in measurement, that is, not necessarily improve accuracy. This is because if an error occurred prior to multiple plating steps, all of which were derived from the same potentially faulty dilution series, then that error, that is, the proposed pre-plating bias, would be common to all resulting plating steps. It is especially during dilution steps that such error may be introduced during titer determinations. Thus, if an error is made while diluting, but diluting is done only once per titer determination, then no matter how many platings that titer is based upon, there will be no way of knowing that in fact a dilution error has occurred. In other words, with multiple platings from a single dilution series the precision of counts can be high while your confidence in the accuracy of those counts, that is, in terms of the impact of biases stemming from the dilution series itself, will be low. Just as with individual data points, these issues are of less of a concern if experiments are inexpensive and easy to perform since in this case each data point can be based on only a single dilution series (and, to reiterate, less is gained from performing multiple platings from that individual dilution series vs. plating from multiple, individual dilution series). If experiments are expensive, however, then it behooves the researcher to strive for as much independence among data points as can be relatively conveniently achieved. Thus, if more than one plating is indicated per time point in an experiment, then so too would more than one dilution series be warranted, that is, one dilution series per plate-count determination (though multiple platings at multiple dilutions is certainly a reasonable practice per individual dilution series, just not re-plating the same dilution from the same dilution series multiple times unless absolutely necessary). Improvement in accuracy similarly might also be achieved by designing experiments so that the magnitude of the dilution series – e.g., tenfold versus 10,000-fold – in fact is reduced. At a minimum, a researcher should be aware of when potentially costly or misleading shortcuts are being taken in terms of experimental design, and a lack of independence of individual titer determinations unquestionably is one such short cut.

Utility of Larger Volumes While serial dilutions can be used to reduce phage titers prior to plating, so as to avoid TNTC, ultimately some phage-containing volume needs to be plated to produce plaques. Statistically it is preferable to add a larger volume from a given dilution versus a smaller one. In non-technical terms, this is because the error

558

S. T. Abedon and T. I. Katsaounis

associated with a larger volume of PFUs is expected to be smaller than for a smaller plated volume, all else held constant. If we consider a given volume to consist of one or more volume “units,” then a larger volume will consist in effect of more volume units than a smaller volume. In statistics, we generally have higher confidence in estimations where more measurements (n) are employed to generate an estimation. That is, since n is found in the denominator of the calculation of variance of the distribution about a mean, the result is an expectation of lower variance, and thus greater precision, when n is larger, as in effect can be achieved by plating larger numbers of volume units and therefore larger volumes. Larger volumes also typically can be handled with lower relative error than smaller volumes. Thus, plating 100 μl will tend to be truer to 100 μl than 10 μl will tend to be true to 10 μl. The former, e.g., as may be typically employed for plate count determinations and the latter, e.g., as may be typically employed for spot count determinations. As a caveat, however, note that adding an additional dilution step in order to plate larger volumes introduces an additional error-generating step, the dilution, thus making that specific practice towards plating larger volumes less worthwhile. It is plating larger volumes from a given dilution that should contribute to greater measurement precision, rather than solely plating larger volumes. Thus, plating larger volumes, or diluting less before plating, ultimately should provide similar if not necessarily identical benefits. Notwithstanding the potential for plating larger volumes to improve the precision associated with plate counts, with larger volumes one cannot similarly claim statistical independence among the samples, with samples in this case being the volume units that are plated (e.g., 1000 μl contains twice as many volume units as 500 μl). Accuracy therefore cannot be assumed to be improved by plating greater volumes. For example, if some prior error resulted in counts being off by, say, tenfold prior to plating, then they would still be off by tenfold no matter the volume plated. What will be improved by plating larger volumes, as noted, is precision, which is another way of stating that the error as plated will be smaller, though any error introduced prior to plating will not be equivalently smaller (i.e., as regarding the second half of the previous paragraph). The latter by contrast may be identified only given multiple, independent platings: the greater the independence of these platings then the greater the potential to identify errors. What these ideas mean in practice is that platings that result in more plaques on a given plate will tend to display lower error, relative to the actual plate count (coefficient of variation), than platings that produce fewer plaques, all else held constant, and this in fact represents the statistical underpinnings, non-technically stated, of the concept of TFTC. For a given volume plated, on the order of 20 plaques produced, for example, will likely be a poorer estimation than, for example, tenfold higher volumes plated that thereby produce on the order of 200 plaques. If you want to make sure that titers are accurate, however, then you will need to strive towards multiple, independent measurements, at least three, and – given circumstances that are warranting – then ideally even more. In addition, keep in mind that all of these considerations are ones that address the determination of plaque counts rather than serving to define how statistics should be employed in the comparison of results obtained either within or between experiments.

Detection of Bacteriophages: Statistical Aspects of Plaque Assay

559

When Comparisons Matter There is a difference between obtaining a useful data point and comparing useful data points. In either case, statistical independence of data points as well as the employment of larger versus smaller volumes or at least greater numbers of plaques per plate will continue to be important. For the sake of making comparisons between data points, however, the size of n becomes much more important. Specifically, at the point where the goal is to identify statistically significant differences between means, particularly as obtained from within individual experiments, then the argument of obtaining at least three platings per data point should be replaced with obtaining instead at least five platings and ideally somewhat more, so that in addition an estimate of the variability of the data values for each set of values can be assessed. Specifically, the smaller that n is then the larger that differences between those means must be before differences may be determined to be statistically significant. In addition, employing medians is not recommended but instead the full distribution of values and their associated arithmetic means, that is, untrimmed means, typically will be used. Thus, when determining a titer, particularly for the first time, good technique will warrant at least three replications – of plate counts rather than spot counts – that are somewhat independent (i.e., different dilution series) and the value itself can reasonably be expressed as a median rather than mean, and without any indication of variability. The more important that value, however, including in terms of making comparisons within experiments or in reducing variability between experiments, then employing somewhat greater values of n can be imperative, e.g., five independent platings or more per titer determination. Note that trimmed means may still be used, for the sake of excluding outlier values, in which case the suggestion would be to base the trimmed mean calculation on at least five data points rather than starting with five data points and then trimming.

Conclusions In this chapter we considered the use of plaquing as a means of phage enumeration, focusing on the data obtained by plaquing experiments and associated statistics. We suggest that use of TNTC and TFTC in fact may be abused by being too dogmatic in their application. We also provide a reminder that spotting-based titer determinations, i.e., drop plaque method, should be viewed as inherently imprecise, i.e., “semiquantitative,” in comparison to titer determinations based on plate counts. We suggest further that medians be employed to generate individual titer determinations based upon multiple plate counts and, related, that at least three plate counts be obtained per titer determination when data are “important” (or more if statistical comparisons between means are to be made). Relevant to define as well is the cost of doing experiments, where easily as well as cheaply run experiments will not necessarily require similar number of plate-count repeats in generating individual titers since whole experiments instead may be repeated numerous times with relative ease. We suggest, regardless, for both an individual datum as well as experiments generally, that efforts be made to maintain independence, such as by employing a

560

S. T. Abedon and T. I. Katsaounis

single dilution series to generate, per specific, individual dilutions, only a single plate count, i.e., such that at least three dilution series should be employed to generate a single titer determination as obtained via median calculation and, as noted, for the sake of obtaining acceptable precision, employ plate counts rather than spot counts. We suggest further that plating larger volumes (or fewer dilutions), such that more plaques are found per plate, can be preferable for reasons that essentially are equivalent to the problem of avoiding TFTC. We additionally draw a distinction between the generation of a single titer determination, in terms of number of suggested plate-count repeats, on the one hand, and results which are to be compared statistically, on the other, where somewhat more independent repeats (five or more) ideally will be performed in the case of the latter and particularly so given observed differences which are relatively small. Lastly, plaque counts are not alone in requiring equivalent handling, e.g., qPCR measurements of phage titers (Anderson et al. 2011) too should be made at least in triplicate or more if they are to be statistically compared.

Cross-References ▶ Phage Infection and Lysis ▶ Detection of Bacteriophages: Phage Plaques ▶ Isolation of Bacteriophages

References Abedon ST, Katsaounis TI (2018) Basic phage mathematics. Methods Mol Biol 1681:3–30 Anderson B, Rashid MH, Carter C, Pasternack G, Rajanna C, Revazishvili T, Dean T, Senecal A, Sulakvelidze A (2011) Enumeration of bacteriophage particles: comparative analysis of the traditional plaque assay and real-time QPCR- and nanosight-based assays. Bacteriophage 1:86–93 Carlson K (2005) Working with bacteriophages: common techniques and methodological approaches. In: Kutter E, Sulakvelidze A (eds) Bacteriophages: biology and application. CRC Press, Boca Raton, pp 437–494 Carlson K, Miller ES (1994) Enumerating phage: the plaque assay. In: Karam JD (ed) Molecular biology of bacteriophage T4. ASM Press, Washington, DC, pp 427–429 DeGroot MH (1975) Probability and statistics. Addison-Wesley, Boston Hyman P, Abedon ST (2009) Practical methods for determining phage growth parameters. Methods Mol Biol 501:175–202 Kropinski AM, Mazzocco A, Waddell TE, Lingohr E, Johnson RP (2009) Enumeration of bacteriophages by double agar overlay plaque assay. Methods Mol Biol 501:69–76 Kutter E (2009) Phage host range and efficiency of plating. Methods Mol Biol 501:141–149 Letarov AV, Kulikov EE (2018) Determination of the bacteriophage host range: culture-based approach. Methods Mol Biol 1693:75–84 Mazzocco A, Waddell TE, Lingohr E, Johnson RP (2009a) Enumeration of bacteriophages by the direct plating plaque assay. Methods Mol Biol 501:77–80 Mazzocco A, Waddell TE, Lingohr E, Johnson RP (2009b) Enumeration of bacteriophages using the small drop plaque assay system. Methods Mol Biol 501:81–85 Moore DS, Notz WI (2014) Statistics, concepts and controversies. W.H. Freeman Co., New York

Detection of Bacteriophages: Electron Microscopy and Visualization David M. Belnap

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electron Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Initial Electron Microscopy of Phages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Staining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metal Shadowing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Negative Staining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thin-Sectioning, Room-Temperature and Cryogenic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Frozen-Hydrated Specimens, Unstained and Negatively Stained . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scanning Electron and Helium-Ion Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scanning Transmission Electron Microscopy, Dark-Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . In Situ, Liquid-Cell Electron Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Analytical Electron Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Two-Dimensional Image Averaging and Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Three Dimensions from Two . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Immuno-EM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

562 563 565 565 571 573 576 580 584 584 585 588 588 592 599 601 611

Abstract

Electron microscopy (EM) is an information-rich, aesthetically satisfying methodology. EM has given us tremendous structural and functional insights into the fascinating world of phages. Bacteriophages were one of the first EM specimens, and phages and EM have enjoyed a warm relationship ever since. Thousands of EM-phage studies have been published. Specimen preparation techniques include

D. M. Belnap (*) School of Biological Sciences and Department of Biochemistry, University of Utah, Salt Lake City, UT, USA e-mail: [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_18

561

562

D. M. Belnap

both staining and nonstaining methods. Care must be taken during sample preparation, and drying of specimens typically results in artifacts. Because of its ease of use and information richness, negative staining is an especially helpful, and the most commonly used, technique. Cryogenic, nonstaining methods are the best way to preserve native structure, but also require more effort. Thin-sectioning methods are useful techniques for phage-host studies. Surfacerendering EM methods are used to image isolated phages, phage-host interactions, and isolated phage DNA. Immuno-labeling allows specific phage components to be located. Computer image processing enables vast improvements in resolution and comprehension through two-dimensional averaging and three-dimensional reconstruction, including through tomography.

Introduction Electron microscopy has given cellular and molecular biology a vast amount of fruitful and invaluable information (see ▶ “Structure and Function of Bacteriophages” chapter). Because phages are typically smaller than the wavelengths of light used in light microscopy and are hence invisible, phage studies are a particular beneficiary of the much smaller wavelengths used in EM. Some spherical and filamentous phages, and some phage components, have been visualized by X-ray diffraction or nuclear magnetic resonance (NMR). However, the bizarre, fascinating, and diverse shapes of phages would likely be completely unknown without EM, even for well-known phages such as the caudoviruses (Fig. 1). X-ray and NMR experiments would likely be more difficult to perform without insights learned from EM. We have gained many functional insights via EM as well. Our knowledge of phages and other viruses would therefore be much poorer without EM. The late Hans-Wolfgang Ackermann noted the tight link between EM and bacteriophages (Ackermann 2012). EM results proved the particulate and viral nature of phages and have aided and continue to aid in understanding their shapes, sizes, components, complexity, variability, diversity, replication, assembly, disassembly, classification, ecology, and evolution. In 2012, he and David Prangishvili noted that 6,284 phages had been examined by EM since 1959; the rate of new bacterial or archaeal virus descriptions was relatively constant at about 100 per year for the previous 20 years (Ackermann 2007; Ackermann and Prangishvili 2012). Assuming that trend has continued, we have EM descriptions now of about 7,000 phages. The body of EM-phage work is immense. As with any other technique, EM work complements and enhances the knowledge gained from other techniques. EM studies are not by themselves the pinnacle or end-all. As with data from any other method, a good EM study confirms results from other experiments and proposes additional questions and hypotheses. However, visualizing objects that have been inferred by other techniques makes EM results particularly joyful. The aesthetic component of EM data is a wonderful and especially rewarding part of EM work.

Detection of Bacteriophages: Electron Microscopy and Visualization

563

Fig. 1 Electron microscopy of three well-studied families in the order Caudovirales: Myoviridae, Podoviridae, and Siphoviridae. These images are of negatively stained specimens – the most common EM technique used in phage studies. Left, a podovirus, bacteriophage ϕ29, between two myoviruses, bacteriophage T2 (Anderson et al. 1966; Kellenberger and Edgar 1971). Right, an example of a siphovirus, bacteriophage Basilisk (Grose et al. 2014). The T2, ϕ29 picture is affectionately called “Family Portrait.” Both T2 and ϕ29 have prolate heads. Note the fibers extending from the T2 tails and serrations in T2 tail sheaths. In ϕ29, a collar, appendages, and an extended tail can be seen. Each Basilisk phage has an icosahedral head and flexible tail. Needle-like structures are seen at the flexible tail tips. Scale bars, 100 nm. (The T2, ϕ29 picture is reprinted with permission from Kellenberger and Edgar (1971), copyright 1971, Cold Spring Harbor Laboratory Press.)

Electron Microscopy Detailed explanations and summaries of electron microscopy are plentiful (e.g., Belnap 2015; Bozzola and Russell 1999a; Castón 2013; Hawkes and Valdrè 1990). To be brief, electron microscopy is a method of observing individual phage particles or other objects. This is different from analysis techniques, such as ultracentrifugation, spectroscopy, calorimetry, or X-ray crystallography, that measure bulk or average properties (see ▶ “Detection of Bacteriophages: Phage Plaques” chapter). Scanning and transmission are the two main types of electron microscopy. In both techniques, an electron beam is aimed at a specimen. Transmission EM (TEM) images come from electrons passing through the specimen as a large diameter beam illuminates an area of interest on the specimen. Scanning EM (SEM) images come from electrons reflected from the specimen as a small diameter beam is scanned across the area of interest. Transmission EM is the predominant form of EM used to study phages. Sometimes the two methods are combined as scanning transmission EM (STEM), and a small beam is scanned across the specimen, but images are formed from the electrons that pass through.

564

D. M. Belnap

Fig. 2 A comparison of transmission light (left) and electron (right) microscopes. Here, light rays or electron beams pass through a specimen and are used to form an image. Light and electron microscopes have similarly functioning components: an illumination source creates a light or electron beam, condenser lenses focus the beam on the specimen stage, the stage holds and moves the specimen, an objective lens forms the image, a projector lens magnifies the image, and a detector allows human visualisation. Each lens is depicted here as a single lens, but in most cases is a system of lenses. Ideally, the condenser lens system produces parallel illumination on the specimen. This gives the highest resolution

Transmission electron microscopes have similar functional components to light microscopes (Fig. 2). But, the two types of microscopes differ in several important ways: • EM lenses are coils of wire, not light-refracting material. Each coil induces an electromagnetic field, which focuses the electron beam passing through a vacuum in the center of the lens. • EM lenses are imperfect. Hence, EM resolutions are far from the theoretical ideal, which essentially has been reached for light microscopy. EM lenses have significant aberrations, some of which can be corrected by tuning the instrument or mathematically correcting the image. An electron microscopist usually spends significant time optimizing the beam and focus settings by tuning lens correctors and beam deflectors. Computer image processing can be used to correct images. • The high-energy electron beam is a blessing and a curse. The wavelength is much shorter than visible light. The shorter wavelength allows smaller objects to be resolved, but the high energy is destructive to fragile biological objects.

Detection of Bacteriophages: Electron Microscopy and Visualization

565

• The inner column of the electron microscope is maintained as a vacuum (pressures of 10–5 Pa or lower). The electron beam cannot be controlled sufficiently otherwise and would be degraded by interactions with gas molecules. • Compared to the colors observed because of light absorption, reflection, or fluorescence, colorless electron images primarily report relative density or mass thickness of imaged objects. In other words, intensity in an electron image corresponds to thickness or atomic mass. For example, a 50-nm-diameter gold particle is more intense than a 50-nm-diameter phage head. An 80-nm-diameter phage head will be easier to see than one that is 50 nm in diameter. • Because of the high vacuum required for passage of the electron beam, nearly all EM specimens to date are solid. This means biological specimens are usually not viewed in a native liquid-water environment. Sample preparation is the most critical part of EM. Not only must phage or other samples be carefully obtained, but that normally aqueous sample must be turned into a solid. Therefore, phage specimens are dried and stained, embedded in plastic resin, or frozen in a thin film. Despite these treatments, images of these specimens have given us an incredible amount of knowledge about the phage world.

Initial Electron Microscopy of Phages The first electron micrographs of bacteriophages were published in 1940 (Ackermann 2011a, 2012; Kruger et al. 2000). Two phage studies were published in the same issue of Naturwissenschaften. These micrographs were not the first published EM images of viruses, but they appeared shortly after the first pictures of non-phage viruses were published. Helmut Ruska presented pictures of Escherichia coli and bacteriophage (Ackermann 2011c; Ruska 1940), and Pfankuch and Kausche (Ackermann 2011b; Pfankuch and Kausche 1940) published pictures of isolated bacteriophage (Fig. 3). Early EM work was sometimes aided by osmium staining (Kruger et al. 2000), but the development of robust, easy-to-use staining techniques enabled electron microscopy to revolutionize phage biology (Ackermann 2011a). Since the 1950s, EM techniques have been thoroughly documented in numerous books and journal articles (Harris 2015). More than 20 different EM methods have been used to prepare and image phages (Table 1).

Staining Unstained biological particles are difficult to see by EM. The densities of phage proteins and nucleic acid are only marginally different from that of the surrounding specimen support (e.g., carbon film). Therefore, increasing the density of phage

566

D. M. Belnap

Fig. 3 Three of the first published electron micrographs of bacteriophage (Ackermann 2011a; Kruger et al. 2000). Top left, coliphages on sharply delimited bacterial surface (Ackermann 2011c; Ruska 1940). Top right, highly purified suspension of coliphages (Ackermann 2011b; Pfankuch and Kausche 1940). These two images were parts of the first two simultaneous publications of bacteriophage EM in 1940. Bottom, 2 years later, this image of E. coli in suspension of bacteriophage was published (Luria and Anderson 1942). The head and tail structure of the phages is clearly seen. (Top figures reprinted by permission from Springer Nature GmbH (Pfankuch and Kausche 1940; Ruska 1940), copyright 1940. Bottom figure, courtesy Proceedings of the National Academy of Sciences of the United States of America.)

particles, or increasing the density of the volume around a phage particle, makes phage easier to see. Stains are metal coatings or metal salts and are used to enhance contrast. When stain is used, the dominant contribution to an image comes from electrons scattered from the metal stain, not from the specimen. In addition, staining mitigates damage from the electron beam. Metals are, of course, more robust than the organic molecules of biological specimens. As can easily be observed, unstained specimens usually are destroyed by the electron beam, sometimes within seconds (e.g., Thomas et al. 2008; Wu et al. 2012). Stained specimens usually last much longer. Three types of metal staining are typically used, metal shadowing (Fig. 4a), negative staining (Fig. 4b), and positive staining (used in thin-sectioning work).

Detection of Bacteriophages: Electron Microscopy and Visualization

567

Table 1 Electron microscopy methods used in phage studies, ordered approximately as given in the text

Method Metal shadowing, unidirectional

Metal shadowing, rotary

Phage Isolated and particles host Prosa x x Enhanced contrast 3D appearance Robust specimen

x

Metal shadowing, freeze fracture

Negative staining

x

x

x

x

Thin-sectioning, room-temperature

x

Thin-sectioning, cryogenic, freezesubstitution

x

Thin-sectioning, Tokuyasu

x

Thin-sectioning, cryogenic

x

Enhanced contrast May have 3D appearance Robust specimen Characterization of DNA and DNA-protein complexes Enhanced contrast May have 3D appearance Robust specimen Interaction of phage with membranes See inner membrane features Rapid preparation High contrast Detailed images Robust specimen

Consa Lengthy preparation Dehydrated specimen Metal coating on specimen Potential sample distortion Limited resolution Only exposed surfaces seen Lengthy preparation Dehydrated specimen Metal coating on specimen Potential sample distortion Limited resolution Only exposed surfaces seen Lengthy and delicate preparation Metal coating on specimen Potential sample distortion Limited resolution Only exposed surfaces seen

Dehydrated specimen Potential sample distortion Metal salt surrounds specimen Limited resolution Positive staining may obscure features Only stain-accessible surfaces seen Intracellular views Lengthy preparation Robust specimen Dehydrated specimen High contrast Potential sample distortion Limited resolution Intracellular views Lengthy preparation Structure better preserved Dehydrated specimen Robust specimen Potential sample distortion High contrast Limited resolution Intracellular views Lengthy preparation Structure better preserved Dehydrated specimen Robust specimen Potential sample distortion High contrast Intracellular views Lengthy preparation Hydrated specimen Difficult preparation Native or near-native Potential sample distortion structure Maintaining cold temperature Full structure (inside and Radiation damage (low dose out) imaging required) Unstained specimen Fragile specimen Low-contrast images Noisy images (continued)

568

D. M. Belnap

Table 1 (continued)

Method Frozen-hydrated, unstained (cryogenic EM or cryo-EM)

Phage Isolated and particles host Prosa x x Hydrated specimen Unstained specimen Native or near-native structure Full structure (inside and out)

Frozen-hydrated, negatively stained

x

x

Scanning EM

x

x

Scanning He-ion microscopy

x

x

Scanning transmission EM, dark-field

x

x

Consa Lengthy preparation Difficult preparation Maintaining cold temperature Radiation damage (low dose imaging required) Beam-induced specimen movement Air-water interface may denature macromolecule Fragile specimen Low-contrast images Noisy images Hydrated specimen Lengthy preparation Approaches native Difficult preparation structure Maintaining cold temperature Higher contrast than Heavy metal salt in specimen unstained specimen Radiation damage (low dose Reduced beam sensitivity imaging should be used) than unstained specimen Fragile specimen Air-water interface may denature macromolecule High contrast Lengthy preparation Large specimens can be Dehydrated specimen viewed Metal coating on specimen Appearance of 3D Potential sample distortion Robust specimen Limited resolution Observe surface features Only exposed surfaces seen High contrast Dehydrated specimen Higher resolution than Potential sample distortion scanning EM Limited resolution Unstained specimen Only exposed surfaces seen (conductive coating not needed) Large specimens can be viewed Appearance of 3D Observe surface features High contrast Dehydrated specimen Molecular weight data Potential sample distortion Unstained specimen Overlapping macromolecules Can relate mass to spatial may not be recognizable position (continued)

Detection of Bacteriophages: Electron Microscopy and Visualization

569

Table 1 (continued)

Method In situ, liquid cell

Analytical EM

Two-dimensional image averaging

Phage Isolated and particles host Prosa x x Hydrated specimen Native or near-native structure Imaging of dynamic processes Can change environment to see effects x x Chemical information relative to spatial position Unstained specimens Enhance contrast of elements in stained or unstained specimens x Improved resolution Improved signal-to-noise ratio Can separate conformational variants

Three-dimensional x imaging, stereo Three-dimensional x image reconstruction, “single-particle analysis”

x

Consa Limited resolution Radiation damage (low dose imaging required) Fragile specimen Specimen movement Low-contrast images Noisy images Specimen may need to be thin (for electron energy loss spectroscopy) Dehydrated specimen (unless liquid-cell or cryogenic specimen)

Need multiple identical images of object Features smeared if input images are poorly aligned Computational resources needed Use any specimen Not everyone can see stereo People who can may still need special glasses Improved resolution Image processing can be Improved signal-to-noise lengthy ratio Computational resources Near-atomic resolution needed possible Features smeared if input 3D structure images are poorly aligned determination Requires many identical Can manipulate and objects dissect 3D structure maps Requires random orientations Full 3D structure (inside and out) if images are of cryo-specimens Can separate conformational variants (continued)

570

D. M. Belnap

Table 1 (continued) Phage Isolated and Method particles host Prosa Three-dimensional x x View in vivo structures image Can get structure of reconstruction, non-averageable tomography (nonidentical) objects Improved resolution Improved signal-to-noise ratio 3D structure determination Can manipulate and dissect 3D structure maps Full 3D structure (inside and out) if images are of cryo-specimens Three-dimensional x x Improved resolution image Improved signal-to-noise reconstruction, ratio tomography, Near-atomic resolution sub-tomogram possible averaging 3D structure determination Can manipulate and dissect 3D structure maps Full 3D structure (inside and out) if images are of cryo-specimens Can separate conformational variants Compensates for tomographic artifact from tilting restriction Immuno-labeling x x Identify phage component Can be used with any other EM method Metal labels easily seen Could be used without secondary and metal label (i.e. primary antibody only) Affinity grids, x x Concentrate sample on antibody, or other grid surface affinity label Purify phage from small amount of starting sample

Consa Microscope stage must be well aligned Image processing can be lengthy Computational resources needed Tilting restricted, so data missing from angles that cannot be accessed (artifact in 3D result) Features smeared if input images are poorly aligned Image processing can be lengthy Computational resources needed Features smeared if input volumes are poorly aligned Requires many identical objects Requires random orientations Only works for portion of tomographically reconstructed volume

Identifying and making antibody may be difficult Labeling can be difficult if antigen conformation is compromised (e.g., in thin sections)

Finding and making specific antibody or affinity-binding entity may be difficult Must attach antibody or other affinity agent to grid surface

Chemical fixation could be used for any of the specimen preparation methods, even cryogenic ones. That could be a disadvantage in that the sample may be changed from its native state. On the other hand, fixation can be an advantage. For example, the treatment often produces higher resolution in 2D averaging and 3D reconstruction of cryogenic specimens because fixation prevents some molecular motions and dissociations

a

Detection of Bacteriophages: Electron Microscopy and Visualization

571

Metal Shadowing Metal shadowing involves placing a thin metal coating over a biological object (Bozzola and Russell 1999c; Chandler 1986; Hendricks 2014). An air-dried, freeze-dried, or critical-point-dried specimen is placed in a vacuum, and metal is evaporated onto the specimen (Fig. 4a). The metal follows the shape of the object. When a specimen is shadowed, the metal vapor comes from one direction. The specimen is either held still or rotated. Holding still causes the metal to only coat from one direction, leaving some of the specimen uncoated and giving the impression of a shadow. If the metal grain size is small and the coating thin, reasonably high resolutions may be obtained. The greatest advantage of metal shadowing is enhanced contrast. Disadvantages include long preparation times and potential distortions from sample drying and from imprecise metal coating. Metal shadowing EM has been an important technique in phage work (Fig. 5). For example, T4-infected E. coli was found to be almost completely filled with T4 bacteriophage (Wyckoff 1948). An image of expelled DNA surrounding a single T2 phage conveys spectacularly the feat of DNA packaging – a relatively large amount of DNA (49  4 μm in length) fits into a container with width and length dimensions 500–600 times smaller (Bradley 1967; Kleinschmidt et al. 1962). When contracted, phage MM tail sheaths had the same helical symmetry as phage T4 tail sheaths, but the symmetry differed when sheaths were extended (Müller et al. 1994). In addition, Fraser and Williams (1953) used this method to show that bacteriophages T3 and T7 were not tailless spheres but had short, stubby tails and hexagonal, “geometrical” heads. Freeze-drying allowed the short, stubby tails to be seen clearly. Tails were less clear in air-dried specimens. They also used metal shadowing to describe the morphology of T-phages (Williams and Fraser 1953). Kay and Bradley (1962) used metal shadowing to characterize bacteriophage ϕR. Metal shadowing was used to characterize phage T5 DNA (Saigo 1975) and measure the length of phage ϕ29 DNA (Anderson et al. 1966). The method was used to confirm the presence of a DNA loop at the replication fork, in a study of phage M13 DNA and phage T7 replication proteins (Park et al. 1998). With the advent of negative staining, usage of metal shadowing in phage studies declined. One form of metal shadowing, freeze-fracture EM (Severs 2007), is useful for studying phage-host interactions. Freeze fracture involves rapidly freezing the specimen, fracturing the specimen, and then coating the fractured surfaces with metal (typically platinum-carbon). The tendency of the fracture to split membranes into half-membrane leaflets allows membrane structure, and phage-host interactions, to be observed in a unique way. Freeze-fracture EM was used to study host interactions of phages ϕ6 (Bamford and Lounatmaa 1978), fd (Bayer and Bayer 1986), T4 (Tarahovsky et al. 1991), T5 (Plançon et al. 1997), and a Methylomirabilis-infecting phage (Gambelli et al. 2016).

572

D. M. Belnap

Fig. 4 Three methods used to prepare and image purified phage samples via EM. A cartoon is shown on the left and an example picture on the right. (a) Metal shadowing. Here the specimen is placed in an evaporative chamber. A vacuum is created. Then, a sample of metal (e.g., platinum) is heated by an electrode to produce metal vapor. Metal vapor is deposited on the specimen (Bozzola and Russell 1999c). If the specimen is held in place, the vapor is unidirectional and produces the appearance of shadows where metal is not deposited. In turn, this gives a three-dimensional appearance in the image. If the specimen is rotated (in plane of support film) during metal deposition, metal will be deposited evenly on the specimen. Rotary metal shadowing also gives a three-dimensional appearance in images. (b) Negative staining. A phage sample is placed on a support film and allowed to adsorb. Excess sample is blotted away, and the support film usually is washed to remove unbound components. Lastly, a solution of heavy metal salt is applied, blotted, and allowed to dry. This leaves phage particles encased in dried, metal salt (Bozzola and Russell 1999c; Bradley 1962; Hayat and Miller 1990). Often particles are distorted by this process, as illustrated in the cartoon. (c) Unstained, frozen-hydrated specimen. A phage sample is placed on a support film (usually perforated), blotted to make a thin film of solution, and then plunge frozen in a cryogen (usually liquid ethane or propane). Electron micrographs: a, top, phage T4 with contracted tails (Müller et al. 1994); a, bottom, phage T2 with expelled DNA (Bradley 1967; Kleinschmidt

Detection of Bacteriophages: Electron Microscopy and Visualization

573

Negative Staining Negative staining (Bozzola and Russell 1999c; Bremer et al. 1992; De Carlo and Harris 2011; Harris 1997; Harris and De Carlo 2014; Hayat and Miller 1990; Ohi et al. 2004) is the most commonly used technique for examining phages by EM. Ackermann and Tiekotter (2012) said negative staining is “arguably the technically simplest and most important single method in virology.” Preparing specimens is easy, fast, and inexpensive. Even if another EM technique is planned (e.g., cryogenic EM), negative staining is the most reliable EM technique for quickly assessing phage samples for such things as sample purity, aggregation state, and particle integrity. Well-prepared specimens have high contrast. Usually, many structural details can be seen. Images of negatively stained specimens are informationrich. Negative staining involves coating an object with a heavy metal salt (Fig. 4b). The biological sample provides a template, and the metal salt forms a “cast” around the biological object. The cast is then imaged and gives a representation of the biological object, hence the designation “negative stain.” If an object is porous, the metal salt will also penetrate to internal regions. In contrast to directly labeling a phage as in other staining techniques (positive staining), negative stain increases the density of the volume surrounding a phage particle. Preparing a negatively stained specimen involves a few simple steps that can be completed in less than 5 min. Variations of the method abound (e.g., Ackermann 2009; Bozzola and Russell 1999c; Harris 1997; Harris and De Carlo 2014; Hayat 1986; Hayat and Miller 1990; Ohi et al. 2004) but follow a basic pattern: • A solution of phage is placed on a support film, usually thin carbon, held by a grid support. During a prescribed time, e.g., one-half to 5 min, particles are adsorbed to the carbon support. The specimen is blotted with filter paper to remove excess sample. • The specimen is washed with water or buffer to remove unbound materials and then is blotted with filter paper to remove excess wash solution. This washing step is sometimes omitted.

ä Fig. 4 (continued) et al. 1962); b and c, phage 9NA (Casjens et al. 2014; Wollin et al. 1981) (S. Casjens, E. Gilcrease, and D. Belnap, unpublished, same sample used to prepare both specimens and both images are shown at same magnification). A note about imaging conventions in TEM. For metal shadowing experiments (a), the most dense region (metal coating) is usually shown as white, with the shadows black, to give a three-dimensional impression. For images of negatively stained specimens (b), the metal salt is usually shown as black to highlight the particle as white. For unstained, frozen-hydrated experiments, particles are often shown as black as in c, but may also be shown as white. The reader should be aware that an author may choose any convention for any EM image. Imaging software makes switching conventions (also known as inverting contrast) a trivial operation. (Micrographs in Panel a (right, top and bottom) reprinted from Müller et al. (1994), copyright 1994, and Kleinschmidt et al. (1962), copyright 1962, respectively, with permission of Elsevier.)

574

D. M. Belnap

Fig. 5 Examples of metal shadowing in phage studies. (a) Rotary-shadowed bacteriophage T2 particle with its DNA expelled; the DNA is spread on a Langmuir trough (Bradley 1967;

Detection of Bacteriophages: Electron Microscopy and Visualization

575

• The specimen is placed in a solution of a metal salt for a prescribed time, e.g., 15 s to 1 min, followed by blotting with filter paper and air- or vacuum-drying. The specimen can then be imaged. Uranyl acetate is commonly used for phage specimens, but other metal salts can be used and have advantages. These include uranyl formate, phosphotungstate, ammonium molybdate, and more than a dozen others (Bozzola and Russell 1999c; Bremer et al. 1992; De Carlo and Harris 2011; Harris 1997; Harris and De Carlo 2014; Ohi et al. 2004). Negative staining is an excellent technique, but limitations include drying effects, chemical effects, limited resolution, and viewing of only stain-accessible surfaces. The physical effects of drying and the chemical effects of the stain likely limit the resolution obtainable. Sometimes negatively stained particles clearly appear crushed, collapsed, or distorted. Particle flattening is a common effect observed in negatively stained specimens (Hayat and Miller 1990; Ohi et al. 2004). Air-drying squashes a “specimen whether it is a cell or a single macromolecule” (Chandler 1986). In a test of phage viability, phages had a high survival rate if negative-stain salts were mixed in the sample solution; however, if phage were dried in a vacuum with or without the presence of negative-stain salt, phages were not viable (Bradley 1962). The use of trehalose in combination with a uranyl acetate or ammonium molybdate stain appears to provide at least some protection to the specimen (Harris et al. 1995). Structure also may be affected by pH. The pH of stains varies (Bremer et al. 1992). The pH of many stain solutions is not physiological (Bremer et al. 1992). For example, uranyl acetate has a low pH. Non-uranyl stains such as tungstate and ammonium molybdate can be neutralized (Harris and De Carlo 2014; Ohi et al. 2004). pH changes as the specimen dries (Bradley 1962). Nevertheless, negative staining works well on a wide range of specimens (Bremer et al. 1992; Ohi et al. 2004), and studies of model macromolecules have documented excellent correlation between results of negative-stain EM and other techniques such as X-ray crystallography and cryogenic EM (Harris et al. 2001; Steven and Navia 1980; Stoops et al. 1992). In one case, an enzyme complex in negative stain was 20% smaller than the frozen-hydrated (cryogenic EM) complex (Stoops et al. 1992). Harris et al. (2001) stated, “Transmission electron microscopy of negatively stained specimens, despite limited resolution, has the potential to reveal a valid representation of surface features.”

ä Fig. 5 (continued) Kleinschmidt et al. 1962). Scale bar, 1 μm. (b) Infection of E. coli by bacteriophage T4 (Wyckoff 1948). Specimens were unidirectionally shadowed. Left, soon after infection, T4 particles can be seen around the cell periphery. Right, longer after infection, T4 particles filled the cell almost completely. (c) Phage T4 tails (top) with contracted tail sheaths and phage MM tails (bottom) with both contracted and extended tail sheaths (Müller et al. 1994). Specimens were unidirectionally shadowed. Scale bar, 100 nm. (Panel a reprinted from Kleinschmidt et al. (1962), copyright 1962; Panel b reprinted from Wyckoff (1948), copyright 1948; and Panel c reprinted from Müller et al. (1994), copyright 1994; all reprinted with permission of Elsevier.)

576

D. M. Belnap

Negative staining has given us many striking, beautiful, and informative images of phages (e.g., Fig. 6). After all the work of growing and purifying phage, seeing them as negatively stained specimens is a most rewarding task! The number of examples is several thousand and growing (Ackermann and Prangishvili 2012). Negative staining lets us see fine structural details in head-tail phage (Fig. 6). The method helps characterize structural components, as it did in 1959 for bacteriophage T4 (Brenner et al. 1959). We see diversity and detail in phage from normal and extreme environments (Fig. 6) (Bradley 1967; Müller et al. 1994; Rachel et al. 2002). The development of negative staining between 1954 and 1959 (Brenner and Horne 1959; Hayat and Miller 1990) allowed people to “visualize viruses in unprecedented clarity and revolutionized virology and our understanding of viruses” (Ackermann 2011a). More than any other structural technique, negative staining facilitates our understanding of the phage world. Less commonly, negative staining EM can be used to visualize some phage-host interactions. For example, by this method, Bayer and Bayer (1986) observed the filamentous phage fd and the isometric phage MS2 simultaneously adsorbed to a single E. coli pilus.

Thin-Sectioning, Room-Temperature and Cryogenic Thin-section experiments are extremely useful for imaging tissues and cells. This technique is widely used in cell biology studies. Therefore, for understanding phages in their cellular context, a thin-section experiment should be considered. This is its primary advantage. Disadvantages of thin-sectioning are lengthy preparation time, harsh preparation procedure, and limited resolution. Though not a commonly used technique in phage EM studies, thin-sectioning has been used to observe phage-cell interactions (Fig. 7). Lenk et al. (1975) used this method to study interactions of phage P22 with its Salmonella host. They determined that proheads assembled in the cytoplasm and not on the membrane and that the scaffolding protein, important for capsid assembly, likely is organized into a shell on the inside of the capsid (Fig. 7b). A different mechanism of assembly was observed by thin-sectioning when T4 proheads were observed at the protoplasmic membrane of E. coli (Kellenberger et al. 1968; Simon 1972). In a study on the effect of mutations in bacteriophage T4 genes, infected E. coli cells were examined by thinsection EM (Keller et al. 1988). Mature DNA-filled T4 heads could be seen within infected cells. Prohead T4 particles bound to membranes were also observed. Some T4 mutants produced aberrant heads, including polyheads (long tubular structures) within cells. Another T4 study observed a bridge between the inner and outer E. coli membranes (Tarahovsky et al. 1991). Bayer and Bayer (1986) observed that bacteriophage fd extrusion occurs at membrane adhesion sites. Dai et al. (2010) observed phage BPP-1 infecting Bordetella bacteria along the full length of the cell, not just at the cell poles where binding of purified attachment protein was strongest. Direct observation of phage-induced cell lysis was shown in Methylomirabilis cells by thinsectioning (Gambelli et al. 2016).

Detection of Bacteriophages: Electron Microscopy and Visualization

577

Fig. 6 Nine examples of negatively stained phages. Clockwise from upper left panel: Acidianus bottle-shaped virus (ABV), an archaeal virus isolated from an Italian hot, acidic spring (Häring et al. 2005); bacteriophage 201ϕ2-1, a member of the family Myoviridae (Thomas et al. 2008); another myovirus, bacteriophage NC-G (Gogokhia et al. 2019); bacteriophage Utah, a member of the Siphoviridae family (E. Gilcrease, S. Casjens, and D. Belnap, unpublished) (Leavitt et al. 2017);

578

D. M. Belnap

Details are described elsewhere (e.g., Bozzola 2014b; Bozzola and Russell 1999b, c, d; Ellis 2014; Hayat 1986), but the preparation of thin-section specimens involves six basic processes: • Fixation. First, the sample is placed in a buffered, primary-fixative solution, e.g., 2.5% glutaraldehyde and 1% paraformaldehyde. After an appropriate incubation time, the primary fixative is removed, and the sample is incubated in a secondary fixative, typically 1–2% osmium tetroxide (OsO4). The goal is to preserve structure through chemical cross-linking. Uranyl acetate is typically added after OsO4 treatment as an additional fixative, but the OsO4 and uranyl acetate treatments at this stage also provide some staining of the sample. • Dehydration. Next, water is removed from the sample by treatment with increasing fractions of ethanol, e.g., 30%, 50%, 70%, 95%, and 100%. The goal is to replace water with a solvent that is compatible with both the aqueous cell and the hydrophobic embedding media. • Infiltration. First, the sample is treated with a transition solvent, e.g., acetone or propylene oxide. Next, the transition solvent is replaced with plastic resin. As with dehydration, an increasing mixture of epoxy to transition solvent is used, e.g., 50%, 75%, 80%, and 100%. The goal is to completely embed the sample in resin. • Polymerization. The embedded specimen and surrounding resin are placed in a mold, and the specimen is oriented for desired sectioning. Then, the sample is allowed to harden, usually by heating in an oven. The goal is to have a solid, plastic-embedded sample. • Sectioning. The embedded, hardened sample is now placed in an ultramicrotome, and thin slices (e.g., 33–90 nm) are cut. These “thin sections” are placed on a grid support. The goal is to make specimens that are thin enough to be viewed in the transmission electron microscope.

ä Fig. 6 (continued) bacteriophage NC-B, a member of the family Podoviridae (Gogokhia et al. 2019); Great Salt Lake halophage BN (B. Nelson, M. Domek, D. Belnap, unpublished); another Great Salt Lake virus, halophage CW02 (Shen et al. 2012); bacteriophage fd, a filamentous phage (Wang et al. 2006); and bacteriophage 9NA (S. Casjens, E. Gilcrease, and D. Belnap, unpublished) (Casjens et al. 2014; Wollin et al. 1981). For bottle-shaped ABV virions, thin filaments are clearly seen at the “base” of the “bottle.” Fibers appear to decorate the tail sheath of phage 201ϕ2-1. Phage NC-G is related to bacteriophage T4 and resembles it. The siphovirus, phage Utah, has a long, noncontractile tail with a curly fiber at the end. The podovirus phage NC-B has an elongated head. Ring-like objects decorate the siphovirus-like tail of halophage BN. Halophage CW02 is morphologically similar to podoviruses and was characterized as a member of the T7-like supergroup. The picture of phage fd clearly shows the flexibility of these virions. Salmonella phage 9NA has the typical isometric head, flexible and noncontractile tail, and brushy tail tip of the Siphoviridae. All scale bars, 100 nm. (The picture of Acidianus bottle-shaped virus is reprinted from Häring et al. (2005), copyright 2005, with permission of the American Society of Microbiology. The phage fd picture is reprinted from Wang et al. (2006), copyright 2006, with permission from Elsevier.)

Detection of Bacteriophages: Electron Microscopy and Visualization

579

Fig. 7 Infection of Salmonella by phage P22 as observed by thin-sectioning EM. P22 particles can be seen within cells and attached to the outer membrane. Full (black), partially full (dark gray), and empty (light gray) particles can be seen. Staining of the highly packed DNA in P22 makes full particles conspicuous. (a) P22 heads accumulated in infected cells because the mutant P22 strains were deficient in head completion (left) or tail formation (right) (Lenk et al. 1975). (b) Infection with a P22 mutant strain deficient in DNA encapsulation; P22 proheads are visible (see inset at higher magnification), and scaffolding protein was deduced to form the inner ring (Lenk et al. 1975). (c, d) Infection of P22 strain H5 into wild-type Salmonella (Salmonella enterica serovar Typhimurium strain LT2). In Panel c, samples were prepared 50, 70, and 50 min after infection, respectively, left to right. Bars, 200 nm. In Panel d, sections of a sample prepared by the Tokuyasu technique 60 min after infection. The inner and outer cell membranes are clearly visible in the second image from the left. Note the extensive wrinkling of the Salmonella cells in all four images. Wrinkling is a common artifact of the Tokuyasu method. Bars, 200 nm. (Panels a and b reprinted from Lenk et al. (1975), copyright 1975, with permission from Elsevier. Panels c and d by Willisa Liou (unpublished data).)

580

D. M. Belnap

• Staining. Unless stained, features in plastic-embedded specimens will be difficult to see via TEM. After sectioning, uranyl acetate and lead citrate are often added for additional staining. These compounds are thought to interact with phosphate, amino, hydroxyl, and OsO4-derivative groups. (In the fixative stage, OsO4 reacts primarily with lipids and not only produces chemical links; it also adds density and hence is a stain.) The staining by osmium tetroxide, uranyl acetate, and lead citrate is positive staining – i.e., the stain interacts with a specific chemical group and becomes a tag for that group. The specimen is ready for imaging by TEM. Rapid freezing provides a faster rate of immobilization than chemical fixation because molecules and cellular components are simultaneously immobilized. This cryogenic method involves first rapidly freezing (in milliseconds) the sample, usually at high pressures (2  108 Pa). Freezing in this ultrafast manner prevents formation of ice crystals. (Liquid nitrogen is typically the cryogen used.) To obtain thin sections following freezing, one alternative is to then exchange fixation and dehydration solvents at sub-zero temperatures (He and He 2014; Mielanczyk et al. 2014). This exchange is known as freeze-substitution. Then the sample is brought to room temperature for resin-embedding, sectioning, staining, and imaging. A second, difficult alternative is to keep the sample at liquid-nitrogen temperatures (approximately 180  C) as it is cut on an ultramicrotome or milled in a focused-ion beam scanning electron microscope and as the specimen is imaged by TEM (Al-Amoudi et al. 2004; Chlanda and Sachse 2014; Mielanczyk et al. 2014; Villa et al. 2013). This alternative preserves native or near-native structure, though artifacts can be produced from sectioning the sample. The Tokuyasu technique is a third alternative that involves freezing (Liou et al. 1996; Mielanczyk et al. 2014; Tokuyasu 1973) (Fig. 7d). In this method, chemically fixed samples are cryo-protected by sucrose and then frozen in liquid nitrogen but at normal atmospheric pressure. Sections are cut at low temperature (approximately 100  C) and subsequently thawed to room temperature. This method is commonly used to better preserve both antigenicity and membrane structure, but the method could also be used to improve structural integrity for non-immunological experiments (Liou et al. 2008). Shrinkage of cells is a common artifact, however (Fig. 7d).

Frozen-Hydrated Specimens, Unstained and Negatively Stained EM samples do not need to be dehydrated, just solid; therefore, phage samples can be frozen. Freezing eliminates the potentially structure-altering process of drying or replacing water with a substrate like plastic resin. After sample preparation, the cold specimens are imaged by transmission EM. This technique is commonly called cryogenic (T)EM or cryo-(T)EM. Frozen-hydrated, unstained phages are in a native (or near-native) environment. This technique is the best current way to preserve biological objects and keep them as a solid. The freezing is so rapid that water remains in a disordered or vitreous state,

Detection of Bacteriophages: Electron Microscopy and Visualization

581

Fig. 8 Three examples of frozen-hydrated phage specimens viewed by cryo-TEM methods. (a) Bacteriophage 9NA (Casjens et al. 2014; Wollin et al. 1981). Top panel, lower magnification view showing phage on perforated (holey) carbon support film. Phages are randomly dispersed on the support with many more in the bottom-left hole than in the other two holes. Several are also on the carbon film with some having darker “halos” around them indicating thicker ice. The phage heads are clearly seen at this magnification as punctate spots. Some tails are apparent as wisps in the perforations. Scale bar, 1 μm. Bottom panel, higher magnification view of the region outlined by the box in the top panel. Tails are easily visible at this magnification as well as a fork-like feature at the tail tips. Scale bar, 200 nm. (S. Casjens, E. Gilcrease, D. Belnap, unpublished). A few pieces of contamination are seen in these two pictures as dark or gray spots. This level of contamination is acceptable and common in cryo-TEM images. In the bottom panel, a few phage particles are obscured by contamination. (b) Bacteriophage Utah (Leavitt et al. 2017). Full and empty capsids are visible, as well as serrations in the tail and at the edge of heads. In the middle of the full and empty heads, respectively, packed DNA and capsomeres are visible. Scale bar, 50 nm. (E. Gilcrease, S. Casjens, and D. Belnap, unpublished). (c) Halophage BN isolated from the Great Salt Lake. Note the appendages visible on the tail. An isometric particle and portions of an empty phage particle are also visible. Scale bar, 50 nm. (B. Nelson, M. Domek, D. Belnap, unpublished). These four images were underfocused to enhance contrast and are not “close-to-focus” images

582

D. M. Belnap

similar to liquid water. (In slower forming crystalline ice, the water molecules order themselves before solidifying. This destroys native structures.) As long as liquidnitrogen temperatures are maintained, the sample will remain vitreous. Combined with the averaging inherent in two- or three-dimensional analysis, frozen-hydrated specimens have enabled the highest-resolution results. The preservation of native structure is the most important advantage of frozen-hydrated specimens. Preserving native structure facilitates three-dimensional reconstruction because undistorted views from different orientations can be joined computationally. An additional advantage includes the ability to see the entire structure, not just surface features. TEM images of frozen-hydrated phage are projections of the entire particle (Fig. 8). Everything inside and outside is superimposed onto the two-dimensional image. Everything inside and outside can therefore be reconstructed by computer image processing. Details are found elsewhere (e.g., Dubochet et al. 1988; Grassucci et al. 2007; Guo and Jiang 2014; Harris 1997), but preparing a cryogenic, frozen-hydrated phage specimen for TEM involves four steps: • A small volume of sample, usually purified phage particles (e.g., 2–5 μL), is placed on a support film, usually perforated, held by a metal grid support. Holey carbon is most common, but holey gold, continuous carbon, or ultra-thin carbon (on top of holey carbon) support films also are used. Grids are usually made of copper or gold. The goal is to have particles evenly spread over the support and especially in holes if perforated support film is used. The perforated film supports the ice and allows imaging of particles through the holes (Figs. 4c and 8a). • The specimen is blotted with filter paper to leave a thin film of sample solution on the grid support. This action has three goals: (1) allow rapid freezing, (2) be thin enough to be transparent to the electron beam, and (3) have a single layer of particles (i.e., particles are not stacked one upon another, Fig. 4c). • Rapidly plunge the specimen (grid, support film, and thin film of sample solution) into a cryogen. The cryogen is usually liquid ethane or propane. The goal is to produce vitreous ice. • The unstained, frozen-hydrated specimen is transferred into liquid nitrogen and is kept at liquid-nitrogen temperatures until imaging is completed. The goal is to preserve the vitreous ice state. Vitreous ice maintains the native environment and is transparent to electrons, the opposite of crystalline ice. Cellular samples (e.g., phage-host preparations) prepared by rapid freezing (see “Thin-Sectioning, Room-Temperature and Cryogenic” section above) are also frozen-hydrated specimens and are treated in a similar manner after high-pressure freezing unless freeze-substitution is used. Disadvantages are length of preparation time, maintaining cryogenic temperatures, use of low-dose imaging techniques, noisy images, and poor image contrast. • Significant time may be needed to prepare suitable specimens with ice that is not too thick, too thin, or too contaminated. If ice is too thin, phage particles will not

Detection of Bacteriophages: Electron Microscopy and Visualization

•

•

• •

583

be present. If the ice is too thick, the beam will not penetrate the specimen, the image will be corrupted, or particles may overlap. The ideal is to have ice that is slightly thicker than the maximum size of the particles (Fig. 4c). Vitreous ice can be defiled by contaminants in the cryogen, exposure to water vapor, or effects of freeze-drying (Grassucci et al. 2007). Contamination can be seen, and one’s goal is to avoid obscuring the objects of interest (e.g., Fig. 8a). If one’s goal is to do general characterization of phage shape, complexity, and size, a carefully prepared negative-stain specimen should give reliable and faster results. Cryogenic specimens are usually a second step with the goal of computing a 3D structure. Cryo-EM is done after negative-stain results are satisfactory showing adequate numbers of particles and good particle integrity. Keeping specimens at cryogenic (i.e., liquid nitrogen) temperatures can be challenging. Although the phase-transition temperature is 137  C or higher (Dubochet et al. 1988), ideally, to safely avoid phase changes in the vitreous ice, the temperature should not exceed about 150  C until imaging is completed. Therefore, cold storage and handling can take significant effort. Unstained specimens are easily damaged by the electron beam (e.g., Thomas et al. 2008; Wu et al. 2012). Therefore, when imaging, minimal electron doses (i.e., “low-dose” techniques) must be used. This means that an area of interest is found at low magnification (e.g., 1000–5000; Fig. 8a, top panel), focusing is done at higher magnification adjacent to the area of interest (e.g., 1.5 to 3 μm away), and the final picture is recorded using the first electrons to interact with the area of interest at the desired magnification. The sensitive nature of an unstained specimen also means it should be imaged with as few electrons as possible. Imaging with few electrons means that the signal-to-noise ratio of these images will be low. This is only overcome by averaging multiple views of particles that are identical. Because the unstained specimen and its solvent have similar densities, the image will have low contrast. Defocusing (usually underfocus) is used to improve contrast, but too much defocus gives a poor image. In-focus or close-to-focus images usually give an image where the object of interest may be very difficult to see. For example, large phage heads are usually easy to see, but the thinner tails may be difficult to see in images of unstained, frozen-hydrated specimens (cf. Fig. 8).

In spite of these challenges, the effort to make unstained, frozen-hydrated, cryogenic specimens is well worth the results that have been obtained (Fig. 8), especially when these images are combined with two-dimensional averaging or three-dimensional image reconstruction. Most images of cryogenic specimens are analyzed by these 2D or 3D processes. However, important details can often be seen in unaveraged 2D micrographs. For example, a serrated capsid edge was seen in images of a spindleshaped virus, and this observation was instrumental in developing a multistart, varyingradius helix model of spindle-shaped archaeal virus capsids (Hochstein et al. 2018). A combination of negative staining and frozen hydration has been used to improve contrast in frozen-hydrated specimens and to avoid the common dehydration effects of negative staining (Adrian et al. 1998; De Carlo and Harris 2011; Harris

584

D. M. Belnap

and De Carlo 2014; Ohi et al. 2004). This method generally follows the protocol for preparing frozen-hydrated specimens. The exception is, after the sample is placed on a support film and before blotting, the sample is immersed in a solution of heavy metal salt (e.g., ammonium molybdate). This method has given some impressive results. A disadvantage is that some complexes have fallen apart in the high ionic strength stain solution. If the frozen specimen is crystalline, diffraction patterns, electron images, or both are obtained. This includes two-dimensional crystals (reviewed in Belnap 2015) and three-dimensional microcrystals or nanocrystals (Nannenga and Gonen 2019).

Scanning Electron and Helium-Ion Microscopy Scanning electron microscopy (SEM) involves the detection of reflected electrons as a small beam is scanned across a specimen. To produce enough signal, samples are coated with metal (Bozzola 2014a). In this sense, sample preparation is similar to that done for metal shadowing. Surface features are detected, and resulting images usually have a three-dimensional appearance. Larger specimens than those for TEM can be imaged. Disadvantages include typically lower resolution than TEM, long preparation times, distortions caused by sample drying and metal coating, and surface-only views. As noted by Ackermann (2012), SEM has had limited usefulness in bacteriophage studies. Nevertheless, past results have given similar views to those obtained by metal shadowing (Broers et al. 1975; Hermann and Müller 1991; Wendelschafer-Crabb et al. 1975) (Fig. 9a). Scanning helium-ion microscopy (SHIM) gives images similar to those from SEM but has significant advantages compared to SEM (Almeida et al. 2018; Joens et al. 2013; Leppänen et al. 2017). SHIM instruments operate similarly to SEM instruments, except a small beam of helium ions rather than electrons is scanned across the specimen. In SHIM, resolution is significantly improved (0.5 nm). Most importantly, biological samples do not need to be coated with metal. The latter removes the problem of metal coatings obscuring subtle surface features. Unstained biological objects can be imaged with this method. For example, direct imaging of bacterial lawns and phage plaques is possible, so phage can be observed in their native environment (Leppänen et al. 2017) (Fig. 9b). Although chemical fixation, dehydration, and drying were needed (Leppänen et al. 2017) and thus artifacts from these processes may be present, the ability to visualize unstained phage in their natural environment is a tremendous step forward. For studies of surface features of phage or of phage-host interactions, SHIM should be considered.

Scanning Transmission Electron Microscopy, Dark-Field Scanning transmission electron microscopy (STEM) can give molecular weight information of unstained particles (Belnap 2015; Thomas et al. 1994; Wall and Hainfeld 1986). A small beam is scanned across the specimen as in SEM, but, in

Detection of Bacteriophages: Electron Microscopy and Visualization

585

Fig. 9 Bacteriophage T4 viewed by scanning microscopy. (a) A single bacteriophage T4 viewed by scanning electron microscopy (Broers et al. 1975). The head, collar (white arrows), tail, and base plate can be seen. Two rodlike structures (possibly tail fibers) can be seen (black arrows) and are partially obscured by the metal coating. Small protrusions (s) can be seen coming from the base plate. (b) T4 phage infecting E. coli cells viewed via scanning helium-ion microscopy (Leppänen et al. 2017). This is an infection near the edge of a plaque on an agar plate. The specimen was not coated with metal. The sample was fixed with glutaraldehyde in sodium cacodylate buffer and then was dehydrated with ethanol and dried via critical-point drying. (Panel a from Broers et al. (1975) and reprinted with permission from American Association for the Advancement of Science. Panel b reprinted from Leppänen et al. (2017), copyright 2017, with permission of John Wiley & Sons.)

STEM, the electrons that pass through the specimen form the image. In principle, the number of electrons scattered by a molecule should be proportional to its molecular weight. This condition is met in dark-field STEM imaging. Dark-field imaging results when the image is formed only from scattered electrons. Normal TEM images are formed from both scattered and unscattered electrons. Therefore, dark-field STEM experiments can be used to measure molecular mass. For example, Cerritelli et al. (1996) used dark-field STEM to measure the masses of bacteriophage T4 fibers (Fig. 10).

In Situ, Liquid-Cell Electron Microscopy As stated, most electron microscopy is done at high vacuum (e.g., TEM is at pressures of 10–5 Pa or lower). Therefore, samples must be solid. Liquid bacteriophage samples are dried or frozen before being placed in the low-pressure environment of an EM. Liquid or gaseous samples would evaporate or dissipate, respectively, in that low-pressure locale.

586

D. M. Belnap

Fig. 10 Scanning transmission electron microscopy used for mass measurements of bacteriophage T4 tail fibers (Cerritelli et al. 1996). (a) Unstained bacteriophage T4 imaged by STEM. Arrow highlights a long tail fiber. A tobacco mosaic virus particle extends from the bottom right corner. TMV served as a mass calibration standard in this experiment. Scale bar, 50 nm. (b–d) STEM image (left) and histogram of mass measurements (right) of unstained, freeze-dried T4 proximal half fibers, distal half fibers, and long tail fibers, respectively. Scale bar, 50 nm. (Figures reprinted from Cerritelli et al. (1996), copyright 1996, with permission from Elsevier.)

In recent years, observation of liquid and gaseous specimens has become practical (de Jonge et al. 2009; de Jonge and Ross 2011; Gilmore et al. 2013; Wu et al. 2016a). Liquid or gas is placed in a small-volume, sealed container (Fig. 11). The container has windows of amorphous, electron-transparent material, such as silicon nitride, through which the electron beam can pass. Each window is approximately 30 nm thick. The sealed container does not compromise the high vacuum of the electron microscope and allows the specimen to be maintained at, for example, atmospheric pressure. To visualize chemical, physical, or biological changes, other substances or objects can be added to the chamber at specific times, or the temperature of the cell can be changed. A solid object or cell can be placed in the chamber and a secondary liquid or gaseous sample introduced via microfluidics. By treating the inner surface of one or both windows, specimens can be held fast (Fig. 11). This development enables visualization of dynamic processes in a native or nearnative environment. In a methods development project, infection of E. coli by P1 bacteriophage was observed (Kennedy et al. 2016). Imaging of non-phage samples, such as rotavirus (Gilmore et al. 2013; Varano et al. 2015), ferritin (Evans et al. 2012), liposomes (Hoppe et al. 2013), micelles (Parent et al. 2017), nanolipoprotein

Detection of Bacteriophages: Electron Microscopy and Visualization

587

Fig. 11 Liquid-cell transmission electron microscopy. Top, a specimen is sealed in a liquid cell with thin, transparent silicon nitride windows that allow passage of the electron beam and protection of the high vacuum environment within a TEM (de Jonge and Ross 2011). Bottom, the siliconnitride barrier can also be used as a platform to anchor specimens inside the chamber. Here, antibodies are used to anchor macromolecules (Varano et al. 2015)

disks (Evans et al. 2012), and transcribing rotavirus (Dukes et al. 2014), shows that liquid-cell TEM has potential for better understanding phage biology. However, significant care must be taken to be successful, and the resolution obtained is less than that available through cryo-EM. As in cryo-EM, in situ specimens are sensitive to the electron beam, and hence microscopy also results in low-contrast, high-noise images. Notably, one must maintain specimen viability under the intense radiation of the electron beam and avoid blurry images from the motion of objects in liquid. For the former, careful measurement of electron dose may be required (Kennedy et al. 2016). For the latter, immobilizing particles is helpful (Gilmore et al. 2013). Resolution is improved when (narrow-beam) STEM is used rather than (wide-beam) TEM, especially for thicker (>500 nm) specimens. However, resolution is still in the nanometer range, whereas cryo-TEM studies often achieve near-atomic resolutions. Perhaps brighter, pulsed beams (Evans and Browning 2013) and faster, more sensitive cameras will be necessary to improve resolution of dynamic specimens.

588

D. M. Belnap

Analytical Electron Microscopy Analytical EM is used extensively to chemically characterize materials in relation to spatial position. Some electrons passing through the specimen (via TEM or STEM) will lose energy corresponding to elements, bonds, and other chemical characteristics. By filtering electrons with respect to their energy, this phenomenon is used to chemically characterize a specimen (electron energy loss spectroscopy, EELS). Interaction of the electron beam with the specimen (via STEM or SEM) produces X-rays characteristic of the elements in the specimen. The X-rays are detected and mapped to spatial position (energy-dispersive X-ray spectroscopy, EDXS). The STEM imaging method HAADF (high-angle annular dark-field imaging) is sensitive to the spatial positions of elements in a specimen. Analytical EM has been used rarely in phage studies. Nevsten et al. (2012) observed that Mg2+ (DNA counter ion) was approximately 2–4 times higher inside phage lambda heads compared to the surrounding buffer solution. Phosphorus was only detected in the phage head. Neither Mg nor P was found in the tail, suggesting no DNA is present in fully packaged phage lambda tails. Frost and Bazett-Jones (1991) detected phosphate in the single-stranded DNA phage f1. Beyond these studies, analytical EM methods could be used to enhance image contrast of thin-section or other specimens (Özel et al. 1990). The advent of better electron detectors may increase the usefulness of analytical EM for phage studies, particularly for in situ, liquid-cell EM experiments (Hart et al. 2017). Current, multiple-detector systems for EDXS are expanding its capability to assay biological specimens.

Two-Dimensional Image Averaging and Diffraction If imaged particles are identical and are imaged in the same three-dimensional orientation, the images can be aligned and averaged to improve the signal-to-noise ratio (Fig. 12). Averaging of the two-dimensional images enhances superimposed features and diminishes noise or non-superimposed features. Particle images are selected from an electron micrograph, which shows an entire field of view. Selection involves extracting the particle image and some background from the larger image as a new, separate image. Usually many are selected. Then particle images may be processed in at least two ways: • All imaged particles are assumed to be the same and to have the same orientation. Images are aligned translationally and rotationally before they are summed. • Imaged particles are assumed to have different orientations, different conformations, or both. Images are classified using algorithms that determine the degree of similarity between images, and images in the large set are put into subsets corresponding to that similarity (e.g., Frank 2006; Glaeser et al. 2007). (Some different orientations, conformations, or objects are distinct enough and can be classified visually, i.e., by human eyes.) Images are aligned rotationally and translationally and then summed.

Detection of Bacteriophages: Electron Microscopy and Visualization

589

Fig. 12 The 13-step process of two-dimensional averaging and three-dimensional image reconstruction via transmission electron microscopy of phage specimens. (1) Phage particles are produced and purified by various means. The phage does not have to be absolutely pure, but the sample must be clean enough so that phage particles can be distinguish and not obscured in the electron micrographs. (2) A small volume (e.g., 2–5 μl) of purified phage is applied to a TEM grid coated with a support film. For negative-stain specimens, the support film is typically continuous carbon. For frozen-hydrated specimens, the support film is usually perforated carbon (“holey carbon”). (3) Negatively stained and frozen-hydrated specimens are known as “thin-film techniques” (Harris 1997). Blotting leaves a thin film of the purified sample in both negatively stained and frozenhydrated preparations. (3a) Negative staining is accomplished by washing the blotted grid and after applying a heavy metal salt (i.e., negative stain) solution. Blotting is done after washing and applying the metal salt. The specimen is then dried. (3b) For frozen-hydrated specimens, the grid is rapidly plunged into a cryogen (e.g., liquid ethane or propane). This produces vitreous ice. The specimen is then transferred to liquid nitrogen. (4) Specimens are transferred to the microscope. Negative-stain specimens are usually imaged at room temperature. Frozen-hydrated specimens must be maintained at liquid nitrogen temperatures until imaging is completed. (5) Images are recorded in the TEM. Low-dose methods (minimal electron exposure) are critical to obtain highresolution images of frozen-hydrated specimens. Low dose can also be used for negatively stained

590

D. M. Belnap

Spectacular improvements can be observed, especially for noisy images of frozenhydrated specimens (Fig. 13). Two-dimensional classification is especially helpful for determining different conformations or oligomeric states because these differences can be very difficult to see in unaveraged images – even in the higher contrast images of negatively stained particles (Fig. 13b). Negatively stained specimens can be very helpful for classification experiments. Ohi et al. (2004) suggest negative stain may be even more useful than frozen-hydrated specimens because negatively stained particles usually have a few preferential orientations on the support film, compared to often more random orientations in vitreous ice. The increased contrast in negative stain also is helpful. STEM (Cerritelli et al. 1996) and SEM (Hermann and Müller 1991) images also have been aligned and averaged. Another form of 2D image averaging is to compute diffraction patterns from repeating motifs within two-dimensional images. Lepault et al. (1987) used diffraction patterns computed from images of frozen-hydrated phage lambda and T4 to investigate DNA packaging. In their study of conformational changes in the T4 capsid, Kocsis et al. (1997) also used diffraction patterns. Optical or computer filtering of diffraction patterns was used to clearly see hexagonal subunits in phage T4 heads and baseplates (Crowther and Klug 1975).

ä Fig. 12 (continued) specimens. Software enables semiautomatic data collection. Usually TEM images are recorded slightly underfocused. (6) Recorded images are examined by eye and by algorithms which determine resolution and image-correction parameters. These results can be used to determine if a micrograph is suitable for image processing. Good images have low astigmatism, low drift, and high resolution. Underfocusing produces an artifact that is modeled by the contrast transfer function (CTF). CTF parameters for each micrograph are determined and used for image correction. CTF modeling is also helpful in assessing image quality. (7) Individual particle images are selected and extracted from the micrographs. (8) Images are adjusted to be statistically similar, i.e., to have the same or similar background, average, and deviation. The set of images can now be used for two-dimensional classification and averaging and for three-dimensional image reconstruction. (9) Two-dimensional class averages can be determined by classification methods. Classification joins subsets of similar images and aligns and averages images within the subset. Alignment includes translational (x,y) alignment and alignment of one rotational, in-plane angle. Most studies of negatively stained specimens end at this step. (10) Three-dimensional image reconstruction begins with this step. Input is either individual particle images or average images. In this step, images are aligned for 3D, i.e., two translational (x,y) and three rotational (ϕ,θ,ψ) parameters (Heymann et al. 2005, 2006) for each image are determined. (11) Once these values are known, the 3D image reconstruction can be computed. (12) Various techniques are applied to assess the quality of the reconstruction, e.g., resolution. The image reconstruction can be manipulated and displayed in programs such as UCSF Chimera (Goddard et al. 2007) used here. After the initial reconstruction is computed, steps 10–12 are usually repeated many times to refine the alignments of the imaged particles. The structure is considered finished when no further improvements (e.g., improvements in resolution) are seen. (13) After the final reconstruction is computed, it is analyzed further and interpreted, often through the use of models. The structure is deposited in the EM Data Bank and Protein Data Bank (if model coordinates are produced), and the work is published. Example pictures used here are of bacteriophage P22, structure EMD-1220 in the EM Data Bank (Lander et al. 2006)

Detection of Bacteriophages: Electron Microscopy and Visualization

591

Fig. 13 Features of bacteriophages T7 and 9NA enhanced by two-dimensional image alignment and averaging. (a) 2D image averaging of tailless T7 mutant structures that helped show that DNA is packaged by spooling around a central axis within the capsid (Cerritelli et al. 1997). Background, image of frozen-hydrated preparation of T7 tailless mutant. Some particles contain DNA, and others are empty (thin-walled particles). The DNA-filled particles were noted to lie with the connector-core vertex perpendicular to the specimen plane (e.g., one black dot) or with that vertex parallel to the specimen plane (e.g., two black dots). DNA strands can be seen in the midst of the DNA-filled particles. Scale bar, 50 nm. Inset, upper left corner, an average of 77 aligned images of DNA-filled particles (core vertex perpendicular to image plane). The DNA packing is enhanced because of the averaging and is clearly seen to be aligned around the central axis. (b) Classification of images of negatively stained T7 helicase/primase (Ohi et al. 2004). Two distinct oligomeric states were found. Background, a field of view of T7 helicase/primase. Scale bar, 50 nm. Inset, (1) hexameric and (2) heptameric forms. Scale bar, 30 nm. (c) 2D classification of images of frozen-hydrated bacteriophage 9NA (Casjens et al. 2014; Wollin et al. 1981) baseplates. Left panel, a micrograph of phage 9NA with baseplates marked by arrows. Scale bar, 100 nm. Baseplate images (2,768) were selected and then classified with the program RELION (Scheres 2012). Right 12 panels, 12 good classes were obtained. Scale bar, 25 nm. (S. Casjens, E. Gilcrease, D. Belnap, unpublished). In this figure, note the different conventions used for the highest densities. In a and c (left panel), black represents the highest densities. Whereas, in b and c (right 12 panels), white represents highest densities. Both conventions are used in papers displaying cryo-EM data. (Panel a reprinted from Cerritelli et al. (1997), copyright 1997, with permission from Elsevier. Panel b reprinted from Ohi et al. (2004), copyright 2004, with permission of Biological Procedures Online.)

592

D. M. Belnap

Three Dimensions from Two Only two-dimensional images can be recorded in an electron microscope. TEM images are 2D projection images of the imaged three-dimensional object. In other words, each point on the 2D image represents the sum of densities perpendicular to that point. Transmission electron images are analogous to medical X-ray images, which also record the mass thickness of the imaged tissues. Three strategies are used to get three dimensions from 2D EM images: surface imaging methods, stereo imaging, and three-dimensional image reconstruction. Metal shadowing, scanning electron microscopy, and scanning helium-ion microscopy image the surfaces of a specimen. The resulting image gives the impression of three dimensions (Figs. 4a, 5, and 9). Some images of negatively stained specimens also have a 3D appearance. The human brain perceives three dimensions when each eye views the same object from slightly different directions, e.g., 6–12 apart. Stereo imaging exploits this phenomenon. Two-dimensional images of the same object are recorded in the electron microscope by tilting the specimen, e.g., views from two angles 6–12 apart. When a reader views the two views simultaneously (one image per eye), a three-dimensional appearance can be seen. Taking stereoscopic pictures of the same specimen has been done for thin-sectioned (Bayer and Bayer 1986) and metalshadowed and other specimens (Anderson 1952). Taking stereo views in the microscope is not the only way to use this method. If one displays a three-dimensional object in a 3D rendering program, the output display is in 2D. Stereo views can be generated by at least some of these programs. This enables a reader to see stereo on a printed page or on a computer screen (Fig. 14). Three-dimensional structure can be determined, or reconstructed, by recording several views of a specimen, each from a different orientation, and using computer image processing to recombine the images into a 3D structure (Fig. 12) (e.g., Belnap 2015; Frank 2006; Glaeser et al. 2007; Guo and Jiang 2014; He and He 2014; Obr and Schur 2019; Sigworth 2016; Wan and Briggs 2016). Once view orientations and origins are determined (image alignment), the 3D structure is computed from the aligned images. The mathematical principle allowing this is known as the projection, or central-slice, theorem (Fig. 15). Interestingly, just as phage EM was part of the development of electron microscopy (i.e., instrumentation and specimen preparation), bacteriophage T4 structural studies were part of the development of image analysis and three-dimensional reconstruction techniques (Crowther 2004). For example, the first 3D reconstruction from electron micrographs was of the bacteriophage T4 tail (DeRosier and Klug 1968). Views from different orientations are recorded in one of two ways, and this depends on whether multiple identical objects or one unique object will be viewed. • If many are viewed, each object is assumed to be identical or nearly so. Ideally, the set of objects will lie in all possible orientations. This is termed “singleparticle analysis” (SPA) because the imaged particles are free-standing, unordered, or “unattached” and because many views of identical particles are considered as multiple views of the same particle (Fig. 12).

Detection of Bacteriophages: Electron Microscopy and Visualization

593

Fig. 14 Three-dimensional reconstructions of extremophilic archaeal phages. Pictures are shown in stereo. (a) Lemon-shaped haloarchaeal virus His1 (Hong et al. 2015). (b) Hyperthermoacidophilic Sulfolobus turreted icosahedral virus (Veesler et al. 2013), at truncated resolution

• If only one unique object is available, that object is rotated in the microscope over a large range, e.g., 70 to +70 at 1–3 intervals, and images are recorded at each step (Fig. 16). This is the tomographic method. (Ideally one would tilt the object to 90 , but the sample holder or specimen grid obscures the specimen at angles larger than about 70 .)

594

D. M. Belnap

Fig. 15 Illustration of the projection or central-section theorem. A three-dimensional object (bottom left) is projected into two dimensions (top left) by summing the densities perpendicular to the view direction. In other words, each intensity in the 2D projection image is proportional to the sum of the densities in the 3D object perpendicular to the view direction. Here, the 3D object, bacteriophage P22 (Lander et al. 2006), is projected along three orthogonal axes. The 3D object and 2D projection images are in real space. Real space 2D or 3D data can be transformed into frequency space (right) via the Fourier transform, a mathematical operation. The 2D Fourier (or frequency) representations (top right) correspond to the central sections of the 3D Fourier representation perpendicular to the view directions (bottom right). The inverse Fourier transform operation converts the 3D Fourier representation into the 3D object (bottom left). Therefore, the Fourier representation of a 2D projection of a 3D object is the central section of the 3D Fourier representation perpendicular to the view direction. Images from the transmission electron microscope are 2D projection images (i.e., top left images). Once the 2D images are aligned (i.e., origins and orientations determined), these images can be transformed into central sections of the 3D Fourier representation. If enough different views are present, then the 3D Fourier representation will be filled, and the inverse Fourier transform will give an accurate representation of the average 3D structure

Detection of Bacteriophages: Electron Microscopy and Visualization

595

Fig. 16 Three-dimensional, tomographic electron microscopy. (1) A specimen is prepared and loaded into the electron microscope. Specimen preparation may be similar to that for “singleparticle analysis” projects, or it may be different. Specimens may also be prepared by thinsectioning. As in SPA, the specimens may be cryogenic or non-cryogenic. (2) A suitable region of interest must be found. This region must be “tiltable” to the high angles used. In other words, the region of interest must be viewable at high tilt angle. Regions near edges of the specimen grid or holder may not be. (3) The specimen is tilted in the microscope and images recorded at specified intervals (e.g., 60 to +60 ). This step is unique to tomographic experiments. The specimen may be tilted in SPA experiments, but usually only to one or two other angles besides 0 . (4) The 3D reconstruction or tomogram is determined from the images in the tilt series. This procedure is similar to the process for SPA. (5) After the 3D tomogram is computed, sub-volumes may be extracted for further analysis. These volumes may be similar structures so that all can be averaged together, or the volumes may differ and be separated by classification methods, e.g., Hu et al. (2013). (6) The sub-volumes are in different orientations (ϕ,θ,ψ) and x,y, and z positions. They must be aligned (analogous to the process used in SPA). Once aligned the sub-volumes are summed together to give a sub-tomogram average structure. The sub-tomogram example here is phage T7 interacting with the outer membrane of E. coli (Hu et al. 2013)

In both tomography and SPA, the view orientations are determined, and then the 2D images are mathematically combined to reconstruct the three-dimensional structure. Image analysis for tomography has many similarities to SPA, but includes some unique aspects, particularly the ability to further refine structures by aligning and averaging 3D sub-volumes (Figs. 16 and 17a). This technique is called sub-

596

D. M. Belnap

Fig. 17 Examples of tomographic, 3D structures. In these cases, the specimens were frozenhydrated. (a) Two stages in the interaction of phage T7 with E. coli mini-cells shown as surface

Detection of Bacteriophages: Electron Microscopy and Visualization

597

tomogram averaging (Wan and Briggs 2016; Obr and Schur 2019). For those identical components that are averaged in this manner, the resolution is enhanced significantly. For images of helical specimens, helical symmetry is used and is a tool to get multiple views and to improve averaging and resolution (Egelman 2010; Fromm and Sachse 2016; Sachse 2015). Diffraction patterns and images also can be collected from 2D or 3D crystalline specimens, and, in these cases, crystallography algorithms are used to “reconstruct” those structures (Glaeser et al. 2007; Nannenga and Gonen 2019). Not only does 3D image reconstruction allow structure to be observed in a more realistic context; 3D reconstruction produces a higher-resolution result than is seen in the two-dimensional images. The combination of images is a form of averaging and hence reduces noise, thereby increasing the observed signal and resolution. This is especially true for images of frozen-hydrated specimens, which are of necessity low in signal and high in noise (to avoid radiation damage). 3D reconstructions, including tomographic ones, are commonly computed from images of frozen-hydrated specimens, but reconstructions can be computed from images of other specimens. 3D reconstructions have been computed from images of negatively stained particles (DeRosier and Klug 1968; Harris et al. 2001; Messaoudi et al. 2003; Stoops et al. 1992) and from tilt-series images of thin-sectioned specimens (Gambelli et al. 2016). If the resolution is high enough, the final step in a three-dimensional EM project is to build a model of the structure. Ideally, this is an atomic model that can be built by tracing a polypeptide or polynucleotide chain through the high-resolution EM structure (e.g., Fig. 18). If this cannot be done and high-resolution components from another EM experiment or an NMR or X-ray crystallography experiment are available, a “pseudo-atomic model” may be built by taking and fitting the component models into the EM density (e.g., Aksyuk et al. 2009, 2012; Dai et al. 2010; Parent et al. 2012; Shen et al. 2012). Additionally, 3D EM structures of isolated components

ä Fig. 17 (continued) renderings (Hu et al. 2013): T7 head white, T7 tail fibers yellow, T7 tail orange, and cell membranes blue. Both structures were computed by tomography, followed by sub-volume extraction, classification, and averaging. Left, T7 tail fibers bound to the outer membrane. Right, T7 tail fibers bound to outer membrane and also penetration of the cell membranes by an extended T7 tail. (b, c) Phage P1 particles interacting with an E. coli mini-cell (Liu et al. 2011). Panel b shows a surface rendering. Three P1 particles (red) have contracted tails. One (blue) has an extended tail conformation. The mini-cell is shown in green. Panel c shows a slice through a 3D tomogram. P1 particle on left has contracted tail, and one on right is extended. Bars, 200 nm. (d) Slice through tomographic reconstruction of Sulfolobus solfataricus infected with Sulfolobus turreted icosahedral virus (Fu et al. 2010). Note the virus particle adjacent to a pyramidal-shaped structure in the cell membrane, here seen as a V-shaped portion of the membrane. Bar, 200 nm. Panels c and d again show the two different ways EM density is represented. In c, viral and cellular components are white with a dark background. In d, the opposite is true. (Panels b and c are reprinted from Liu et al. (2011), copyright 2011, with permission from Elsevier. Data for Panel d was obtained from the Caltech Electron Tomography Database (Ortega et al. 2019).)

598

D. M. Belnap

Fig. 18 Two phage structures at near-atomic resolution. (a) Bacteriophage Sf6 (Zhao et al. 2017). Here, six histidine residues (His 186) surround density at the center of a capsid hexamer. The inner density was modeled as a chloride ion, due to its abundance in the solution. The resolution of the Sf6 reconstruction was computed to be 2.9 Å. (b) Bacteriophage P22 (Hryc et al. 2017). An αhelical interface between two subunits within a capsid hexamer is shown. For each helix, residues at the ends and one in middle are labeled. The resolution is 3.3 Å

can be fitted into tomograms, where each component can be viewed in a cellular or other context. The method of computing three-dimensional image reconstructions from images of frozen-hydrated phage, phage and host, and other specimens has grown dramatically in usefulness and power in the last several years. Resolution has increased dramatically and now is approaching atomic resolution for many specimens including bacteriophage Sf6 (Fig. 18a) (Zhao et al. 2017), phage P22 (Fig. 18b) (Hryc et al. 2017), and Sulfolobus turreted icosahedral virus (Veesler et al. 2013). Even some tomographic studies, via sub-tomogram averaging, have reached near-atomic resolution (Schur 2019). Classification techniques, which are helpful for sorting conformations in two dimensions, also sort three-dimensional conformations (Scheres et al. 2007; Sigworth 2016). Both sub-tomograms and SPA reconstructions can be sorted by conformation (e.g., Fig. 17a) during the 3D reconstruction process (Figs. 12 and 16). This 3D classification has powerful potential because most, if not all, macromolecules in any biological specimen are not identical. Multiple conformations are likely the norm, not the exception. The floodgates of opportunity and knowledge are opening through technical developments in cryo-EM and through increased interest. Even structures of small phage components can now be solved to high resolution by electron microscopy (e.g., Gao et al. 2019). Three-dimensional, cryogenic EM has helped us learn valuable phage biology. Three-dimensional cryogenic EM (SPA) has led to insights into the maturation of bacteriophages, including HK97 (Cardone et al. 2014; Conway et al. 1995, 2001), T5 (Preux et al. 2013), T7 (Ionel et al. 2011), 80α (Spilman et al. 2011), and P22

Detection of Bacteriophages: Electron Microscopy and Visualization

599

(Parent et al. 2010a, b; Teschke and Parent 2010). 3D cryo-EM has given insights into other phage processes including P22 DNA packaging (Lander et al. 2006), P22 genome release (McNulty et al. 2018), SPP1 connector function (Orlova et al. 2003), the contractile mechanism of the T4 tail (Aksyuk et al. 2009, 2012), T4 baseplate assembly and function (Yap et al. 2016), and the packaging of ssRNA in phages Qβ (Gorzelnik et al. 2016) and MS2 (Dai et al. 2017; Koning et al. 2016). The structures of numerous bacteriophage heads have been solved by 3D cryo-EM, including those of extremophilic phages His1 (Hong et al. 2015) (Fig. 14a), Sulfolobus turreted icosahedral virus (Veesler et al. 2013) (Fig. 14b), Sulfolobus turreted icosahedral virus 2 (Happonen et al. 2010), P23-45 (Bayfield et al. 2019), and others (Hartman et al. 2019). Asymmetric structures solved by 3D EM have improved knowledge of genome packaging and ejection (Parent et al. 2018) and the detailed organization of phage P68 (Hrebík et al. 2019). Helical methods have been used to solve structures of filamentous phage fd (Wang et al. 2006) and IKe (Xu et al. 2019) and tail structures of T4 (Zheng et al. 2017) and ϕRSL1 (Effantin et al. 2013). Tomographic studies allow phage-host interactions to be studied in greater detail. For example, these studies have shown that Sulfolobus turreted icosahedral virus assembly was accompanied by formation of a pyramidal-shaped structure in the host cell membrane (Fu et al. 2010) (Fig. 17d). In Sulfolobus islandicus rod-shaped virus 2 infection, the pyramidal structure was also seen and found to have sevenfold symmetry and be a portal for cell lysis (Daum et al. 2014). Phage 201ϕ2-1 assembly in Pseudomonas includes the formation of a nucleus-like compartment (Chaikeeratisak et al. 2017). Bacteriophage PRD1 uses a tube formed from its inner membrane to deliver its genome into a host cell (Peralta et al. 2013). Acidianus tailed spindle virus has smooth, continuous density from its spindle-shaped capsid to its tubular tail (Hochstein et al. 2018). Other studies using tomography include ε15 (Chang et al. 2010) and P22 (Wang et al. 2019) infection of Salmonella; maturation of Syn5 inside its cyanobacterial host (Dai et al. 2013); T4 (Hu et al. 2015), T7 (Hu et al. 2013) (Fig. 17a), and P1 (Liu et al. 2011) (Fig. 17b, c) infection of E. coli; and ϕCb13 and ϕCbK infection of Caulobacter crescentus (Guerrero-Ferreira et al. 2011). Together with insights from SPA experiments, tomographic studies are giving us incredible knowledge of phage-host biology (Bertin et al. 2011; Guerrero-Ferreira and Wright 2013). Tomography has also been used to study isolated phage (Guerrero-Ferreira and Wright 2013).

Immuno-EM Antibodies, including antigen-binding antibody fragments (e.g., Fabs), to phage components are also used in phage-EM experiments. This labeling is used primarily to identify phage components in electron micrographs (Fig. 19). Most commonly, an antibody to a phage component is incubated with the sample, followed by treatment with a secondary antibody or protein A conjugated to a heavy metal particle. The second label has affinity for the first antibody (not the phage) and has a metal label, typically a conjugated gold particle a few to several nanometers in

600

D. M. Belnap

Fig. 19 Examples of phage immuno-EM. (a) Phage A118 particles immunogold labeled with antibodies to proteins gp19 (left), gp17 (middle), and gp16 (right) and negatively stained (Bielmann et al. 2015). (b–d) Immunogold labeling of bacteriophage BPP-1 major tropism determinant protein (Dai et al. 2010). Negatively stained two-dimensional views are shown in Panels b and c. Panel d is a slice through a 3D tomographic reconstruction made by tilting the particle shown in Panel c. The arrow highlights two gold particles connected to the same tail fiber. (Panel a is reprinted from (Bielmann et al. 2015), copyright 2015, with permission from Elsevier. Panels b-d from (Dai et al. 2010).)

diameter. In electron micrographs, the high-contrast metal label is easily seen (Fig. 19). Immuno-labeling was used to model phage A118 baseplate organization (Fig. 19a) (Bielmann et al. 2015), identify the major capsid and major tail proteins of phage p2 (Labrie et al. 2012), deduce protein functions for Shiga toxin 2 (Stx2)-

Detection of Bacteriophages: Electron Microscopy and Visualization

601

transducing phage Sp5 (Mondal et al. 2016), identify proteins in the “horn” of marine cyanophage Syn5 (Raytcheva et al. 2014), locate proteins in the baseplate of phage T4 (Watts et al. 1990), and locate proteins in thermophilic phage GVE2 (Wang and Zhang 2008; Wu et al. 2009). Primary antibodies can be used without a secondary, heavy metal label. Aebi et al. (1977) used antibodies to reveal location and conformational changes of proteins within phage T4 capsomeres. Both primary antibodies alone and primary antibodies with secondary metal labels were used to show that maturation of the T4 capsid involved radical reorganization of the major capsid protein, including translocation of specific parts of the protein from the inside to the outside or vice versa (Steven et al. 1991). Multiple specimen preparation and imaging methods can be used to image antibody-phage complexes. The most common method is negative staining of isolated virus-antibody complexes (e.g., Gowen et al. 2003) (Fig. 19). Dark-field imaging was used to study negatively supercoiled bacteriophage ϕX174 DNA with antibodies to Z-DNA (Revet et al. 1984). Metal shadowing was used to study phage T4 capsid maturation (Kistler et al. 1978). Scanning EM was used to visualize antibodies bound to phage Tu II-46 (Hermann et al. 1991). Scanning transmission EM (STEM) was used to help locate components of the phage T4 baseplate (Watts et al. 1990). Thin-sectioning was used to observe interactions of immuno-labeled phage fd with its E. coli host (Bayer and Bayer 1986). Two-dimensional image processing methods were used to study T4 (Aebi et al. 1977). A three-dimensional tomographic structure helped locate the major tropism determinant protein of phage BPP-1 (Dai et al. 2010) (Fig. 19b–d). Three-dimensional structures could also be computed from images of frozen-hydrated, antibody-labeled phage (primary antibody only), as has been done for non-phage viruses (e.g., Lin et al. 2013; Prasad et al. 1990). Analytical EM techniques could be used to enhance the contrast of metal labels (Özel et al. 1990). As with other macromolecules, phage particles can be difficult to isolate in concentrations high enough to be visualized confidently by electron microscopy. A promising technique is to attach antibodies or other affinity-purifying entities to the grid surface and use this to purify phage. This “affinity grid” method was used successfully with phages T7 and T3 (Yu et al. 2014, 2016).

Conclusion Bacteriophages are unlike what we see in the macroscopic world with our normal, unaided vision. As proof of this, note the response of people who are not used to seeing phage. When showing phage to visitors in my laboratory, I never tire of seeing their surprise and awe. Although I am no longer surprised by the heads, tails, and so forth, I delight in seeing a new phage under the microscope. That experience is joyful

602

D. M. Belnap

and never tiring. If not for other pressing priorities and finite disk storage space, I could spend many hours recording images! What a wonderful privilege to observe these fascinating biological entities! An old adage says “a picture is worth a thousand words.” If this is true, as pictures of expelled phage DNA (Fig. 5a), whole phage (Fig. 6), phage and host (Figs. 5b, 7, and 9b), and numerous other pictures attest, then spending time to get quality EM images is worth the effort. Hans-Wolfgang Ackermann, who examined at least 1800 phage, wrote much about phage EM, and tallied the vast number of phage papers, lamented that many published phage images were of poor quality (Ackermann 2007, 2012, 2014; Ackermann and Prangishvili 2012; Ackermann and Tiekotter 2012). Poor operation of the microscope was given as the primary problem and suggestions for improvement were given. Both maintenance and training should be improved. The microscope should be well aligned by trained personnel. Ackermann and others praised the outstanding work of another great phage microscopist, Anna Sergeyevna Tikhonenko, who worked in the former Soviet Union (Popenko et al. 2013). They noted she published “the first bacteriophage atlas” (Tikhonenko 1970), and most of her “micrographs were of exceptional quality.” To honor Ackermann’s and Tikhonenko’s legacy, I highlight seven recommendations to help phage researchers improve their microscopy. Because negative staining is the prominent method used in phage electron microscopy, I illustrate with images of negatively stained specimens. But these points apply to other EM techniques, as well: • Use Fourier representations to assess image quality. Most electron microscopes are equipped with digital cameras. Digital images (including scans of film or plate images) facilitate easy computation of Fourier, or frequency, representations (e.g., Fig. 15). Fourier-transform, power spectra, or fast-Fourier transform (FFT) algorithms turn an image into a frequency representation (usually amplitudes vs. spatial frequency). Such algorithms are likely included in scientific imaging programs or microscope camera software. Compute Fourier representations of EM images and use this capability to assess if your microscope is working well. When the field of view is underfocused, the Fourier pattern computed from a square image should be round and concentric and should extend to a reasonable resolution (Fig. 20). • Check and correct astigmatism. Astigmatism means lenses focus differently for different axes within the image plane (e.g., x- and y-axes). This commonly needs adjustment and is a common reason for poor images. To collect high-quality images, astigmatism should be checked at the beginning of an imaging session – at a minimum. Slight astigmatism is often not noticeable unless a Fourier representation is computed, but large astigmatic offsets give images a distinct, almostshaky appearance (Fig. 21). Astigmatism is difficult to get perfect, and so slight astigmatism is acceptable (Fig. 21b). Image processing algorithms exist to determine and correct astigmatism, but these computer programs are meant to correct slight, not extreme astigmatism.

Detection of Bacteriophages: Electron Microscopy and Visualization

603

Fig. 20 An example of a good micrograph as assessed by Fourier-transform methods. Top, image of phage T7. Bottom, power spectrum (a type of Fourier representation) of T7 image. Note the concentric (Thon) rings. These should be circular. The rings extend to the edge of the representation indicating that image resolution extends to the maximum for this magnification and detector. If the Fourier pattern for an image resembles this one, then one can be confident that the microscope is well tuned and is working well

• Check focus setting. Generally, transmission EM images are recorded slightly underfocused (0.5–2.0 μm). This focus region is usually optimal for seeing important features, such as phage components, in transmission electron micrographs. Overfocused and far-underfocused images are not desired because features are smeared or out of focus. Features are often difficult to resolve in at-focus

604

D. M. Belnap

Fig. 21 Examples of astigmatic and corrected images and the Fourier-transform pattern observed for each. (a) Four phage Basilisk (Grose et al. 2014) images ranging from no or corrected astigmatism (top) to most astigmatic (bottom), images on left and Fourier-transform patterns on right. Note the circular Fourier-transform (power spectrum) pattern on top right. This indicates astigmatism has been corrected. Astigmatism is modest in second image from top, and may not be

Detection of Bacteriophages: Electron Microscopy and Visualization

•

•

•

•

605

images (Fig. 22). Specimens should be viewed at the eucentric position. This position along the column axis is where a well-tuned microscope works optimally. Eliminate drift. Drift occurs when a specimen moves within the microscope while an image is recorded. Specimens in a well-functioning microscope should not drift beyond its specifications (Fig. 23). A variety of factors can cause drift, from unstable specimen-support films to temperature gradients within the microscope. The microscope should be repaired if it is at fault. A more stable support film may be necessary if the film is unstable. Sometimes just waiting after moving the specimen stage allows drifting to stop (Fig. 23). Use of objective aperture. Inserting the objective aperture of the transmission electron microscope generally improves contrast in the images (Fig. 24). This is often helpful in improving images. However, sometimes features may be clearer without the aperture (Fig. 24). Choose appropriately stained field of view. If one views thin sections of stained material, positive staining is desired (Fig. 7). The opposite is true for negatively stained specimens. Negatively stained phage specimens often have regions where staining is positive and not negative, and these regions can overlap or be adjacent (Fig. 25). If the purpose is to view purified phage via negative staining, positively stained areas generally have a poorer appearance than negatively stained fields of view because the stain obscures features. If a microscopist happens upon a positively stained field of view, often he or she can find negatively stained regions by searching the specimen grid. If not, the person should try applying the heavy metal salt solution for a shorter time. Minimize radiation damage. Radiation damage is a well-known problem in cryogenic EM work, but even some negatively stained specimens are sensitive to the electron beam (Fig. 26). For such specimens, the low-dose techniques used in cryogenic EM may be necessary. Unless exploiting radiation damage for a useful purpose, as has been done in some phage studies (Cheng et al. 2014; Wu et al. 2012, 2016b), radiation damage should be avoided or minimized.

ä Fig. 21 (continued) noticeable, except in the Fourier-transform pattern. For best imaging, this should be corrected. Often improvement can be seen. The bottom two examples have extreme astigmatism. Note the “shaky” appearance in the astigmatic images and the distorted pattern in the Fourier representations. (b) Bacteriophage 9NA (S. Casjens, E. Gilcrease, and D. Belnap, unpublished) images with Fourier-transform pattern inset. Left example has slight astigmatism (Fourier pattern is slightly elliptical) but is acceptable. Uncorrected, extreme astigmatism is shown in the two examples to the right. Note the shaky appearance and loss of resolution in the 9NA tails compared to the left image

606

D. M. Belnap

Fig. 22 A series of images of the same field of view of bacteriophage Basilisk (Grose et al. 2014) taken from 10 μm underfocus ( 10, top row) to 10 μm overfocus (+10, bottom row). On left is image and on right is Fourier-transform (power spectrum) representation. Note, the Thon-ring (concentric-ring) pattern disappears at focus (0), and it has a smaller internal circle at focus levels farther from focus. Note smearing of features in highly underfocused ( 4, 10) and overfocused (+4, +10) images. Tail striations are more clearly seen in slightly underfocused images ( 0.3 to 2) than at focus (0)

Detection of Bacteriophages: Electron Microscopy and Visualization Fig. 23 Examples of drift with images of bacteriophage Utah (S. Casjens, E. Gilcrease, and D. Belnap, unpublished). Inset on each image is the Fourier-transform (power spectrum) representation of the image. Top, an image where the specimen moved significantly during the exposure. The image is blurred, and the Fourier pattern is truncated in the direction of the specimen movement (here horizontal). Middle, the same field of view a few minutes after moving the specimen stage to this position. Note the sharpness of the image and the complete circles in the Fourier representation. (This image is slightly astigmatic, and so it would be further improved by correcting astigmatism.) Bottom, another image with less drift than the top image. The Fourier pattern is less truncated, and the blurring is less obvious

607

608

D. M. Belnap

Fig. 24 The same field of view of bacteriophage Basilisk (Grose et al. 2014) without (top) and with (bottom) an objective aperture in the transmission electron microscope. The image taken with the aperture in place has higher contrast. This may or may not be an advantage. For example, striations in the tails are clearer without the aperture. (Both images were recorded at the same focus setting)

These suggestions should help a phage researcher get better TEM images. Additional points are found in Ackermann’s and Tiekotter’s writings (Ackermann 2009, 2012; Ackermann and Tiekotter 2012; Tiekotter and Ackermann 2009). In particular, they note that poor EM images often are the result of poor samples, such as crude extracts. Bacterial debris can lead to erroneous conclusions. Some cell debris resembles phage or phage parts. Therefore, sample preparation often needs improvement to get better images. Whether describing (Bradley 1967; Tikhonenko 1970; Wurtz 1992), comparing (Parent et al. 2012; Pietilä et al. 2014), classifying (Ackermann 2003), or surveying (Ackermann and Nguyen 1983) phages, electron microscopy is a definitive and critical technique and well worth the effort.

Detection of Bacteriophages: Electron Microscopy and Visualization Fig. 25 Bacteriophage Basilisk (Grose et al. 2014) stained positively (top left of each panel) and negatively (bottom right of each panel). Lower magnification view (top panel) and higher magnification view (bottom panel). Some particles appear to be intermediate between positively and negatively stained (bottom panel). The tails are negatively stained, even those attached to positively stained heads

609

610

D. M. Belnap

Fig. 26 Radiation damage in a negatively stained specimen. Three images of the same bacteriophage Utah (S. Casjens, E. Gilcrease, and D. Belnap, unpublished) particles taken sequentially in order top, middle, and bottom. The first (top) image shows a relatively even background. The second (middle) image shows some new features in the background that were induced by the electron beam. However, the phages still are the most prominent features in the image. The last (bottom) image shows prominent, extensive features in what was the background. These new features now compete with the phage particles for prominence. Interestingly, tail striations are clear in the first picture, diminished in the second picture, and obscured in the third picture

Acknowledgments I thank Linda Nikolova and Willisa Liou for help with the thin-sectioning portion of this chapter and Brian Van Devener for help with the in situ liquid-cell and analytical EM portions. I thank my editors for their helpful suggestions. I thank those who provided samples for me to image and permission to use the images, and I thank those who provided images from their own work. Eddie B. Gilcrease and Sherwood R. Casjens provided samples of bacteriophages Utah and 9NA. Julianne H. Grose provided samples of bacteriophages Basilisk and T7. Edward H. Egelman provided an image of bacteriophage fd. Willisa Liou infected cells, prepared thin-

Detection of Bacteriophages: Electron Microscopy and Visualization

611

section specimens, and recorded the images used in Fig. 7c, d; Kelly T. Hughes and Christopher E. Wozniak provided Salmonella and phage P22 that Willisa used in her infection experiment. Matthew J. Domek and Brent Nelson provided halophage BN samples. Matthew Domek provided halophage CW02 samples. Lasha Gogokhia and June Round provided samples of phage NC-B and NC-G. Julie Thomas and Philip Serwer provided phage 201ϕ2-1 samples. Miika Leppänen and Ilari J. Maasilta provided an image of E. coli and bacteriophage T4. Jun Liu provided images of bacteriophage P1 and E. coli. Jochen Klumpp provided images of antibody-labeled phage A118. Melanie Ohi and Thomas Walz are thanked for an image of T7 helicase/primase. Paul Jardine and Dwight Anderson are thanked for an image of bacteriophages T2 and ϕ29. I thank Khim Karki, Jordan Moering, Madeline Dukes, and Sherwood Casjens for helpful suggestions.

References Ackermann H-W (2003) Bacteriophage observations and evolution. Res Microbiol 154:245–251 Ackermann H-W (2007) 5500 phages examined in the electron microscope. Arch Virol 152:227–243 Ackermann H-W (2009) Basic phage electron microscopy. Methods Mol Biol 501:113–126 Ackermann H-W (2011a) The first phage electron micrographs. Bacteriophage 1:225–227 Ackermann H-W (2011b) Pfankuch E, Kausche GA. Isolation and supra-microscopic representation of a bacteriophage. Naturwissenschaften 1940; 28:46. Bacteriophage 1:186–187 Ackermann H-W (2011c) Ruska H. Visualization of bacteriophage lysis in the hypermicroscope. Naturwissenschaften 1940; 28:45-6. Bacteriophage 1:183–185 Ackermann H-W (2012) Bacteriophage electron microscopy. Adv Virus Res 82:1–32 Ackermann H-W (2014) Sad state of phage electron microscopy. Please shoot the messenger. Microorganisms 2:1–10 Ackermann H-W, Nguyen T-M (1983) Sewage coliphages studied by electron microscopy. Appl Environ Microbiol 45:1049–1059 Ackermann H-W, Prangishvili D (2012) Prokaryote viruses studied by electron microscopy. Arch Virol 157:1843–1849 Ackermann H-W, Tiekotter KL (2012) Murphy’s law – if anything can go wrong, it will: problems in phage electron microscopy. Bacteriophage 2:122–129 Adrian M, Dubochet J, Fuller SD, Harris JR (1998) Cryo-negative staining. Micron 29:145–160 Aebi U, ten Heggeler B, Onorato L, Kistler J, Showe MK (1977) New method for localizing proteins in periodic structures: Fab fragment labeling combined with image processing of electron micrographs. Proc Natl Acad Sci U S A 74:5514–5518 Aksyuk AA, Leiman PG, Kurochkina LP, Shneider MM, Kostyuchenko VA, Mesyanzhinov VV, Rossmann MG (2009) The tail sheath structure of bacteriophage T4: a molecular machine for infecting bacteria. EMBO J 28:821–829 Aksyuk AA, Leiman PG, Kurochkina LP, Shneider MM, Kostyuchenko VA, Mesyanzhinov VV, Rossmann MG (2012) Corrigendum: the tail sheath structure of bacteriophage T4: a molecular machine for infecting bacteria. EMBO J 31:3507 Al-Amoudi A, Chang J-J, Leforestier A, McDowall A, Salamin LM, Norlén LPO, Richter K, Blanc NS, Studer D, Dubochet J (2004) Cryo-electron microscopy of vitreous sections. EMBO J 23:3583–3588 Almeida GM, Leppänen M, Maasilta IJ, Sundberg L-R (2018) Bacteriophage imaging: past, present and future. Res Microbiol 169:488–494 Anderson TF (1952) Stereoscopic studies of cells and viruses in the electron microscope. Am Nat 86:91–100 Anderson DL, Hickman DD, Reilly BE (1966) Structure of Bacillus subtilis bacteriophage ϕ29 and the length of ϕ29 deoxyribonucleic acid. J Bacteriol 91:2081–2089 Bamford DH, Lounatmaa K (1978) Freeze-fracturing of Pseudomonas phaseolicola infected by the lipid-containing bacteriophage φ6. J Gen Virol 39:161–170

612

D. M. Belnap

Bayer ME, Bayer MH (1986) Effects of bacteriophage fd infection on Escherichia coli HB11 envelope: a morphological and biochemical study. J Virol 57:258–266 Bayfield OW, Klimuk E, Winkler DC, Hesketh EL, Chechik M, Cheng N, Dykeman EC, Minakhin L, Ranson NA, Severinov K, Steven AC, Antson AA (2019) Cryo-EM structure and in vitro DNA packaging of a thermophilic virus with supersized T=7 capsids. Proc Natl Acad Sci U S A 116:3556–3561 Belnap DM (2015) Electron microscopy and image processing: essential tools for structural analysis of macromolecules. Curr Protoc Protein Sci 82:17.2.1–17.2.61 Bertin A, de Frutos M, Letellier L (2011) Bacteriophage–host interactions leading to genome internalization. Curr Opin Microbiol 14:492–496 Bielmann R, Habann M, Eugster MR, Lurz R, Calendar R, Klumpp J, Loessner MJ (2015) Receptor binding proteins of Listeria monocytogenes bacteriophages A118 and P35 recognize serovarspecific teichoic acids. Virology 477:110–118 Bozzola JJ (2014a) Conventional specimen preparation techniques for scanning electron microscopy of biological specimens. Methods Mol Biol 1117:133–150 Bozzola JJ (2014b) Conventional specimen preparation techniques for transmission electron microscopy of cultured cells. Methods Mol Biol 1117:1–19 Bozzola JJ, Russell LD (1999a) Electron microscopy: principles and techniques for biologists, 2nd edn. Jones and Bartlett Publishers, Sudbury Bozzola JJ, Russell LD (1999b) Specimen preparation for transmission electron microscopy. In: Electron microscopy: principles and techniques for biologists, 2nd edn. Jones and Bartlett Publishers, Sudbury, pp 16–71 Bozzola JJ, Russell LD (1999c) Specimen staining and contrast methods for transmission electron microscopy. In: Electron microscopy: principles and techniques for biologists, 2nd edn. Jones and Bartlett Publishers, Sudbury, pp 120–147 Bozzola JJ, Russell LD (1999d) Ultramicrotomy. In: Electron microscopy: principles and techniques for biologists, 2nd edn. Jones and Bartlett Publishers, Sudbury, pp 72–118 Bradley DE (1962) A study of the negative staining process. J Gen Microbiol 29:503–516 Bradley DE (1967) Ultrastructure of bacteriophages and bacteriocins. Bacteriol Rev 31:230–314 Bremer A, Henn C, Engel A, Baumeister W, Aebi U (1992) Has negative staining still a place in biomacromolecular electron microscopy? Ultramicroscopy 46:85–111 Brenner S, Horne RW (1959) A negative staining method for high resolution electron microscopy of viruses. Biochim Biophys Acta 34:103–110 Brenner S, Streisinger G, Horne RW, Champe SP, Barnett L, Benzer S, Rees MW (1959) Structural components of bacteriophage. J Mol Biol 1:281–292 Broers AN, Panessa BJ, Gennaro JF Jr (1975) High-resolution scanning electron microscopy of bacteriophages 3C and T4. Science 189:637–639 Cardone G, Duda RL, Cheng N, You L, Conway JF, Hendrix RW, Steven AC (2014) Metastable intermediates as stepping stones on the maturation pathways of viral capsids. mBio 5: e02067–14 Casjens SR, Leavitt JC, Hatfull GF, Hendrix RW (2014) Genome sequence of Salmonella phage 9NA. Genome Announc 2:e00531–14 Castón JR (2013) Conventional electron microscopy, cryo-electron microscopy and cryo-electron tomography of viruses. Subcell Biochem 68:79–115 Cerritelli ME, Wall JS, Simon MN, Conway JF, Steven AC (1996) Stoichiometry and domainal organization of the long tail-fiber of bacteriophage T4: a hinged viral adhesin. J Mol Biol 260:767–780 Cerritelli ME, Cheng N, Rosenberg AH, McPherson CE, Booy FP, Steven AC (1997) Encapsidated conformation of bacteriophage T7 DNA. Cell 91:271–280 Chaikeeratisak V, Nguyen K, Khanna K, Brilot AF, Erb ML, Coker JKC, Vavilina A, Newton GL, Buschauer R, Pogliano K, Villa E, Agard DA, Pogliano J (2017) Assembly of a nucleus-like structure during viral replication in bacteria. Science 355:194–197

Detection of Bacteriophages: Electron Microscopy and Visualization

613

Chandler DE (1986) Rotary shadowing with platinum-carbon in biological electron microscopy: a review of methods and applications. J Electron Microsc Tech 3:305–335 Chang JT, Schmid MF, Haase-Pettingell C, Weigele PR, King JA, Chiu W (2010) Visualizing the structural changes of bacteriophage epsilon15 and its Salmonella host during infection. J Mol Biol 402:731–740 Cheng N, Wu W, Watts NR, Steven AC (2014) Exploiting radiation damage to map proteins in nucleoprotein complexes: the internal structure of bacteriophage T7. J Struct Biol 185:250–256 Chlanda P, Sachse M (2014) Cryo-electron microscopy of vitreous sections. Methods Mol Biol 1117:193–214 Conway JF, Duda RL, Cheng N, Hendrix RW, Steven AC (1995) Proteolytic and conformational control of virus capsid maturation: the bacteriophage HK97 system. J Mol Biol 253:86–99 Conway JF, Wikoff WR, Cheng N, Duda RL, Hendrix RW, Johnson JE, Steven AC (2001) Virus maturation involving large subunit rotations and local refolding. Science 292:744–748 Crowther RA (2004) Viruses and the development of quantitative biological electron microscopy. IUBMB Life 56:239–248 Crowther RA, Klug A (1975) Structural analysis of macromolecular assemblies by image reconstruction from electron micrographs. Annu Rev Biochem 44:161–182 Dai W, Hodes A, Hui WH, Gingery M, Miller JF, Zhou ZH (2010) Three-dimensional structure of tropism-switching Bordetella bacteriophage. Proc Natl Acad Sci U S A 107:4347–4352 Dai W, Fu C, Raytcheva D, Flanagan J, Khant HA, Liu X, Rochat RH, Haase-Pettingell C, Piret J, Ludtke SJ, Nagayama K, Schmid MF, King JA, Chiu W (2013) Visualizing virus assembly intermediates inside marine cyanobacteria. Nature 502:707–710 Dai X, Li Z, Lai M, Shu S, Du Y, Zhou ZH, Sun R (2017) In situ structures of the genome and genome-delivery apparatus in a single-stranded RNA virus. Nature 541:112–116 Daum B, Quax TEF, Sachse M, Mills DJ, Reimann J, Yildiz Ö, Häder S, Saveanu C, Forterre P, Albers S-V, Kühlbrandt W, Prangishvili D (2014) Self-assembly of the general membraneremodeling protein PVAP into sevenfold virus-associated pyramids. Proc Natl Acad Sci U S A 111:3829–3834 De Carlo S, Harris JR (2011) Negative staining and cryo-negative staining of macromolecules and viruses for TEM. Micron 42:117–131 de Jonge N, Ross FM (2011) Electron microscopy of specimens in liquid. Nat Nanotechnol 6:695–704 de Jonge N, Peckys DB, Kremers GJ, Piston DW (2009) Electron microscopy of whole cells in liquid with nanometer resolution. Proc Natl Acad Sci U S A 106:2159–2164 DeRosier DJ, Klug A (1968) Reconstruction of three dimensional structures from electron micrographs. Nature 217:130–134 Dubochet J, Adrian M, Chang J-J, Homo J-C, Lepault J, McDowall AW, Schultz P (1988) Cryoelectron microscopy of vitrified specimens. Q Rev Biophys 21:129–228 Dukes MJ, Thomas R, Damiano J, Klein KL, Balasubramaniam S, Kayandan S, Riffle JS, Davis RM, McDonald SM, Kelly DF (2014) Improved microchip design and application for in situ transmission electron microscopy of macromolecules. Microsc Microanal 20:338–345 Effantin G, Hamasaki R, Kawasaki T, Bacia M, Moriscot C, Weissenhorn W, Yamada T, Schoehn G (2013) Cryo-electron microscopy three-dimensional structure of the jumbo phage FRSL1 infecting the phytopathogen Ralstonia solanacearum. Structure 21:298–305 Egelman EH (2010) Reconstruction of helical filaments and tubes. Methods Enzymol 482:167–183 Ellis EA (2014) Staining sectioned biological specimens for transmission electron microscopy: conventional and en bloc stains. Methods Mol Biol 1117:57–72 Evans JE, Browning ND (2013) Enabling direct nanoscale observations of biological reactions with dynamic TEM. Microscopy 62:147–156 Evans JE, Jungjohann KL, Wong PCK, Chiu P-L, Dutrow GH, Arslan I, Browning ND (2012) Visualizing macromolecular complexes with in situ liquid scanning transmission electron microscopy. Micron 43:1085–1090

614

D. M. Belnap

Frank J (2006) Three-dimensional electron microscopy of macromolecular assemblies: visualization of biological molecules in their native state, 2nd edn. Oxford University Press, New York Fraser D, Williams RC (1953) Details of frozen-dried T3 and T7 bacteriophages as shown by electron microscopy. J Bacteriol 65:167–170 Fromm SA, Sachse C (2016) Cryo-EM structure determination using segmented helical image reconstruction. Methods Enzymol 579:307–328 Frost LS, Bazett-Jones DP (1991) Examination of the phosphate in conjugative F-like pili by use of electron spectroscopic imaging. J Bacteriol 173:7728–7731 Fu C-y, Wang K, Gan L, Lanman J, Khayat R, Young MJ, Jensen GJ, Doerschuk PC, Johnson JE (2010) In vivo assembly of an archaeal virus studied with whole-cell electron cryotomography. Structure 18:1579–1586 Gambelli L, Cremers G, Mesman R, Guerrero S, Dutilh BE, Jetten MSM, Op den Camp HJM, van Niftrik L (2016) Ultrastructure and viral metagenome of bacteriophages from an anaerobic methane oxidizing Methylomirabilis bioreactor enrichment culture. Front Microbiol 7:1740 Gao Y, Cui Y, Fox T, Lin S, Wang H, de Val N, Zhou ZH, Yang W (2019) Structures and operating principles of the replisome. Science 363:eaav7003 Gilmore BL, Showalter SP, Dukes MJ, Tanner JR, Demmert AC, McDonald SM, Kelly DF (2013) Visualizing viral assemblies in a nanoscale biosphere. Lab Chip 13:216–219 Glaeser RM, Downing K, DeRosier D, Chiu W, Frank J (2007) Electron crystallography of biological macromolecules. Oxford University Press, New York Goddard TD, Huang CC, Ferrin TE (2007) Visualizing density maps with UCSF Chimera. J Struct Biol 157:281–287 Gogokhia L, Buhrke K, Bell R, Hoffman B, Brown DG, Hanke-Gogokhia C, Ajami NJ, Wong MC, Ghazaryan A, Valentine JF, Porter N, Martens E, O’Connell R, Jacob V, Scherl E, Crawford C, Stephens WZ, Casjens SR, Longman RS, Round JL (2019) Expansion of bacteriophages is linked to aggravated intestinal inflammation and colitis. Cell Host Microbe 25:285–299 Gorzelnik KV, Cui Z, Reed CA, Jakana J, Young R, Zhang J (2016) Asymmetric cryo-EM structure of the canonical Allolevivirus Qβ reveals a single maturation protein and the genomic ssRNA in situ. Proc Natl Acad Sci U S A 113:11519–11524 Gowen B, Bamford JKH, Bamford DH, Fuller SD (2003) The tailless icosahedral membrane virus PRD1 localizes the proteins involved in genome packaging and injection at a unique vertex. J Virol 77:7863–7871 Grassucci RA, Taylor DJ, Frank J (2007) Preparation of macromolecular complexes for cryoelectron microscopy. Nat Protoc 2:3239–3246 Grose JH, Belnap DM, Jensen JD, Mathis AD, Prince JT, Burnett SH, Breakwell DP (2014) The genomes, proteomes, and structure of three novel phages that infect the Bacillus cereus group and carry putative virulence factors. J Virol 88:11846–11860 Guerrero-Ferreira RC, Wright ER (2013) Cryo-electron tomography of bacterial viruses. Virology 435:179–186 Guerrero-Ferreira RC, Viollier PH, Ely B, Poindexter JS, Georgieva M, Jensen GJ, Wright ER (2011) Alternative mechanism for bacteriophage adsorption to the motile bacterium Caulobacter crescentus. Proc Natl Acad Sci U S A 108:9963–9968 Guo F, Jiang W (2014) Single particle cryo-electron microscopy and 3-D reconstruction of viruses. Methods Mol Biol 1117:401–443 Happonen LJ, Redder P, Peng X, Reigstad LJ, Prangishvili D, Butcher SJ (2010) Familial relationships in hyperthermo- and acidophilic archaeal viruses. J Virol 84:4747–4754 Häring M, Rachel R, Peng X, Garrett RA, Prangishvili D (2005) Viral diversity in hot springs of Pozzuoli, Italy, and characterization of a unique archaeal virus, Acidianus bottle-shaped virus, from a new family, the Ampullaviridae. J Virol 79:9904–9911 Harris JR (1997) Negative staining and cryoelectron microscopy: the thin film techniques. BIOS Scientific Publishers, Oxford Harris JR (2015) Transmission electron microscopy in molecular structural biology: a historical survey. Arch Biochem Biophys 581:3–18

Detection of Bacteriophages: Electron Microscopy and Visualization

615

Harris JR, De Carlo S (2014) Negative staining and cryo-negative staining: applications in biology and medicine. Methods Mol Biol 1117:215–258 Harris JR, Gebauer W, Markl J (1995) Keyhole limpet haemocyanin: negative staining in the presence of trehalose. Micron 26:25–33 Harris JR, Schröder E, Isupov MN, Scheffler D, Kristensen P, Littlechild JA, Vagin AA, Meissner U (2001) Comparison of the decameric structure of peroxiredoxin-II by transmission electron microscopy and X-ray crystallography. Biochim Biophys Acta 1547:221–234 Hart JL, Lang AC, Leff AC, Longo P, Trevor C, Twesten RD, Taheri ML (2017) Direct detection electron energy-loss spectroscopy: a method to push the limits of resolution and sensitivity. Sci Rep 7:8243 Hartman R, Munson-McGee J, Young MJ, Lawrence CM (2019) Survey of high-resolution archaeal virus structures. Curr Opin Virol 36:74–83 Hawkes PW, Valdrè U (eds) (1990) Biophysical electron microscopy: basic concepts and modern techniques. Academic Press, London Hayat MA (1986) Basic techniques for transmission electron microscopy. Academic Press, Orlando Hayat MA, Miller SE (1990) Negative staining. McGraw-Hill, New York He W, He Y (2014) Electron tomography for organelles, cells, and tissues. Methods Mol Biol 1117:445–483 Hendricks GM (2014) Metal shadowing for electron microscopy. Methods Mol Biol 1117:73–93 Hermann R, Müller M (1991) High resolution biological scanning electron microscopy: a comparative study of low temperature metal coating techniques. J Electron Microsc Tech 18:440–449 Hermann R, Schwartz H, Müller M (1991) High precision immunoscanning electron microscopy using Fab fragments coupled to ultra-small colloidal gold. J Struct Biol 107:38–47 Heymann JB, Chagoyen M, Belnap DM (2005) Common conventions for interchange and archiving of three-dimensional electron microscopy information in structural biology. J Struct Biol 151:196–207 Heymann JB, Chagoyen M, Belnap DM (2006) Corrigendum to “common conventions for interchange and archiving of three-dimensional electron microscopy information in structural biology” [J. Struct. Biol. 151 (2005) 196–207]. J Struct Biol 153:312 Hochstein R, Bollschweiler D, Dharmavaram S, Lintner NG, Plitzko JM, Bruinsma R, Engelhardt H, Young MJ, Klug WS, Lawrence CM (2018) Structural studies of Acidianus tailed spindle virus reveal a structural paradigm used in the assembly of spindle-shaped viruses. Proc Natl Acad Sci U S A 115:2120–2125 Hong C, Pietilä MK, Fu CJ, Schmid MF, Bamford DH, Chiu W (2015) Lemon-shaped halo archaeal virus His1 with uniform tail but variable capsid structure. Proc Natl Acad Sci U S A 112:2449–2454 Hoppe SM, Sasaki DY, Kinghorn AN, Hattar K (2013) In-situ transmission electron microscopy of liposomes in an aqueous environment. Langmuir 29:9958–9961 Hrebík D, Štveráková D, Škubník K, Füzik T, Pantůček R, Plevka P (2019) Structure and genome ejection mechanism of Staphylococcus aureus phage P68. Sci Adv 5:eaaw7414 Hryc CF, Chen D-H, Afonine PV, Jakana J, Wang Z, Haase-Pettingell C, Jiang W, Adams PD, King JA, Schmid MF, Chiu W (2017) Accurate model annotation of a near-atomic resolution cryo-EM map. Proc Natl Acad Sci U S A 114:3103–3108 Hu B, Margolin W, Molineux IJ, Liu J (2013) The bacteriophage T7 virion undergoes extensive structural remodeling during infection. Science 339:576–579 Hu B, Margolin W, Molineux IJ, Liu J (2015) Structural remodeling of bacteriophage T4 and host membranes during infection initiation. Proc Natl Acad Sci U S A 112:E4919–E4928 Ionel A, Velázquez-Muriel JA, Luque D, Cuervo A, Castón JR, Valpuesta JM, Martín-Benito J, Carrascosa JL (2011) Molecular rearrangements involved in the capsid shell maturation of bacteriophage T7. J Biol Chem 286:234–242 Joens MS, Huynh C, Kasuboski JM, Ferranti D, Sigal YJ, Zeitvogel F, Obst M, Burkhardt CJ, Curran KP, Chalasani SH, Stern LA, Goetze B, Fitzpatrick JAJ (2013) Helium ion microscopy (HIM) for the imaging of biological samples at sub-nanometer resolution. Sci Rep 3:3514

616

D. M. Belnap

Kay D, Bradley DE (1962) The structure of bacteriophage ϕR. J Gen Microbiol 27:195–200 Kellenberger E, Edgar RS (1971) Structure and assembly of phage particles. In: Hershey AD (ed) The bacteriophage lambda. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp 271–295 Kellenberger E, Eiserling FA, Boy de la Tour E (1968) Studies on the morphopoiesis of the head of phage T-even III. The cores of head-related structures. J Ultrastruct Res 21:335–360 Keller B, Dubochet J, Adrian M, Maeder M, Wurtz M, Kellenberger E (1988) Length and shape variants of the bacteriophage T4 head: mutations in the scaffolding core genes 68 and 22. J Virol 62:2960–2969 Kennedy E, Nelson EM, Tanaka T, Damiano J, Timp G (2016) Live bacterial physiology visualized with 5 nm resolution using scanning transmission electron microscopy. ACS Nano 10:2669–2677 Kistler J, Aebi U, Onorato L, ten Heggeler B, Showe MK (1978) Structural changes during the transformation of bacteriophage T4 polyheads: characterization of the initial and final states by freeze-drying and shadowing Fab-fragment-labelled preparations. J Mol Biol 126:571–589 Kleinschmidt AK, Lang D, Jacherts D, Zahn RK (1962) Darstellung und längenmessungen des gesamten desoxyribonucleinsäure-inhaltes von T2-bakteriophagen. Biochim Biophys Acta 61:857–864 Kocsis E, Greenstone HL, Locke EG, Kessel M, Steven AC (1997) Multiple conformational states of the bacteriophage T4 capsid surface lattice induced when expansion occurs without prior cleavage. J Struct Biol 118:73–82 Koning RI, Gomez-Blanco J, Akopjana I, Vargas J, Kazaks A, Tars K, Carazo JM, Koster AJ (2016) Asymmetric cryo-EM reconstruction of phage MS2 reveals genome structure in situ. Nat Commun 7:12524 Kruger DH, Schneck P, Gelderblom HR (2000) Helmut Ruska and the visualisation of viruses. Lancet 355:1713–1717 Labrie SJ, Tremblay DM, Moisan M, Villion M, Magadán AH, Campanacci V, Cambillau C, Moineau S (2012) Involvement of the major capsid protein and two early-expressed phage genes in the activity of the lactococcal abortive infection mechanism AbiT. Appl Environ Microbiol 78:6890–6899 Lander GC, Tang L, Casjens SR, Gilcrease EB, Prevelige P, Poliakov A, Potter CS, Carragher B, Johnson JE (2006) The structure of an infectious P22 virion shows the signal for headful DNA packaging. Science 312:1791–1795 Leavitt JC, Heitkamp AJ, Bhattacharjee AS, Gilcrease EB, Casjens SR (2017) Genome sequence of Escherichia coli tailed phage Utah. Genome Announc 5:e01494–16 Lenk E, Casjens S, Weeks J, King J (1975) Intracellular visualization of precursor capsids in phage P22 mutant infected cells. Virology 68:182–199 Lepault J, Dubochet J, Baschong W, Kellenberger E (1987) Organization of double-stranded DNA in bacteriophages: a study by cryo-electron microscopy of vitrified samples. EMBO J 6:1507–1512 Leppänen M, Sundberg L-R, Laanto E, Almeida GMdF, Papponen P, Maasilta IJ (2017) Imaging bacterial colonies and phage–bacterium interaction at sub-nanometer resolution using heliumion microscopy. Adv Biosys 1:1700070 Lin J, Cheng N, Hogle JM, Steven AC, Belnap DM (2013) Conformational shift of a major poliovirus antigen confirmed by immuno-cryogenic electron microscopy. J Immunol 191:884–891 Liou W, Geuze HJ, Slot JW (1996) Improving structural integrity of cryosections for immunogold labeling. Histochem Cell Biol 106:41–58 Liou W, Sung Y-J, Tao M-H, Lo SJ (2008) Morphogenesis of the hepatitis B virion and subviral particles in the liver of transgenic mice. J Biomed Sci 15:311–316 Liu J, Chen C-Y, Shiomi D, Niki H, Margolin W (2011) Visualization of bacteriophage P1 infection by cryo-electron tomography of tiny Escherichia coli. Virology 417:304–311

Detection of Bacteriophages: Electron Microscopy and Visualization

617

Luria SE, Anderson TF (1942) The identification and characterization of bacteriophages with the electron microscope. Proc Natl Acad Sci U S A 28:127–130 McNulty R, Cardone G, Gilcrease EB, Baker TS, Casjens SR, Johnson JE (2018) Cryo-EM elucidation of the structure of bacteriophage P22 virions after genome release. Biophys J 114:1295–1301 Messaoudi C, Boudier T, Lechaire J-P, Rigaud J-L, Delacroix H, Gaill F, Marco S (2003) Use of cryo-negative staining in tomographic reconstruction of biological objects: application to T4 bacteriophage. Biol Cell 95:393–398 Mielanczyk L, Matysiak N, Michalski M, Buldak R, Wojnicz R (2014) Closer to the native state. Critical evaluation of cryo-techniques for transmission electron microscopy: preparation of biological samples. Folia Histochem Cytobiol 52:1–17 Mondal SI, Islam MR, Sawaguchi A, Asadulghani M, Ooka T, Gotoh Y, Kasahara Y, Ogura Y, Hayashi T (2016) Genes essential for the morphogenesis of the Shiga toxin 2-transducing phage from Escherichia coli O157:H7. Sci Rep 6:39036 Müller M, Engel A, Aebi U (1994) Structural and physicochemical analysis of the contractile MM phage tail and comparison with the bacteriophage T4 tail. J Struct Biol 112:11–31 Nannenga BL, Gonen T (2019) The cryo-EM method microcrystal electron diffraction (MicroED). Nat Methods 16:369–379 Nevsten P, Evilevitch A, Wallenberg R (2012) Chemical mapping of DNA and counter-ion content inside phage by energy-filtered TEM. J Biol Phys 38:229–240 Obr M, Schur FKM (2019) Structural analysis of pleomorphic and asymmetric viruses using cryoelectron tomography and subtomogram averaging. Adv Virus Res 105:117–159 Ohi M, Li Y, Cheng Y, Walz T (2004) Negative staining and image classification – powerful tools in modern electron microscopy. Biol Proced Online 6:23–34 Orlova EV, Gowen B, Dröge A, Stiege A, Weise F, Lurz R, van Heel M, Tavares P (2003) Structure of a viral DNA gatekeeper at 10 Å resolution by cryo-electron microscopy. EMBO J 22:1255–1262 Ortega DR, Oikonomou CM, Ding HJ, Rees-Lee P, Alexandria, Jensen GJ (2019) ETDB-Caltech: a blockchain-based distributed public database for electron tomography. PLoS One 14:e0215531 Özel M, Pauli G, Gelderblom HR (1990) Electron spectroscopic imaging (ESI) of viruses using thin-section and immunolabelling preparations. Ultramicroscopy 32:35–41 Parent KN, Khayat R, Tu LH, Suhanovsky MM, Cortines JR, Teschke CM, Johnson JE, Baker TS (2010a) P22 coat protein structures reveal a novel mechanism for capsid maturation: stability without auxiliary proteins or chemical crosslinks. Structure 18:390–401 Parent KN, Sinkovits RS, Suhanovsky MM, Teschke CM, Egelman EH, Baker TS (2010b) Cryoreconstructions of P22 polyheads suggest that phage assembly is nucleated by trimeric interactions among coat proteins. Phys Biol 7:045004 Parent KN, Gilcrease EB, Casjens SR, Baker TS (2012) Structural evolution of the P22-like phages: comparison of Sf6 and P22 procapsid and virion architectures. Virology 427:177–188 Parent LR, Bakalis E, Ramírez-Hernańdez A, Kammeyer JK, Park C, de Pablo J, Zerbetto F, Patterson JP, Gianneschi NC (2017) Directly observing micelle fusion and growth in solution by liquid-cell transmission electron microscopy. J Am Chem Soc 139:17140–17151 Parent KN, Schrad JR, Cingolani G (2018) Breaking symmetry in viral icosahedral capsids as seen through the lenses of X-ray crystallography and cryo-electron microscopy. Viruses 10:67 Park K, Debyser Z, Tabor S, Richardson CC, Griffith JD (1998) Formation of a DNA loop at the replication fork generated by bacteriophage T7 replication proteins. J Biol Chem 273:5260–5270 Peralta B, Gil-Carton D, Castaño-Díez D, Bertin A, Boulogne C, Oksanen HM, Bamford DH, Abrescia NGA (2013) Mechanism of membranous tunnelling nanotube formation in viral genome delivery. PLoS Biol 11:e1001667 Pfankuch E, Kausche GA (1940) Isolierung und übermikroskopische abbildung eines bakteriophagen. Naturwissenschaften 28:46

618

D. M. Belnap

Pietilä MK, Demina TA, Atanasova NS, Oksanen HM, Bamford DH (2014) Archaeal viruses and bacteriophages: comparisons and contrasts. Trends Microbiol 22:334–344 Plançon L, Chami M, Letellier L (1997) Reconstitution of FhuA, an Escherichia coli outer membrane protein, into liposomes. J Biol Chem 272:16868–16872 Popenko VI, Kutter EM, Ackermann H-W (2013) Anna S. Tikhonenko. Bacteriophage 3:e23646 Prasad BVV, Burns JW, Marietta E, Estes MK, Chiu W (1990) Localization of VP4 neutralization sites in rotavirus by three-dimensional cryo-electron microscopy. Nature 343:476–479 Preux O, Durand D, Huet A, Conway JF, Bertin A, Boulogne C, Drouin-Wahbi J, Trévarin D, Pérez J, Vachette P, Boulanger P (2013) A two-state cooperative expansion converts the procapsid shell of bacteriophage T5 into a highly stable capsid isomorphous to the final virion head. J Mol Biol 425:1999–2014 Rachel R, Bettstetter M, Hedlund BP, Häring M, Kessler A, Stetter KO, Prangishvili D (2002) Remarkable morphological diversity of viruses and virus-like particles in hot terrestrial environments. Arch Virol 147:2419–2429 Raytcheva DA, Haase-Pettingell C, Piret J, King JA (2014) Two novel proteins of cyanophage Syn5 compose its unusual horn structure. J Virol 88:2047–2055 Revet B, Zarling DA, Jovin TM, Delain E (1984) Different Z DNA forming sequences are revealed in ϕX174 RFI by high resolution darkfield immuno-electron microscopy. EMBO J 3:3353–3358 Ruska H (1940) Die sichtbarmachung der bakteriophagen lyse im übermikroskop. Naturwissenschaften 28:45–46 Sachse C (2015) Single-particle based helical reconstruction – how to make the most of real and Fourier space. AIMS Biophys 2:219–244 Saigo K (1975) Denaturation mapping and chromosome structure in bacteriophage T5. Virology 68:166–172 Scheres SHW (2012) RELION: implementation of a Bayesian approach to cryo-EM structure determination. J Struct Biol 180:519–530 Scheres SHW, Gao H, Valle M, Herman GT, Eggermont PPB, Frank J, Carazo J-M (2007) Disentangling conformational states of macromolecules in 3D-EM through likelihood optimization. Nat Methods 4:27–29 Schur FKM (2019) Toward high-resolution in situ structural biology with cryo-electron tomography and subtomogram averaging. Curr Opin Struct Biol 58:1–9 Severs NJ (2007) Freeze-fracture electron microscopy. Nat Protoc 2:547–576 Shen PS, Domek MJ, Sanz-García E, Makaju A, Taylor RM, Hoggan R, Culumber M, Oberg C, Breakwell DP, Prince JT, Belnap DM (2012) Sequence and structural characterization of Great Salt Lake bacteriophage CW02, a member of the T7-like supergroup. J Virol 86:7907–7917 Sigworth FJ (2016) Principles of cryo-EM single-particle image processing. Microscopy 65:57–67 Simon LD (1972) Infection of Escherichia coli by T2 and T4 bacteriophages as seen in the electron microscope: T4 head morphogenesis. Proc Natl Acad Sci U S A 69:907–911 Spilman MS, Dearborn AD, Chang JR, Damle PK, Christie GE, Dokland T (2011) A conformational switch involved in maturation of Staphylococcus aureus bacteriophage 80α capsids. J Mol Biol 405:863–876 Steven AC, Navia MA (1980) Fidelity of structure representation in electron micrographs of negatively stained protein molecules. Proc Natl Acad Sci U S A 77:4721–4725 Steven AC, Bauer AC, Bisher ME, Robey FA, Black LW (1991) The maturation-dependent conformational change of phage T4 capsid involves the translocation of specific epitopes between the inner and the outer capsid surfaces. J Struct Biol 106:221–236 Stoops JK, Baker TS, Schroeter JP, Kolodziej SJ, Niu X-D, Reed LJ (1992) Three-dimensional structure of the truncated core of the Saccharomyces cerevisiae pyruvate dehydrogenase complex determined from negative stain and cryoelectron microscopy images. J Biol Chem 267:24769–24775 Tarahovsky YS, Khusainov AA, Deev AA, Kim YV (1991) Membrane fusion during infection of Escherichia coli cells by phage T4. FEBS Lett 289:18–22

Detection of Bacteriophages: Electron Microscopy and Visualization

619

Teschke CM, Parent KN (2010) ‘Let the phage do the work’: using the phage P22 coat protein structures as a framework to understand its folding and assembly mutants. Virology 401:119–130 Thomas D, Schultz P, Steven AC, Wall JS (1994) Mass analysis of biological macromolecular complexes by STEM. Biol Cell 80:181–192 Thomas JA, Rolando MR, Carroll CA, Shen PS, Belnap DM, Weintraub ST, Serwer P, Hardies SC (2008) Characterization of Pseudomonas chlororaphis myovirus 201ϕ2-1 via genomic sequencing, mass spectrometry, and electron microscopy. Virology 376:330–338 Tiekotter KL, Ackermann H-W (2009) High-quality virus images obtained by transmission electron microscopy and charge coupled device digital camera technology. J Virol Methods 159:87–92 Tikhonenko AS (1970) Ultrastructure of Bacterial Viruses (trans: Haigh B). Plenum Press, New York Tokuyasu KT (1973) A technique for ultracryotomy of cell suspensions and tissues. J Cell Biol 57:551–565 Varano AC, Rahimi A, Dukes MJ, Poelzing S, McDonald SM, Kelly DF (2015) Visualizing virus particle mobility in liquid at the nanoscale. Chem Commun 51:16176–16179 Veesler D, Ng T-S, Sendamarai AK, Eilers BJ, Lawrence CM, Lok S-M, Young MJ, Johnson JE, Fu C-y (2013) Atomic structure of the 75 MDa extremophile Sulfolobus turreted icosahedral virus determined by CryoEM and X-ray crystallography. Proc Natl Acad Sci U S A 110:5504–5509 Villa E, Schaffer M, Plitzko JM, Baumeister W (2013) Opening windows into the cell: focused-ionbeam milling for cryo-electron tomography. Curr Opin Struct Biol 23:771–777 Wall JS, Hainfeld JF (1986) Mass mapping with the scanning transmission electron microscope. Annu Rev Biophys Biophys Chem 15:355–376 Wan W, Briggs JAG (2016) Cryo-electron tomography and subtomogram averaging. Methods Enzymol 579:329–367 Wang Y, Zhang X (2008) Characterization of a novel portal protein from deep-sea thermophilic bacteriophage GVE2. Gene 421:61–66 Wang YA, Yu X, Overman S, Tsuboi M, Thomas GJ Jr, Egelman EH (2006) The structure of a filamentous bacteriophage. J Mol Biol 361:209–215 Wang C, Tu J, Liu J, Molineux IJ (2019) Structural dynamics of bacteriophage P22 infection initiation revealed by cryo-electron tomography. Nat Microbiol 4:1049–1056 Watts NRM, Hainfeld J, Coombs DH (1990) Localization of the proteins gp7, gp8 and gp10 in the bacteriophage T4 baseplate with colloidal gold:F(ab)2 and undecagold: Fab’ conjugates. J Mol Biol 216:315–325 Wendelschafer-Crabb G, Erlandsen SL, Walker DH Jr (1975) Conditions critical for optimal visualization of bacteriophage adsorbed to bacterial surfaces by scanning electron microscopy. J Virol 15:1498–1503 Williams RC, Fraser D (1953) Morphology of the seven T-bacteriophages. J Bacteriol 66:458–464 Wollin R, Eriksson U, Lindberg AA (1981) Salmonella bacteriophage glycanases: endorhamnosidase activity of bacteriophages P27, 9NA, and KB1. J Virol 38:1025–1033 Wu S, Liu B, Zhang X (2009) Identification of a tail assembly gene cluster from deep-sea thermophilic bacteriophage GVE2. Virus Genes 38:507–514 Wu W, Thomas JA, Cheng N, Black LW, Steven AC (2012) Bubblegrams reveal the inner body of bacteriophage ϕKZ. Science 335:182 Wu J, Shan H, Chen W, Gu X, Tao P, Song C, Shang W, Deng T (2016a) In situ environmental TEM in imaging gas and liquid phase chemical reactions for materials research. Adv Mater 28:9686–9712 Wu W, Leavitt JC, Cheng N, Gilcrease EB, Motwani T, Teschke CM, Casjens SR, Steven AC (2016b) Localization of the Houdinisome (ejection proteins) inside the bacteriophage P22 virion by bubblegram imaging. mBio 7:e01152–16 Wurtz M (1992) Bacteriophage structure. Electron Microsc Rev 5:283–309 Wyckoff RWG (1948) The electron microscopy of developing bacteriophage: II. Growth of T4 in liquid culture. Biochim Biophys Acta 2:246–253

620

D. M. Belnap

Xu J, Dayan N, Goldbourt A, Xiang Y (2019) Cryo-electron microscopy structure of the filamentous bacteriophage IKe. Proc Natl Acad Sci U S A 116:5493–5498 Yap ML, Klose T, Arisaka F, Speir JA, Veesler D, Fokine A, Rossmann MG (2016) Role of bacteriophage T4 baseplate in regulating assembly and infection. Proc Natl Acad Sci U S A 113:2654–2659 Yu G, Vago F, Zhang D, Snyder JE, Yan R, Zhang C, Benjamin C, Jiang X, Kuhn RJ, Serwer P, Thompson DH, Jiang W (2014) Single-step antibody-based affinity cryo-electron microscopy for imaging and structural analysis of macromolecular assemblies. J Struct Biol 187:1–9 Yu G, Li K, Jiang W (2016) Antibody-based affinity cryo-EM grid. Methods 100:16–24 Zhao H, Li K, Lynn AY, Aron KE, Yu G, Jiang W, Tang L (2017) Structure of a headful DNA-packaging bacterial virus at 2.9 Å resolution by electron cryo-microscopy. Proc Natl Acad Sci U S A 114:3601–3606 Zheng W, Wang F, Taylor NMI, Guerrero-Ferreira RC, Leiman PG, Egelman EH (2017) Refined cryo-EM structure of the T4 tail tube: exploring the lowest dose limit. Structure 25:1436–1441

Detection of Bacteriophages: Sequence-Based Systems Si^an V. Owen, Blanca M. Perez-Sepulveda, and Evelien M. Adriaenssens

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview of Sequencing Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Short-Read Platforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Long-Read Platforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A Note on Sequence Data and Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Applications of Sequencing-Based Detection Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gene-Based Detection of Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genome Sequencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sequence-Based Identification of Prophages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metagenomics-Based Detection of Bacteriophages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

622 624 624 625 628 629 630 632 633 635 637 638 638

Author contributed equally with all other contributors. Si^an V. Owen and Blanca Perez-Sepulveda S. V. Owen Microbiology Research Group, Institute of Integrative Biology, University of Liverpool, Liverpool, UK Department of Biomedical Informatics and Laboratory of Systems Pharmacology, Harvard Medical School, Boston, MA, USA e-mail: [email protected] B. M. Perez-Sepulveda · E. M. Adriaenssens (*) Microbiology Research Group, Institute of Integrative Biology, University of Liverpool, Liverpool, UK e-mail: [email protected]; [email protected]; [email protected]; [email protected] © Springer Nature Switzerland AG 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_19

621

622

S. V. Owen et al.

Abstract

The invention of sequencing technologies has fundamentally changed molecular biology, including the way we look at bacteriophages. In addition to investigating bacteriophage plaques, electron micrographs, or the phenotypes of different mutants, we are now able to explore the entire genetic potential encoded in their genomes. As of July 2018, over 6,000 complete phage genome sequences were published in public databases, with over 28,000 partial sequences available. In this chapter, we give an overview of the latest technologies that can be used to determine phage genome sequences, ranging from short-read platforms which generally give multiple gigabases of sequence data to long-read technologies which have the potential to sequence a bacteriophage genome in one single read. We then look at applications of sequencing technologies in detecting bacteriophages, from a single gene, over entire genomes, to the community level.

Introduction The recent advances in sequencing technologies have profoundly changed our understanding of bacteriophage biology and genomics. Phage genomics, the study of the nucleic acid content or full gene complement of a phage, has demonstrated the remarkable genetic diversity of the phage world. This diversity is so high that phages with similar morphologies can have no nucleotide sequence similarity and very limited amino acid similarities of orthologous proteins (Krupovic et al. 2011). As a result, sequencing of specific phage marker genes or the entire genome is more accurate in detecting the exact variant or isolate of a phage than electron microscopic- or host range-based methods. Sequencing is becoming increasingly important in phage research. For example, phage taxonomy has started to move away from morphology-based classification based on the presence/absence, type, and length of the tail toward a genomebased strategy more accurately reflecting the full diversity present in the virosphere (Ackermann 2011; Adriaenssens and Brister 2017; Adriaenssens et al. 2018). In phage therapy – the use of phages as therapeutic agents to control bacterial infections – sequencing of the entire genome is encouraged to ensure that the phages in question do not encode potentially damaging genes (e.g., toxins, lysogeny-related genes, antibiotic resistance genes) (Alavidze et al. 2016; Abedon 2017). Sequencing has also provided an extra dimension to viral ecology, offering a greater resolution of their diversity and unearthing viral dark matter, i.e., viral/ phage sequence fragments and genes without any database homologs (Filée et al. 2005; Hatfull 2015). In the first part of this chapter, we give an overview of available sequencing technologies for phage research, made up of short-read platforms and long-read platforms (Table 1). For these platforms, we review the different methodological steps and discuss some of the advantages and disadvantages of each technology. While the information presented here was up-to-date to the time of writing,

Detection of Bacteriophages: Sequence-Based Systems

623

Table 1 Overview of technologies used in phage sequencing

a

Sequencing technology Sanger

Read length (bp) 800–1,000 single read

Illumina MiSeq

150–350 paired-end reads

Sequencing by synthesis, with detection of fluorescently labeled nucleotides

Ion Torrent PGM

300–400 paired reads

Detection of protons released during DNA synthesis

PacBio

Up to 60,000 single reads

Real-time detection of labeled nucleotide incorporation by a tethered DNA polymerase

Oxford Nanopore

Up to 1,000,000 single reads

Real-time detection of nucleotide-mediated current changes as nucleic acid molecule passes through a tethered nanopore protein

Characteristics Capillary electrophoresis of labeled chain-terminating dideoxynucleotides

Pros/consa High accuracy and resolution at the nucleotide level Low throughput High sequencing cost per base Low amount of starting material (ng) allows the sequencing of bacteriophages from single plaques PCR-free library preparation reduces the introduction of sequencing errors Widely used, allowing access in most laboratories at a low cost per read Nextera-based library preparation will not be able to sequence phage genome ends Ability to detect more variants compared to Illumina MiSeq, but with more false positives Short sequencing run time Long-read lengths allow repeat region resolution Additional data on DNA-chemical modifications is generated Poor accuracy at single read level High amount of starting material required (μg) Large hardware cost, large machine Long-read lengths allow repeat region resolution Additional data on DNA-chemical modifications can be generated Base calling in real time permits instant sequence interrogation (i.e., results are available as they are generated, not only once the machine has finished) Low hardware cost, portable machine allows for anytimeanywhere sequencing Only technology that allows direct sequencing of RNA molecules Poor accuracy at single read level High amount of starting material required (ug)

Sequencing yield was not considered as a factor because of small genome sizes of phage

624

S. V. Owen et al.

these technologies evolve rapidly, and it is highly likely that new platforms or optimized technological approaches will come to the forefront in the next few years. The second part of the chapter deals with the application of sequencing technology in the detection of bacteriophages. We have divided these applications according to level of complexity, from detecting a single gene of a phage, over sequencing the complete genome, to detection of prophages in bacterial genomes, and even to simultaneous detection of entire communities of phages.

Overview of Sequencing Methods Short-Read Platforms Historically, bacteriophage sequencing has been performed using Sanger technology. This sequencing technology uses capillary electrophoresis to detect precisely the selective incorporation of labeled chain-terminating dideoxynucleotides during in vitro replication of DNA. However, this approach can be expensive, depends in many cases on cloning the phage DNA, and does not allow separation of contaminating host DNA sequence (Klumpp et al. 2012). High-throughput (or “next-generation”) short-read sequencing has become the preferred method for sequencing phages, mostly due to the low cost per base and high output compared to other platforms (Rihtman et al. 2016). Currently, popular platforms for sequencing bacteriophages include Illumina ® MiSeq and Ion Torrent Personal Genome Machine (Ion PGMTM). These sequencing platforms differ mostly in their chemistry but are generally based on the same three steps: library preparation, amplification, and sequencing of dsDNA. For RNA phages, cDNA needs to be generated by reverse transcription, either before the library preparation step or by using library preparation kits that are optimized for RNA sequencing and can provide information on whether the genome is made up of dsRNA or ssRNA and the sense of the single-stranded RNA. During library preparation, the extracted phage DNA is quality checked before fragmentation into random overlapping parts, which can be done by either mechanical (e.g., sonication) or chemical methods (e.g., the nuclease fragmentase (New England Biolabs)). The amount of input DNA for library preparation is a crucial step, especially in the study of phages, where limiting factors are either phage propagation to large numbers of particles or varying genome sizes. Several kits (e.g., Nextera XT (Illumina)) have been designed to significantly decrease the required amount of input DNA (to 1 ng) and reduce hands-on time by using “tagmentation,” a process that fragments, size-selects, and tags the input DNA. These new kits increase sequencing efficiency as they produce greater output data in a shorter overall time (Marine et al. 2011). However, the use of Nextera kits comes with the trade-off that the defined ends of linear phage genomes will be missed and is, therefore, not recommended when the aim of sequencing is to generate a highquality reference genome (Kot et al. 2014).

Detection of Bacteriophages: Sequence-Based Systems

625

After fragmentation, the pieces of DNA are size-selected and ligated with adaptors specific for each platform. During the following amplification step, the DNA is copied several times in its specific position, creating reaction centers or clusters that allow the sequencer to distinguish the input DNA from background noise. The amplification step is carried out by annealing to complementary adapters attached either to beads in micelle droplets (emulsion PCR; e.g., 454 pyrosequencing and Ion Torrent), to solid plates (e.g., Illumina sequencing), or by creating nanoballs that are then placed in a flow cell (Complete Genomics (BGI)). The most notable differences between platforms can be observed during the sequencing step, which can be by synthesis or instead by ligation. Platforms based on sequencing by synthesis include Illumina (formerly Solexa) which detects fluorescently labeled nucleotides; Ion Torrent, which detects a change in pH; and 454 pyrosequencing, which senses the amount of light generated due to pyrophosphate release during nucleotide incorporation. Alternatively, SOLiD is a platform based on sequencing by ligation of a labeled probe to the target DNA (Goodwin et al. 2016). The development of high-throughput sequencing technologies decreased sequencing costs and time compared to construction of clone libraries (Loman et al. 2012). The 454 pyrosequencing platform, the first platform to be optimized for high-throughput phage sequencing, has been recently discontinued (Henn et al. 2010; Marine et al. 2011). Currently, the most widely used platform for phage sequencing is Illumina technology, with its benchtop sequencer, MiSeq, particularly popular for phages, and the high-throughput machine HiSeq used for bacterial, eukaryotic, and metagenomes. These machines can provide large amounts of highquality sequencing data in a reduced time when compared to other technologies. Due to the small size of phage genomes relative to other organisms, and the large amount of data generated, short-read platforms, especially Illumina, thus far are the preferred method for bacteriophage metagenomics, genome sequencing, and re-sequencing or, most recently, to detect termini and packaging mechanisms (Garneau et al. 2017).

Long-Read Platforms Long-read sequencing represents the most recent and transformative development in sequencing technology, and these technologies are collectively referred to as thirdgeneration sequencing (TGS). Whereas short-read sequencing platforms may generate sequence reads of up to 1 kb, the capability of long-read sequencing platforms is now approaching read lengths of 1 Mb (which is approximately three times longer than the longest known phage genome). Unlike short-read sequencing, which has become dominated by a single platform, two platforms utilizing very different technologies are used for long-read sequencing. Pacific Biosciences (PacBio) was the first company to make long-read sequencing technology available to the mass market, with the release of its RS platform in 2011, shortly followed by the widely adopted RS-II platform in 2013. PacBio sequencing

626

S. V. Owen et al.

technology is comparable to Illumina short-read technology in that it is polymerasedependent; sequence data result from the detection of base incorporation by a DNA polymerase enzyme onto a growing nucleotide chain. However, whereas Illumina sequencing relies on detection of base incorporation within a clonal population of concomitantly amplified DNA fragments, PacBio sequencers capture incorporation signals from single DNA molecules. This feat is made possible by physical anchoring of polymerase enzymes within narrow wells, which allows video recording of laser excitation of each fluorescently labeled nucleotide in direct contact with the anchored polymerase during DNA synthesis. Although the technology is intrinsically error-prone due to dependence on a polymerase enzyme and signal noise resulting from unincorporated nucleotides, a high degree of accuracy is achieved by using hairpin DNA adapters to create circular templates, which are sequenced continuously until polymerase function declines. Repetitive sequencing of the same DNA fragment allows random errors to be detected downstream and results in the output of high-quality sequence data. The maximum read length of a PacBio sequencer is dependent on the life of individual polymerase enzymes, thought to be between 10 and 60 kb (Rhoads and Au 2015). Sequencing takes place on SMRT Cells, which are chips containing 150,000 anchored polymerase wells. The second PacBio sequencing platform, Sequel, was released in 2015 and increased the capacity of each sequencing run nearly seven times (Sequel SMRT Cells contain one million polymerase wells), generating 5–10 gb of data per run. PacBio sequence library preparation protocols require double-stranded DNA, and therefore this technology is only directly applicable to dsDNA phages. There are currently no reports of PacBio technology to sequence ssDNA or RNA phage genomes. However, RNA may be reverse-transcribed into cDNA for sequencing on a PacBio instrument (Tseng and Underwood 2013), and a second strand DNA synthesis step may facilitate the sequencing of ssDNA phage genomes. A frequently cited advantage of PacBio technology is the additional generation of chemical modification data along with sequence data (Rhoads and Au 2015). DNA-chemical modifications, such as the addition of methyl groups to cytosine residues (DNA methylation), cause characteristic kinetic changes in nucleotide incorporation rates by the polymerase and can be detected by automated analysis of the kinetic pattern of the DNA synthesis reaction. Identification of DNA modifications can be of particular interest for bacteriophages, as certain phage groups are known to incorporate modified bases into their genomes, potentially involved in escaping host restriction modification systems (Klumpp et al. 2010; Adriaenssens et al. 2012; Lee et al. 2018). The second long-read sequencing technology in widespread use is being developed by Oxford Nanopore Technologies (ONT), and their prototype platform, MinION, was released in 2014 (Ip et al. 2015). Unlike the majority of DNA sequencing technologies, ONT does not detect nucleotide addition during DNA synthesis but instead directly detects the nucleotide composition of a single-stranded DNA or RNA molecule. The technology employs anchored pore proteins (nanopores), each under an electric current. Nucleic acid molecules are threaded

Detection of Bacteriophages: Sequence-Based Systems

627

and natively transported through the pore protein through the action of a coupled motor protein, and each nucleotide on the molecule causes characteristic and detectable current disruptions which are translated into sequence. ONT uses nucleic acid adaptors and tethers to facilitate the threading of single-stranded molecules into individual nanopore proteins. Individual reads of ~200,000 bp are reported (Ip et al. 2015), but reads of >1,000,000 bp have been reported anecdotally. On the MinION platform, sequencing takes place on flow cells harboring 512 nanopore channels and capable of generating 10–20 gb of sequence data. ONT has recently released two high-throughput sequencing platforms based on the same technology as the MinION, the GridION and the PromethION, which run multiple flow cells or use flow cells with an increased number of nanopore channels. Though commercial ONT sequencing is one of the purported applications of the larger GridION and PromethION platforms, at the time of writing, ONT is not commercially available at the majority of sequencing facilities, and currently the primary users of ONT are research labs in possession of MinION devices. The primary advantage of ONT is the portability and usability of the platforms. All platforms can be run on benchtops, using very simple DNA library preparation protocols, and sequence data can be interpreted in real time. Indeed, the portability of ONT sequencing technology was recently demonstrated by the sequencing of the phage lambda genome using a MinION device onboard the International Space Station (Castro-Wallace et al. 2017). Consequently, the obvious application of ONT is circumstances where rapid outputs are required such as in field or diagnostic settings. A major limitation of ONT sequencing is that it requires a high amount of input DNA, which may limit its use in low-biomass environments such as the ocean virome or for phages which are difficult to amplify. On the other hand, there may be an application for ONT technologies in high-biomass environments such as fecal samples. Perhaps ONT sequencing could be used to monitor the realtime abundance and sequence variation of phages and bacteria during phage therapy trials of gastrointestinal pathogens such as Clostridium difficile. Furthermore, ONT is the only sequencing platform that permits direct sequencing of RNA molecules (as opposed to reverse transcription of RNA into cDNA for sequencing). Like PacBio technology, detection of chemically modified nucleotides is made possible by analysis of signature current changes (Rand et al. 2017). ONT may therefore be highly applicable to the sequencing of RNA phage genomes, and the technology has recently been used to sequence the native RNA genome of an influenza A virus for the first time, identifying chemical modifications that were undetected using cDNA-sequencing strategies (Keller et al. 2018). A disadvantage of both long-read sequencing technologies is the poor accuracy of single reads. As single reads from both technologies represent the sequencing of single molecules, and therefore single-base calls, random error is frequent. Reads from the ONT MinION platform have been reported to have an error rate of 38.2% (Laver et al. 2015), and the error rate for single PacBio reads has been reported to be 11–15% (Rhoads and Au 2015). Therefore, some caution must be taken when utilizing these technologies for applications where accuracy is crucial, such as

628

S. V. Owen et al.

de novo genome assembly. Provided the input DNA is sequenced in enough depth, i.e., each nucleotide in the sequence is represented on multiple independent reads, erroneous base calls can be eliminated by consensus calling. This process is termed “read polishing,” and increased accuracy can also be achieved by combining the sequencing outputs from long- and short-read sequencing technologies to yield “hybrid” assemblies (Phillippy 2017). The primary advantage of long-read sequencing technologies such as PacBio and ONT over short-read technology such as Illumina is the ability to generate reads spanning repetitive sequence regions, thereby greatly improving DNA sequence assembly. Though large repetitive regions, such as sequence duplications and transposable elements, are not common features of phage genomes, there are instances where short-read technology alone is insufficient to complete phage genomes. For example, myoviruses infecting bacteria of the genus Bacillus can exhibit heavily chemically modified DNA which impedes routine short-read sequencing strategies (Klumpp et al. 2010). PacBio sequencing has been used to complete several of these myovirus genomes, showing that it can be a useful alternative for difficult-tosequence phages (Klumpp et al. 2014). Restriction modification systems are known to be important in the interaction between phages and their bacterial hosts (Adams and Burdon 1985), and therefore the secondary use of long-read sequencing technologies to map chemical genome modifications may be an important application of long-read sequencing technology to phage biology in the future.

A Note on Sequence Data and Sharing The platforms described above generate multiple gigabytes of data that need to be processed for quality, assembled and analyzed, but where possible should also be shared to improve reproducibility and promote open science. The most appropriate way for data sharing is through the International Nucleotide Sequence Database Collaboration (INSDC) which links the three main international sequence data organizations, DDBJ (DNA Data Bank of Japan), EMBL-EBI (European Molecular Biology Laboratory-European Bioinformatics Institute), and NCBI (National Center for Biotechnology Information) (Karsch-Mizrachi et al. 2012). Submissions only need to be made to a database of one of these organizations to be shared across all. For the DDBJ and NCBI, unassembled sequencing reads should be deposited in the Sequence Read Archive, shortened to DRA (DDBJ) or SRA (NCBI) (Leinonen et al. 2011b). The associated metadata for studies and samples are collected as BioProjects and BioSamples, respectively (Barrett et al. 2012). The assembled and annotated complete phage genomes can be deposited in GenBank (Benson et al. 2013) or DDBJ annotated sequence submissions. For EMBL-EBI, the European Nucleotide Archive (ENA) accepts all types of sequences and associated metadata through the same submission portal (Leinonen et al. 2011a). Any genomes deposited and released through the above-described resources will become a part of publicly available databases that can be searched through the Basic Local Alignment Search Tool (BLAST) (Johnson et al. 2008).

Detection of Bacteriophages: Sequence-Based Systems

629

NCBI have in recent years developed virus-specific tools that can be used for phage sequence analysis. The NCBI Viral Genomes Resource groups offer specialized resources for analysis of phage genomes, such as curated reference genome databases, sequence comparison tools, protein clusters, and custom downloads (Brister et al. 2015).

Applications of Sequencing-Based Detection Methods Sequencing-based detection of bacteriophages is possible at different levels of complexity (Fig. 1). In this section, we discuss four specific applications of sequencing-based methods, single-gene amplicon sequencing, phage genome sequencing, prophage detection, and metagenomic sequencing of viral communities.

Fig. 1 Overview of possible applications of sequencing-based detection of bacteriophages at different levels of complexity. Sequencing discussed in this chapter describes, in order of rising complexity, the single-gene level, complete phage genome sequencing, prophage detection in bacterial genomes, and metagenomes. More complex systems to be sequenced require higher sequencing yields. Sanger sequencing, therefore, is only appropriate for gene-level detection or sequencing of very small genomes. Illumina MiSeq (Ion Torrent PGM) is ideally suited for phage genome sequencing, which can be complemented with long-read technology for detection of a phage genome in a single read. Prophage detection requires bacterial genome sequencing and specific computational identification tools. Metagenomic-based detection of phages requires high sequencing yields and often has low-input biomass, requiring experiment-specific decisions on the sequencing platform to use (e.g., Illumina HiSeq giving higher yields than MiSeq)

630

S. V. Owen et al.

Gene-Based Detection of Bacteriophages Perhaps the simplest way of detecting bacteriophages is to amplify a gene or gene fragment of the bacteriophage in question and then determine its sequence and taxonomic affiliation, a method called amplicon sequencing. This can be straightforward if the target bacteriophage has a published genome in an INSDC database. Then it is simply a matter of designing primers in a unique location of the genome, performing a PCR amplification and sequencing the PCR fragment. But it is also possible to detect “unknown” phages in a sample by amplifying a signature or marker gene. Unfortunately, there is no such thing as a universal viral/phage marker gene, comparable to the 16S rRNA gene in bacteria or the 18S rRNA gene in eukaryotes (Rohwer and Edwards 2002), which can be used to screen for any bacteriophage. There have been, however, primers developed that amplify signature genes targeting specific phage groups (Adriaenssens and Cowan 2014). In many cases, these are specific for a small group of viruses within a family of tailed phages. Before the rise of short-read sequencing technology, PCR fragments used to be cloned into plasmid vectors, and the single inserts were sequenced with plasmid primers using Sanger sequencing. Currently, sequencing of amplicons representing diverse communities can be done with any of the sequencing platforms described above, but longer reads will lead to better resolution of viral diversity. Processing of the sequencing results is generally performed by clustering sequencing reads into operational taxonomic units (OTUs) based on a threshold similarity score (90–99% identity) or by direct sequence variant comparison. Several pipelines are available for the analysis of amplicon data dealing with quality control, removal of chimeric reads, and OTU clustering, such as MOTHUR (Schloss et al. 2009) and QIIME (Caporaso et al. 2010). The resulting OTUs can then be assigned to a taxonomic group and further investigated using phylogenetics. In the following section, we will give an overview of the phage groups which have been detected in previous studies using PCR amplification and sequencing.

Cyanophages Most phage amplicon studies have been targeted toward discovering the diversity of cyanophages in the environment, i.e., phages infecting cyanobacteria. The signature genes which have been used successfully in the past include structural genes, such as the portal protein (Fuller et al. 1998; Zhong et al. 2002; Short and Suttle 2005) or major capsid protein of myoviruses (Baker et al. 2006), or metabolism-related genes such as the ones coding for photosystem II proteins psbA and psbD (Zeidner et al. 2003; Millard et al. 2004; Clokie et al. 2006; Sullivan et al. 2006; Wang and Chen 2008) or the ATPase phoH (Marston and Amrich 2009; Goldsmith et al. 2011). Some of these marker genes target more specific or less diverse groups than others, for example, the structural protein markers target only a subgroup of phages belonging to the Myoviridae, whereas phoH is present in 40% of cultured marine phages, in certain eukaryotic viruses and in some phages infecting enteric bacteria.

Detection of Bacteriophages: Sequence-Based Systems

631

T4-Like Phages The type isolate for this group of phages with contractile tails, Escherichia phage T4, is an iconic phage with a long history in molecular biology. Its major capsid protein (gp23) has been the basis for most of the primer sets (Filée et al. 2005; Comeau and Krisch 2008; Marston and Amrich 2009; Chow and Fuhrman 2012). Originally used in the marine environment, these gp23 amplicons have been found around the globe in, for example, rice paddy soil (Fujii et al. 2008; Wang et al. 2009a, b), freshwater lakes in Russia (Butina et al. 2010), and even in Antarctica (López-Bueno et al. 2009) and an Arctic glacier (Bellas and Anesio 2013). The detection of this very diverse group using amplicons overlaps with cyanophage detection as many cyanophages with myovirus morphology are related to T4. T7-Like Phages Detection of T7-like phages belonging to the family Podoviridae can be done by targeting the DNA polymerase (polA) gene, with at least nine primer sets published so far (Breitbart et al. 2004; Labonté et al. 2009; Chen et al. 2009; Dekel-Bird et al. 2013). Partial sequences have been detected in a range of habitats around the world, including marine, freshwater, and terrestrial (Breitbart et al. 2004; Chen et al. 2009; Huang et al. 2010). Gokushoviruses ssDNA phages belonging to the subfamily Gokushovirinae, family Microviridae, have been recently shown to be ubiquitous across habitats and geographic regions, based on amplification of the major capsid protein gene (Hopkins et al. 2014; Székely and Breitbart 2016). Using this gene, bloom-bust patterns and fluctuations in microvirus abundance were discovered in two freshwater lakes in France (Zhong et al. 2015). Other Potential Signature Genes There are other genes which are conserved among groups of phages, but which have not yet been investigated using gene-based sequencing approaches. A great resource to find a signature gene for a phage group of interest is the prokaryotic virus orthologous groups (pVOGs) database, formerly known as phage orthologous groups (POGs) (Kristensen et al. 2013; Grazziotin et al. 2017). This database comprises protein clusters (pVOGs) for all sequenced bacterial and archaeal viruses and can therefore be used to identify shared genes between any taxon of interest. For example, the second largest pVOG, VOG4544 terminase large subunit, is found in 83% of the tailed phages (order Caudovirales) (Grazziotin et al. 2017) and has been used in many phylogenetic analyses from isolates to virome studies (Sullivan et al. 2009; Roux et al. 2014). RNA phages all encode an RNA-dependent RNA polymerase, which can act as a signature gene, but only eukaryotic-infecting RNA viruses have been found with published primer sets (Culley et al. 2003; Culley and Steward 2007). A helpful tool for the choice of signature genes and primer design of a phage group of interest is PhiSigns, which is both available as stand-alone tool and web-based application (Dwivedi et al. 2012).

632

S. V. Owen et al.

Genome Sequencing The year 2017 marked 40 years since the first bacteriophage, the ssDNA phage ɸX174, was sequenced (Sanger et al. 1977), followed shortly after by the genomes of the reference phages lambda and T7 (Sanger et al. 1982; Dunn et al. 1983). Due to their small sizes, ranging from approximately 3.5 kb (Friedman et al. 2009) to nearly 500 kbp (Hatfull and Hendrix 2011), phage genomes were sequenced well in advance of the first bacterial genome, in 1995 (Fleischmann et al. 1995). However, the number of complete, i.e., finished or entirely sequenced genomes, and “whole genome shotgun” sequences available in public databases is higher for bacteria than for phages, despite the genome size differences that make the latter easier to sequence and assemble. Still, sequencing of novel phage genomes has increased greatly during the last two decades (Adriaenssens and Brister 2017). The detection of bacteriophages based on their genome sequences could only be achieved after having a repertoire of different completed or close to completion genomes. Likewise, sequencing bacteriophage genomes is crucial to broadening knowledge of their biology, such as metabolic processes and interactions with their host and environment. This is one of the main drawbacks of the currently small number of phage genomes available, which is only a very small fraction compared to the predicted diversity of bacteriophages (Perez Sepulveda et al. 2016). Despite the relatively small number of sequenced phage genomes, their sequences have contributed significantly to the discovery of several phage genes directly involved in the host metabolism (Zeidner et al. 2003; Millard et al. 2004, 2009, 2010; Sullivan et al. 2006; Wang and Chen 2008; Sabehi et al. 2012; Chan et al. 2015). These auxiliary metabolic genes (AMGs) can play important roles in redirecting host metabolism by, for example, guiding carbon flux to the biosynthesis of deoxynucleotides through the pentose phosphate pathway and hence favoring bacteriophage replication (Thompson et al. 2011). As mentioned before, detection of bacteriophages using marker genes can only be used for a reduced number of bacteriophages, but by having a fuller understanding of their genomes, either the range of markers can be extended or potentially the relatively small genomes can be used as markers, allowing detection of supposed “unculturable” bacteriophages. Obtaining bacteriophages for sequencing normally involves the infection of suitable hosts by “plaque assay” in which individual plaques (i.e., visible clearing of bacterial culture that represents the lysis of the host by replication of bacteriophages) (see chapter ▶ “Detection of Bacteriophages: Statistical Aspects of Plaque Assay”) are picked (i.e., selected) and propagated further to increase phage biomass. Bacteriophages may be further concentrated by methods such as polyethylene glycol precipitation and/or cesium chloride density gradient centrifugation. Nucleic acids are then extracted and purified for sequencing using standard methods, such as phenol-chloroform phase separation or the use of commercial kits. The dependence on the use of a sensitive host bacterium has limited the approach to sequencing only those bacteriophages capable of infecting the small proportion of bacteria that can be grown under laboratory conditions, even if those bacteria do not necessarily

Detection of Bacteriophages: Sequence-Based Systems

633

represent the “primary” host. This has become a problem for environments where not many hosts can be cultured. An example of this phenomenon is presented in a study by Brum and colleagues who found only 39 genomes that could be associated to “cultured” bacteriophages compared to an estimated total of 5,476 different dominant bacteriophage populations in the upper ocean (Brum et al. 2015). Another issue related to this method of culturing phages is that these plaques can contain more than one phage, which can be due to spontaneous induction of any host prophages (Henn et al. 2010; Cowley et al. 2015). Having multiples phages in a sample can cause incorrect assemblies and misinterpretation of genomes. During recent years, there have been developments and method optimizations to sequence phage genomes extracted from single plaques, reducing the input material and costs associated with library preparation (Kot et al. 2014; Baym et al. 2015). In line with these advances, Rihtman and colleagues demonstrated successful Illumina high-throughput sequencing and assembly of samples containing multiple bacteriophages; the combination of multiple genomes per library preparation allows for costeffective nucleic acid extraction and sequencing (Rihtman et al. 2016). As a general rule, a read coverage (i.e., the number of sequencing reads generated for each base of the genome) of 30X is adequate for successful assembly of a genome. Due to variation in the properties of different bacteriophages, in the case of multiplexing genomes, a coverage of 100X has been suggested for obtaining a reliable assembly. It is worth noting, however, that the efficiency described when sequencing multiple bacteriophages per library does not necessarily apply to closely related, i.e., similar, bacteriophages. Thus, in order to avoid mistakes caused by possible undetectable mis-assemblies, it is recommended to only multiplex genomes obtained from different hosts. The study of bacteriophage genomes has increased the knowledge of their structure and composition, which consequently allows the development and design of novel methods for detection, including new bioinformatic tools. Phage genomes are normally free from complex sequences than can massively affect genome assembly, such as transposable elements and repetitive sequences (i.e., gene duplications or variable-number tandem repeats). However, re-sequencing of cultured bacteriophages using different sequencing technologies can help to address the potential issues derived from sequence complexity, allowing increased accuracy and better discrimination of correct assemblies by including more data. Additionally, re-sequencing of bacteriophages decreases the needed coverage for proper assembly, thereby providing significant information regarding the evolution of bacteriophage genomes (Puxty et al. 2015).

Sequence-Based Identification of Prophages A prophage represents a stage in the life cycle of a temperate phage wherein the phage genetic material is transmitted vertically with that of the bacterial host, either integrated into the chromosome or existing as a low-copy number plasmid (see ▶ “Temperate Phages, Prophages, and Lysogeny”). Prophages have been shown to

634

S. V. Owen et al.

be important elements of horizontal gene transfer which can contribute significantly to bacterial niche adaptation and population dynamics (Bossi et al. 2003; Fortier and Sekulovic 2013). In particular, prophages have played a crucial role in the evolution of some notable bacterial pathogens, by encoding toxins and virulence factors, such as the Stx toxin of Shiga-toxigenic Escherichia coli, the cholera toxin of Vibrio cholerae, and the C1 neurotoxin of Clostridium botulinum (Brüssow et al. 2004). Consequently, the identification of prophages is an important application of sequencing technologies. As part of the realm of the temperate phage is within the genome of bacteria, temperate phage genome sequencing is a natural by-product of the sequencing of bacterial genomes, and no specific methodological considerations are necessary to sequence bacterial genomes containing prophages. It is likely, therefore, that temperate phages are the most deeply sequenced of all phages. In fact, given that bacteria typically have multiple prophage sequences incorporated into their genomes (Casjens 2003), it could be argued that more genome sequences exist for temperate phages than for their bacterial hosts and therefore any other group of organisms on the planet. However, the challenge of accessing this wealth of temperate phage genome sequence data is in identifying prophages in bacterial genome sequences. Analogous to identifying bacteriophages from the environment using gene markers, temperate phages can be identified from within bacterial sequence space using sequence markers. A number of computational tools have been developed to mine bacterial genome datasets for prophage sequences including Phage_Finder (Fouts 2006), Prophage Finder (Bose and Barber 2006), Prophinder (Lima-Mendez et al. 2008), PHAST (Zhou et al. 2011), PhiSpy (Akhter et al. 2012), VirSorter (Roux et al. 2015), and PHASTER (Arndt et al. 2016). All these tools rely on characteristics of prophage genome sequences, such as identification of sequences with homology to characterized phage genes and identification of direct repeats corresponding to phage attachment (att) sites or regions of DNA with differential GC skew, protein length, or transcription strand directionality. Unfortunately, there are currently several limitations to these approaches. Firstly, most of the listed prophage identification tools are designed to be run on complete (single-contig) bacterial genome sequences. This is because the naturally modular nature of prophages makes it impossible to accurately predict which prophagecontaining contigs belong together when multiple contigs are present. Prophage prediction tools rely heavily on the locality of prophage-signature sequences, but this locality is likely to be arbitrary in unfinished, multi-contig assemblies. Though one of the recent prophage prediction tools, PHASTER, is able to handle multi-contig files, a single functional prophage that assembles as two contigs would likely be assigned as two “incomplete” prophages. As the vast majority of bacterial genome sequence data exists as unfinished, contiguous genomes, accessing the prophage content of these genomes remains challenging. A second limitation lies in inferring functionality of the prophage. Prophage sequences within bacterial genomes can exist in a variety of complex functional states, from fully functional (capable of induction and replication), to defective but capable of resuscitation, to extremely degraded representing a host-domesticated prophage remnant. Distinguishing between functional, nonfunctional, and

Detection of Bacteriophages: Sequence-Based Systems

635

incomplete prophage remnants in silico is extremely difficult. This task is even further complicated by the issue of multi-contig genome assemblies discussed above. How can a remnant prophage island be computationally distinguished from a functional prophage disrupted across separate contigs in an unfinished bacterial genome assembly? Even in complete, single-contig assemblies, mutations that inactivate prophage function can be as subtle as singular noncoding SNPs (Owen et al. 2017), making the accurate computational prediction of prophage function virtually impossible. As well as the sequencing of prophages together with their bacterial hosts, other methods to sequence temperate phages exist. Functional temperate phages may be easily sequenced by amplification on a sensitive host, virion concentration, and purification, as has been described for phage genome sequencing above. Even when putative temperate prophages cannot be replicated, for example, if a sensitive host strain is not available, chemically induced prophage induction (i.e., using SOS response-inducing agents such as mitomycin C, norfloxacin, nalidixic acid, or UV-light exposure) may be used to induce the formation of phage particles (Raya and Hébert 2009) (see ▶ “Temperate Phages, Prophages, and Lysogeny”). Phage particles can then be purified from the culture supernatant for sequencing. From this it can be determined which, if multiple prophage-like sequences are present in the genome, can be induced to form phage particles. An important consideration when sequencing induced phages from culture supernatant is the removal of contaminating bacterial chromosomal DNA. Thorough DNase treatment procedures should therefore be undertaken, such that non-prophage chromosomal DNA can no longer be detected in the sample by PCR, to ensure that only virion-encapsulated phage DNA is sequenced. The advantage of this technique is that it may allow temperate phages to be identified in previously uncharacterized bacteria, in which little functional information can be obtained based on homology to known genes. Sequencing of protein-encapsulated DNA would provide convincing evidence for the existence of novel prophages, even if the sequence had no homology to known phages. Finally, methods have been developed to selectively enrich DNA samples for phages that exist as extrachromosomal elements, for example, as circular or linear plasmids. Certain bacterial genera such Chlamydia and Borrelia which are associated with small genome sizes have been found to harbor extrachromosomal prophage plasmids rather than integrated prophages (Casjens 2003). A method to selectively enrich genomic DNA samples for extrachromosomal prophage elements in Staphylococcus by the use of plasmid purification kits has been reported (Utter et al. 2014). However, the increasing adoption of long-read sequencing technology may represent a more effective way to investigate extrachromosomal prophage elements.

Metagenomics-Based Detection of Bacteriophages It is possible to detect phages in any environment without any prior knowledge about their genome using shotgun metagenomics or more specifically metaviromics. In its

636

S. V. Owen et al.

most basic form, metagenomics is the sequencing of all the nucleic acid extracted from the environment (or any sample of interest). For phages, and viruses in general, these metagenomics protocols get amended to account for the smaller genome sizes of bacteriophages compared to bacteria, fungi, and protists. In many environments with low biomass, a concentration step is necessary to reduce the volume of aquatic input material before sample processing, with concentration generally performed using tangential flow filtration (Vega Thurber et al. 2009) or, for smaller volumes, by using spin filter columns in a benchtop centrifuge (Bolduc et al. 2012). The most commonly used step following concentration is the combination of viral enrichment by size exclusion of the cellular fraction using filters and nuclease treatment for the removal of free-floating DNA and RNA (Vega Thurber et al. 2009; Hall et al. 2014). When using size exclusion to remove bacteria, it is important to keep in mind the size of the phage target, with some of the jumbo myoviruses such as Pseudomonas phages phiKZ and EL and their relatives (head diameter of >120 nm and tail length of ~200 nm) potentially clogging a 0.22 μm filter pore (Hertveldt et al. 2005). Subsequently, the phage (viral) community nucleic acid can be extracted and sequenced using next-generation sequencing platforms. In most cases, researchers will be targeting dsDNA as this genome group represents most known phages. The first use of metaviromics was to explore uncultured marine virus communities and found that a large fraction of these communities were made up of phages (Breitbart et al. 2002). Ever since this seminal paper, the techniques and approaches in metaviromics have been optimized and updated to investigate viral/phage ecology, in a field that has started to come of age (Sullivan 2015; Sullivan et al. 2017). Virtually every habitat sampled and analyzed with metaviromics has shown that phages make up a substantial fraction, if not the majority of identified sequences. This includes from pristine environments in the polar regions (López-Bueno et al. 2009; Zablocki et al. 2014; Aguirre de Cárcer et al. 2015; Adriaenssens et al. 2017), over the global oceans (Huang et al. 2010; Mizuno et al. 2013; Brum et al. 2015), freshwater lakes (Roux et al. 2012; Labonté and Suttle 2013; Skvortsov et al. 2016), and soils (Fierer et al. 2007; Zablocki et al. 2017). Currently, only marine epipelagic habitats have been sampled near saturation, allowing network-based clustering approaches to describe the full diversity, revealing that the most abundant viral clusters represent phages infecting members of the phyla Actinobacteria, Proteobacteria, Bacteroidetes, Cyanobacteria, and Deferribacteres (Roux et al. 2016). Metaviromics has become the go-to method to identify human-associated phage communities. Early studies of the human gut revealed highly diverse and unknown phage communities, with the majority of the known phage signatures belonging to tailed phages of the order Caudovirales (Breitbart et al. 2003). Comparative analyses revealed high interpersonal differences in gut phage communities, but low levels of change over time within the same individual, and a predominance of temperate phages (Reyes et al. 2010, 2012, 2015; Minot et al. 2011; Manrique et al. 2016). Metaviromic sequencing of ultrasmall amounts of DNA also showed unique phage communities on the human skin with differences according to topical sites and large intrapersonal differences (Hannigan et al.

Detection of Bacteriophages: Sequence-Based Systems

637

2015). Other metavirome studies have found viral communities dominated by bacteriophages in bodily fluids, such as saliva and urine, and to a lesser extent in blood where phages might have originated from contamination of the sequencing procedure (Pride et al. 2012; Santiago-Rodriguez et al. 2015; Moustafa et al. 2017). In an alternative approach, metaviromic sequencing analyses have been used to identify the phages present in phage cocktails used in phage therapy treatments and experiments. The first such cocktail analyzed was a Russian cocktail called ColiProteus (Microgen) used against E. coli and Proteus infections (McCallin et al. 2013). This study revealed 17 different phage groups present in the cocktail at different abundances, suggesting that some of the low abundance groups were by-products of phage cocktail production. The second study investigated the Intesti phage cocktail from the Eliava Institute in Georgia, active against a range of enteric bacteria (Zschach et al. 2015) (see chapter ▶ “Current Updates from the LongStanding Phage Research Centers in Georgia, Poland, and Russia”). The metaviromic analysis showed 23 different sequence groups, called phage clusters by the researchers, falling within the families Myoviridae, Siphoviridae, and Podoviridae and an unassigned grouping. These two studies showed that these phage cocktails were more complex than initially assumed and might contain additional (unwanted) phage sequences at low abundances. One of the most interesting discoveries to come from using a metaviromic sequencing approach is the discovery of a highly abundant phage in human gut viromes (Dutilh et al. 2014). The researchers in this study used a cross-assembly approach (assembling multiple datasets together in one contig set) in order to increase contig length and establish a co-occurrence profile of contigs over the different samples (Dutilh et al. 2012). They were able to reconstruct a circular contig of ~97 kb representing a phage genome they labeled crAssphage and verified its existence by long-range PCR and Sanger sequencing. In a comparison with all published metagenomes at the time, this phage was found to make up a significant portion of gut metagenomes [up to 90% of reads of the twin dataset (Reyes et al. 2010)] and represented 1.7% of all sequencing reads from human feces, making crAssphage one of the most abundant phages in publicly available datasets. With more phage genomes sequenced and metagenome data becoming available, it is possible that more of these abundant, unknown phages will be discovered.

Conclusions The development of sequencing technology and the subsequent boom in nextgeneration sequencing platforms (both short- and long-read platforms) has been fundamental in advancing bacteriophage research. These methods have not only contributed to an explosion of genomes in public databases but have also provided an opportunity to exploit sequencing-based methods for bacteriophage detection. At the simplest level of complexity, phages can be detected by the sequencing of a single marker gene. Whole genome sequencing of isolated phages has populated reference databases, while bacterial genome sequencing led to the discovery of

638

S. V. Owen et al.

a plethora of previously undetected prophage genomes. At the community level, shotgun sequencing methods have made it possible for researchers to investigate all bacteriophages in a sample without previous knowledge of its content. In conclusion, sequencing and next-generation sequencing technology has added a new layer to bacteriophage research, opening new avenues of research, from exploitation of genes for biotechnological applications to population ecology.

Cross-References ▶ Current Updates from the Long-Standing Phage Research Centers in Georgia, Poland, and Russia ▶ Detection of Bacteriophages: Statistical Aspects of Plaque Assay ▶ Temperate Phages, Prophages, and Lysogeny

References Abedon ST (2017) Information phage therapy research should report. Pharmaceuticals 10:1–17. https://doi.org/10.3390/ph10020043 Ackermann H-W (2011) Bacteriophage taxonomy. Microbiol Aust 32:90–94 Adams RLP, Burdon RH (1985) The function of DNA methylation in bacteria and phage. In: Molecular biology of DNA methylation. Springer, New York, pp 73–87 Adriaenssens E, Brister JR (2017) How to name and classify your phage: an informal guide. Viruses 9:70. https://doi.org/10.3390/v9040070 Adriaenssens EM, Cowan DA (2014) Using signature genes as tools to assess environmental viral ecology and diversity. Appl Environ Microbiol 80:4470–4480. https://doi.org/10.1128/ AEM.00878-14 Adriaenssens EM, Ackermann H-W, Anany H et al (2012) A suggested new bacteriophage genus: “Viunalikevirus”. Arch Virol 157:2035–2046. https://doi.org/10.1007/s00705-012-1360-5 Adriaenssens EM, Kramer R, Van Goethem MW et al (2017) Environmental drivers of viral community composition in Antarctic soils identified by viromics. Microbiome 5:83. https:// doi.org/10.1186/s40168-017-0301-7 Adriaenssens EM, Wittmann J, Kuhn JH et al (2018) Taxonomy of prokaryotic viruses: 2017 update from the ICTV Bacterial and Archaeal Viruses Subcommittee. Arch Virol 163:1125–1129. https://doi.org/10.1007/s00705-018-3723-z Aguirre de Cárcer D, López-Bueno A, Pearce DA, Alcamí A (2015) Biodiversity and distribution of polar freshwater DNA viruses. Sci Adv 1:e1400127. https://doi.org/10.1126/sciadv.1400127 Akhter S, Aziz RK, Edwards RA (2012) PhiSpy: a novel algorithm for finding prophages in bacterial genomes that combines similarity- and composition-based strategies. Nucleic Acids Res 40:e126. https://doi.org/10.1093/nar/gks406 Alavidze Z, Aminov R, Betts A et al (2016) Silk route to the acceptance and re-implementation of bacteriophage therapy. Biotechnol J 11:595–600. https://doi.org/10.1002/biot.201600023 Arndt D, Grant JR, Marcu A et al (2016) PHASTER: a better, faster version of the PHAST phage search tool. Nucleic Acids Res 44:W16–W21. https://doi.org/10.1093/nar/gkw387 Baker AC, Goddard VJ, Davy J et al (2006) Identification of a diagnostic marker to detect freshwater cyanophages of filamentous cyanobacteria. Appl Environ Microbiol 72:5713–5719. https://doi.org/10.1128/AEM.00270-06

Detection of Bacteriophages: Sequence-Based Systems

639

Barrett T, Clark K, Gevorgyan R et al (2012) BioProject and BioSample databases at NCBI: facilitating capture and organization of metadata. Nucleic Acids Res 40:57–63. https://doi.org/ 10.1093/nar/gkr1163 Baym M, Kryazhimskiy S, Lieberman TD et al (2015) Inexpensive multiplexed library preparation for megabase-sized genomes. PLoS One 10:e0128036. https://doi.org/10.1371/journal. pone.0128036 Bellas CM, Anesio AM (2013) High diversity and potential origins of T4-type bacteriophages on the surface of Arctic glaciers. Extremophiles 17:861–870. https://doi.org/10.1007/s00792-013-0569-x Benson DA, Cavanaugh M, Clark K et al (2013) GenBank. Nucleic Acids Res 41:36–42. https:// doi.org/10.1093/nar/gks1195 Bolduc B, Shaughnessy DP, Wolf YI et al (2012) Identification of novel positive-strand RNA viruses by metagenomic analysis of archaea-dominated Yellowstone hot springs. J Virol 86:5562–5573. https://doi.org/10.1128/JVI.07196-11 Bose M, Barber RD (2006) Prophage Finder: a prophage loci prediction tool for prokaryotic genome sequences. In Silico Biol 6:223–227 Bossi L, Fuentes JA, Mora G, Figueroa-Bossi N (2003) Prophage contribution to bacterial population dynamics. J Bacteriol 185:6467–6471 Breitbart M, Salamon P, Andresen B et al (2002) Genomic analysis of uncultured marine viral communities. Proc Natl Acad Sci U S A 99:14250–14255. https://doi.org/10.1073/ pnas.202488399 Breitbart M, Hewson I, Felts B et al (2003) Metagenomic analyses of an uncultured viral community from human feces. J Bacteriol 185:6220–6223. https://doi.org/10.1128/ JB.185.20.6220 Breitbart M, Miyake JH, Rohwer F (2004) Global distribution of nearly identical phage-encoded DNA sequences. FEMS Microbiol Lett 236:249–256. https://doi.org/10.1111/j.15746968.2004.tb09654.x Brister JR, Ako-Adjei D, Bao Y, Blinkova O (2015) NCBI viral genomes resource. Nucleic Acids Res 43:D571–D577. https://doi.org/10.1093/nar/gku1207 Brum JR, Ignacio-Espinoza JC, Roux S et al (2015) Patterns and ecological drivers of ocean viral communities. Science (80-) 348:1261498. https://doi.org/10.1126/science.1261498 Brüssow H, Canchaya C, Hardt WD (2004) Phages and the evolution of bacterial pathogens: from genomic rearrangements to lysogenic conversion. Microbiol Mol Biol Rev 68:560–602. https:// doi.org/10.1128/MMBR.68.3.560 Butina TV, Belykh OI, Maksimenko SY, Belikov SI (2010) Phylogenetic diversity of T4-like bacteriophages in Lake Baikal, East Siberia. FEMS Microbiol Lett 309:122–129. https://doi. org/10.1111/j.1574-6968.2010.02025.x Caporaso JG, Kuczynski J, Stombaugh J et al (2010) QIIME allows analysis of high-throughput community sequencing data. Nat Methods 7:335–336. https://doi.org/10.1038/nmeth.f.303 Casjens S (2003) Prophages and bacterial genomics: what have we learned so far? Mol Microbiol 49:277–300 Castro-Wallace SL, Chiu CY, John KK et al (2017) Nanopore DNA sequencing and genome assembly on the International Space Station. Sci Rep 7:18022. https://doi.org/10.1038/ s41598-017-18364-0 Chan Y-WW, Millard AD, Wheatley PJ et al (2015) Genomic and proteomic characterization of two novel siphovirus infecting the sedentary facultative epibiont cyanobacterium Acaryochloris marina. Environ Microbiol 17:4239–4252. https://doi.org/10.1111/1462-2920.12735 Chen F, Wang K, Huang S et al (2009) Diverse and dynamic populations of cyanobacterial podoviruses in the Chesapeake Bay unveiled through DNA polymerase gene sequences. Environ Microbiol 11:2884–2892. https://doi.org/10.1111/j.1462-2920.2009.02033.x Chow C-ET, Fuhrman JA (2012) Seasonality and monthly dynamics of marine myovirus communities. Environ Microbiol 14:2171–2183. https://doi.org/10.1111/j.1462-2920.2012.02744.x Clokie MRJ, Millard AD, Mehta JY, Mann NH (2006) Virus isolation studies suggest short-term variations in abundance in natural cyanophage populations of the Indian Ocean. J Mar Biol Assoc UK 86:499–505. https://doi.org/10.1017/S0025315406013403

640

S. V. Owen et al.

Comeau AM, Krisch HM (2008) The capsid of the T4 phage superfamily: the evolution, diversity, and structure of some of the most prevalent proteins in the biosphere. Mol Biol Evol 25:1321–1332. https://doi.org/10.1093/molbev/msn080 Cowley LA, Beckett SJ, Chase-Topping M et al (2015) Analysis of whole genome sequencing for the Escherichia coli O157:H7 typing phages. BMC Genomics 16:271. https://doi.org/ 10.1186/s12864-015-1470-z Culley AI, Steward GF (2007) New genera of RNA viruses in subtropical seawater, inferred from polymerase gene sequences. Appl Environ Microbiol 73:5937–5344. https://doi.org/10.1128/ AEM.01065-07 Culley AI, Lang AS, Suttle CA (2003) High diversity of unknown picorna-like viruses in the sea. Nature 424:1054–1057. https://doi.org/10.1038/nature01886 Dekel-Bird NP, Avrani S, Sabehi G et al (2013) Diversity and evolutionary relationships of T7-like podoviruses infecting marine cyanobacteria. Environ Microbiol 15:1476–1491. https://doi.org/ 10.1111/1462-2920.12103 Dunn JJ, Studier FW, Gottesman M (1983) Complete nucleotide sequence of bacteriophage T7 DNA and the locations of T7 genetic elements. J Mol Biol 166:477–535. https://doi.org/ 10.1016/S0022-2836(83)80282-4 Dutilh BE, Schmieder R, Nulton J et al (2012) Reference-independent comparative metagenomics using cross-assembly: CrAss. Bioinformatics 28:3225–3231. https://doi.org/10.1093/bioinformatics/bts613 Dutilh BE, Cassman N, McNair K et al (2014) A highly abundant bacteriophage discovered in the unknown sequences of human faecal metagenomes. Nat Commun 5:4498. https://doi.org/ 10.1038/ncomms5498 Dwivedi B, Schmieder R, Goldsmith DB et al (2012) PhiSiGns: an online tool to identify signature genes in phages and design PCR primers for examining phage diversity. BMC Bioinform 13:37. https://doi.org/10.1186/1471-2105-13-37 Fierer N, Breitbart M, Nulton J et al (2007) Metagenomic and small-subunit rRNA analyses reveal the genetic diversity of bacteria, archaea, fungi, and viruses in soil. Appl Environ Microbiol 73:7059–7066. https://doi.org/10.1128/AEM.00358-07 Filée J, Tétart F, Suttle CA, Krisch HM (2005) Marine T4-type bacteriophages, a ubiquitous component of the dark matter of the biosphere. Proc Natl Acad Sci U S A 102:12471–12476. https://doi.org/10.1073/pnas.0503404102 Fleischmann R, Adams M, White O et al (1995) Whole-genome random sequencing and assembly of Haemophilus influenzae Rd. Science (80-) 269:496–512. https://doi.org/10.1126/science.7542800 Fortier L-C, Sekulovic O (2013) Importance of prophages to evolution and virulence of bacterial pathogens. Virulence 4:354–365. https://doi.org/10.4161/viru.24498 Fouts DE (2006) Phage_Finder: automated identification and classification of prophage regions in complete bacterial genome sequences. Nucleic Acids Res 34:5839–5851. https://doi.org/ 10.1093/nar/gkl732 Friedman SD, Genthner FJ, Gentry J et al (2009) Gene mapping and phylogenetic analysis of the complete genome from 30 single-stranded RNA male-specific coliphages (family Leviviridae). J Virol 83:11233–11243. https://doi.org/10.1128/JVI.01308-09 Fujii T, Nakayama N, Nishida M et al (2008) Novel capsid genes (g23) of T4-type bacteriophages in a Japanese paddy field. Soil Biol Biochem 40:1049–1058. https://doi.org/10.1016/j. soilbio.2007.11.025 Fuller NJ, Wilson WH, Joint IR, Mann NH (1998) Occurrence of a sequence in marine cyanophages similar to that of T4 g20 and its application to PCR-based detection and quantification techniques. Appl Environ Microbiol 64:2051–2060 Garneau JR, Depardieu F, Fortier L-C et al (2017) PhageTerm: a tool for fast and accurate determination of phage termini and packaging mechanism using next-generation sequencing data. Sci Rep 7:8292. https://doi.org/10.1038/s41598-017-07910-5 Goldsmith DB, Crosti G, Dwivedi B et al (2011) Development of phoH as a novel signature gene for assessing marine phage diversity. Appl Environ Microbiol 77:7730–7739. https://doi.org/ 10.1128/AEM.05531-11

Detection of Bacteriophages: Sequence-Based Systems

641

Goodwin S, McPherson JD, McCombie WR (2016) Coming of age: ten years of next-generation sequencing technologies. Nat Rev Genet 17:333–351. https://doi.org/10.1038/nrg.2016.49 Grazziotin AL, Koonin EV, Kristensen DM (2017) Prokaryotic Virus Orthologous Groups (pVOGs): a resource for comparative genomics and protein family annotation. Nucleic Acids Res 45:D491–D498. https://doi.org/10.1093/nar/gkw975 Hall RJ, Wang J, Todd AK et al (2014) Evaluation of rapid and simple techniques for the enrichment of viruses prior to metagenomic virus discovery. J Virol Methods 195:194–204. https://doi.org/ 10.1016/j.jviromet.2013.08.035 Hannigan GD, Meisel JS, Tyldsley AS et al (2015) The human skin double-stranded DNA virome: topographical and temporal diversity, genetic enrichment, and dynamic associations with the host microbiome. MBio 6:e01578–e01515. https://doi.org/10.1128/mBio.01578-15.Editor Hatfull GF (2015) Dark matter of the biosphere: the amazing world of bacteriophage diversity. J Virol 89:8107–8110. https://doi.org/10.1128/JVI.01340-15 Hatfull GF, Hendrix RW (2011) Bacteriophages and their genomes. Curr Opin Virol 1:298–303. https://doi.org/10.1016/j.coviro.2011.06.009 Henn MR, Sullivan MB, Stange-Thomann N et al (2010) Analysis of high-throughput sequencing and annotation strategies for phage genomes. PLoS One 5:e9083. https://doi.org/10.1371/ journal.pone.0009083 Hertveldt K, Lavigne R, Pleteneva E et al (2005) Genome comparison of Pseudomonas aeruginosa large phages. J Mol Biol 354:536–545. https://doi.org/10.1016/j.jmb.2005.08.075 Hopkins M, Kailasan S, Cohen A et al (2014) Diversity of environmental single-stranded DNA phages revealed by PCR amplification of the partial major capsid protein. ISME J 8:2093–2103. https://doi.org/10.1038/ismej.2014.43 Huang S, Wilhelm SW, Jiao N, Chen F (2010) Ubiquitous cyanobacterial podoviruses in the global oceans unveiled through viral DNA polymerase gene sequences. ISME J 4:1243–1251. https:// doi.org/10.1038/ismej.2010.56 Ip CLC, Loose M, Tyson JR et al (2015) MinION analysis and reference consortium: phase 1 data release and analysis. F1000Research 4:1075. https://doi.org/10.12688/f1000research.7201.1 Johnson M, Zaretskaya I, Raytselis Y et al (2008) NCBI BLAST: a better web interface. Nucleic Acids Res 36:W5–W9. https://doi.org/10.1093/nar/gkn201 Karsch-Mizrachi I, Nakamura Y, Cochrane G (2012) The international nucleotide sequence database collaboration. Nucleic Acids Res 40:D33–D37. https://doi.org/10.1093/nar/gkr1006 Keller MW, Rambo-Martin BL, Wilson MM, et al (2018) Direct RNA sequencing of the complete Influenza A virus genome. bioRxiv. https://doi.org/10.1101/300384 Klumpp J, Lavigne R, Loessner MJ, Ackermann H-W (2010) The SPO1-related bacteriophages. Arch Virol 155:1547–1561. https://doi.org/10.1007/s00705-010-0783-0 Klumpp J, Fouts DE, Sozhamannan S (2012) Next generation sequencing technologies and the changing landscape of phage genomics. Bacteriophage 2:190–199. https://doi.org/10.4161/ bact.22111 Klumpp J, Schmuki M, Sozhamannan S et al (2014) The odd one out: Bacillus ACT bacteriophage CP-51 exhibits unusual properties compared to related Spounavirinae W.Ph. and Bastille. Virology 462:299–308. https://doi.org/10.1016/j.virol.2014.06.012 Kot W, Vogensen FK, Sørensen SJ, Hansen LH (2014) DPS – a rapid method for genome sequencing of DNA-containing bacteriophages directly from a single plaque. J Virol Methods 196:152–156. https://doi.org/10.1016/j.jviromet.2013.10.040 Kristensen DM, Waller AS, Yamada T et al (2013) Orthologous gene clusters and taxon signature genes for viruses of prokaryotes. J Bacteriol 195:941–950. https://doi.org/10.1128/JB.0180112 Krupovic M, Prangishvili D, Hendrix RW, Bamford DH (2011) Genomics of bacterial and archaeal viruses: dynamics within the prokaryotic virosphere. Microbiol Mol Biol Rev 75:610–635. https://doi.org/10.1128/MMBR.00011-11 Labonté JM, Suttle CA (2013) Metagenomic and whole-genome analysis reveals new lineages of gokushoviruses and biogeographic separation in the sea. Front Microbiol 4:404. https://doi.org/ 10.3389/fmicb.2013.00404

642

S. V. Owen et al.

Labonté JM, Reid KE, Suttle CA (2009) Phylogenetic analysis indicates evolutionary diversity and environmental segregation of marine podovirus DNA polymerase gene sequences. Appl Environ Microbiol 75:3634–3640. https://doi.org/10.1128/AEM.02317-08 Laver T, Harrison J, O’Neill PA et al (2015) Assessing the performance of the Oxford Nanopore Technologies MinION. Biomol Detect Quantif 3:1–8. https://doi.org/10.1016/j. bdq.2015.02.001 Lee Y-J, Dai N, Walsh SE et al (2018) Identification and biosynthesis of thymidine hypermodifications in the genomic DNA of widespread bacterial viruses. Proc Natl Acad Sci. https://doi.org/10.1073/pnas.1714812115 Leinonen R, Akhtar R, Birney E et al (2011a) The European nucleotide archive. Nucleic Acids Res 39:31–34. https://doi.org/10.1093/nar/gkq967 Leinonen R, Sugawara H, Shumway M (2011b) The sequence read archive. Nucleic Acids Res 39:2010–2012. https://doi.org/10.1093/nar/gkq1019 Lima-Mendez G, Van Helden J, Toussaint A, Leplae R (2008) Prophinder: a computational tool for prophage prediction in prokaryotic genomes. Bioinformatics 24:863–865. https://doi.org/ 10.1093/bioinformatics/btn043 Loman NJ, Misra RV, Dallman TJ et al (2012) Performance comparison of benchtop highthroughput sequencing platforms. Nat Biotechnol 30:434–439. https://doi.org/10.1038/ nbt.2198 López-Bueno A, Tamames J, Velázquez D et al (2009) High diversity of the viral community from an Antarctic lake. Science 326:858–861. https://doi.org/10.1126/science.1179287 Manrique P, Bolduc B, Walk ST et al (2016) Healthy human gut phageome. Proc Natl Acad Sci 113:10400–10405. https://doi.org/10.1073/pnas.1601060113 Marine R, Polson SW, Ravel J et al (2011) Evaluation of a transposase protocol for rapid generation of shotgun high-throughput sequencing libraries from nanogram quantities of DNA. Appl Environ Microbiol 77:8071–8079. https://doi.org/10.1128/AEM.05610-11 Marston MF, Amrich CG (2009) Recombination and microdiversity in coastal marine cyanophages. Environ Microbiol 11:2893–2903. https://doi.org/10.1111/j.1462-2920.2009.02037.x McCallin S, Alam Sarker S, Barretto C et al (2013) Safety analysis of a Russian phage cocktail: from MetaGenomic analysis to oral application in healthy human subjects. Virology 443:187–196. https://doi.org/10.1016/j.virol.2013.05.022 Millard A, Clokie MRJ, Shub DA, Mann NH (2004) Genetic organization of the psbAD region in phages infecting marine Synechococcus strains. Proc Natl Acad Sci U S A 101:11007–11012. https://doi.org/10.1073/pnas.0401478101 Millard AD, Zwirglmaier K, Downey MJ et al (2009) Comparative genomics of marine cyanomyoviruses reveals the widespread occurrence of Synechococcus host genes localized to a hyperplastic region: implications for mechanisms of cyanophage evolution. Environ Microbiol 11:2370–2387. https://doi.org/10.1111/j.1462-2920.2009.01966.x Millard AD, Gierga G, Clokie MRJ et al (2010) An antisense RNA in a lytic cyanophage links psbA to a gene encoding a homing endonuclease. ISME J 4:1121–1135. https://doi.org/10.1038/ ismej.2010.43 Minot S, Sinha R, Chen J et al (2011) The human gut virome: inter-individual variation and dynamic response to diet. Genome Res 21:1616–1625. https://doi.org/10.1101/gr.122705.111 Mizuno CM, Rodriguez-Valera F, Kimes NE, Ghai R (2013) Expanding the marine virosphere using metagenomics. PLoS Genet 9:e1003987. https://doi.org/10.1371/journal.pgen.1003987 Moustafa A, Xie C, Kirkness E et al (2017) The blood DNA virome in 8,000 humans. PLoS Pathog 13:e1006292. https://doi.org/10.1371/journal.ppat.1006292 Owen SV, Wenner N, Canals R et al (2017) Characterization of the prophage repertoire of African Salmonella Typhimurium ST313 reveals high levels of spontaneous induction of novel phage BTP1. Front Microbiol 8:235. https://doi.org/10.3389/fmicb.2017.00235 Perez Sepulveda B, Redgwell T, Rihtman B et al (2016) Marine phage genomics: the tip of the iceberg. FEMS Microbiol Lett 363:fnw158. https://doi.org/10.1093/femsle/fnw158 Phillippy AM (2017) New advances in sequence assembly. Genome Res 27:xi–xiii. https://doi.org/ 10.1101/gr.223057.117

Detection of Bacteriophages: Sequence-Based Systems

643

Pride DT, Salzman J, Haynes M et al (2012) Evidence of a robust resident bacteriophage population revealed through analysis of the human salivary virome. ISME J 6:915–926. https://doi.org/ 10.1038/ismej.2011.169 Puxty RJ, Perez-Sepulveda B, Rihtman B et al (2015) Spontaneous deletion of an “ORFanage” region facilitates host adaptation in a “photosynthetic” cyanophage. PLoS One 10:e0132642. https://doi.org/10.1371/journal.pone.0132642 Rand AC, Jain M, Eizenga JM et al (2017) Mapping DNA methylation with high-throughput nanopore sequencing. Nat Methods 14:411–413. https://doi.org/10.1038/nmeth.4189 Raya R, Hébert EM (2009) Isolation of phage via induction of lysogens. In: Clokie MR, Kropinski AM (eds) Bacteriophages: methods and protocols. Humana Press, New York, pp 23–32 Reyes A, Haynes M, Hanson N et al (2010) Viruses in the faecal microbiota of monozygotic twins and their mothers. Nature 466:334–338. https://doi.org/10.1038/nature09199 Reyes A, Semenkovich NP, Whiteson K et al (2012) Going viral: next-generation sequencing applied to phage populations in the human gut. Nat Rev Microbiol 10:607–617. https://doi.org/ 10.1038/nrmicro2853 Reyes A, Blanton LV, Cao S et al (2015) Gut DNA viromes of Malawian twins discordant for severe acute malnutrition. Proc Natl Acad Sci 112:201514285. https://doi.org/10.1073/ pnas.1514285112 Rhoads A, Au KF (2015) PacBio sequencing and its applications. Genomics Proteomics Bioinform 13:278–289. https://doi.org/10.1016/j.gpb.2015.08.002 Rihtman B, Meaden S, Clokie MRJ et al (2016) Assessing Illumina technology for the highthroughput sequencing of bacteriophage genomes. PeerJ 4:e2055. https://doi.org/10.7717/ peerj.2055 Rohwer F, Edwards R (2002) The Phage Proteomic Tree: a genome-based taxonomy for phage. J Bacteriol 184:4529–4535. https://doi.org/10.1128/JB.184.16.4529 Roux S, Enault F, Robin A et al (2012) Assessing the diversity and specificity of two freshwater viral communities through metagenomics. PLoS One 7:e33641. https://doi.org/10.1371/journal. pone.0033641 Roux S, Tournayre J, Mahul A et al (2014) Metavir 2: new tools for viral metagenome comparison and assembled virome analysis. BMC Bioinform 15:76. https://doi.org/10.1186/1471-2105-1576 Roux S, Enault F, Hurwitz BL, Sullivan MB (2015) VirSorter: mining viral signal from microbial genomic data. PeerJ 3:e985. https://doi.org/10.7717/peerj.985 Roux S, Brum JR, Dutilh BE et al (2016) Ecogenomics and potential biogeochemical impacts of globally abundant ocean viruses. Nature 537:689–693. https://doi.org/10.1038/nature19366 Sabehi G, Shaulov L, Silver DH et al (2012) A novel lineage of myoviruses infecting cyanobacteria is widespread in the oceans. Proc Natl Acad Sci U S A 109:2037–2042. https://doi.org/10.1073/ pnas.1115467109 Sanger F, Nicklen S, Coulson AR (1977) DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci 74:5463–5467. https://doi.org/10.1073/pnas.74.12.5463 Sanger F, Coulson AR, Hong GF et al (1982) Nucleotide sequence of bacteriophage lambda DNA. J Mol Biol 162:729–773 Santiago-Rodriguez TM, Ly M, Bonilla N, Pride DT (2015) The human urine virome in association with urinary tract infections. Front Microbiol 6:1–12. https://doi.org/10.3389/fmicb.2015.00014 Schloss PD, Westcott SL, Ryabin T et al (2009) Introducing mothur: open-source, platformindependent, community-supported software for describing and comparing microbial communities. Appl Environ Microbiol 75:7537–7541. https://doi.org/10.1128/AEM.01541-09 Short CM, Suttle CA (2005) Nearly identical bacteriophage structural gene sequences are widely distributed in both marine and freshwater environments. Appl Environ Microbiol 71:480–486. https://doi.org/10.1128/AEM.71.1.480-486.2005 Skvortsov T, de Leeuwe C, Quinn JP et al (2016) Metagenomic characterisation of the viral community of Lough Neagh, the largest freshwater lake in Ireland. PLoS One 11:e0150361. https://doi.org/10.1371/journal.pone.0150361

644

S. V. Owen et al.

Sullivan MB (2015) Viromes, not gene markers, for studying double-stranded DNA virus communities. J Virol 89:2459–2461. https://doi.org/10.1128/JVI.03289-14 Sullivan MB, Lindell D, Lee JA et al (2006) Prevalence and evolution of core photosystem II genes in marine cyanobacterial viruses and their hosts. PLoS Biol 4:e234. https://doi.org/10.1371/ journal.pbio.0040234 Sullivan MB, Krastins B, Hughes JL et al (2009) The genome and structural proteome of an ocean siphovirus: a new window into the cyanobacterial “mobilome”. Environ Microbiol 11:2935–2951. https://doi.org/10.1111/j.1462-2920.2009.02081.x Sullivan MB, Weitz JS, Wilhelm S (2017) Viral ecology comes of age. Environ Microbiol Rep 9:33–35. https://doi.org/10.1111/1758-2229.12504 Székely AJ, Breitbart M (2016) Single-stranded DNA phages: from early molecular biology tools to recent revolutions in environmental microbiology. FEMS Microbiol Lett 363:1–9. https://doi. org/10.1093/femsle/fnw027 Thompson LR, Zeng Q, Kelly L et al (2011) Phage auxiliary metabolic genes and the redirection of cyanobacterial host carbon metabolism. Proc Natl Acad Sci 108:E757–E764. https://doi.org/ 10.1073/pnas.1102164108 Tseng E, Underwood JG (2013) Full length cDNA sequencing on the PacBio ® RS. J Biomol Tech 24:S45 Utter B, Deutsch DR, Schuch R et al (2014) Beyond the chromosome: the prevalence of unique extra-chromosomal bacteriophages with integrated virulence genes in pathogenic Staphylococcus aureus. PLoS One 9:e100502. https://doi.org/10.1371/journal.pone.0100502 Vega Thurber R, Haynes M, Breitbart M et al (2009) Laboratory procedures to generate viral metagenomes. Nat Protoc 4:470–483. https://doi.org/10.1038/nprot.2009.10 Wang K, Chen F (2008) Prevalence of highly host-specific cyanophages in the estuarine environment. Environ Microbiol 10:300–312. https://doi.org/10.1111/j.14622920.2007.01452.x Wang G, Hayashi M, Saito M et al (2009a) Survey of major capsid genes (g23) of T4-type bacteriophages in Japanese paddy field soils. Soil Biol Biochem 41:13–20. https://doi.org/ 10.1016/j.soilbio.2008.07.008 Wang G, Murase J, Taki K et al (2009b) Changes in major capsid genes (g23) of T4-type bacteriophages with soil depth in two Japanese rice fields. Biol Fertil Soils 45:521–529. https://doi.org/10.1007/s00374-009-0362-2 Zablocki O, Van Zyl L, Adriaenssens EM et al (2014) High-level diversity of tailed phages, eukaryote-associated viruses and virophage-like elements in the metaviromes of Antarctic soils. Appl Environ Microbiol 80:6888–6897. https://doi.org/10.1128/AEM.01525-14 Zablocki O, Adriaenssens EM, Frossard A et al (2017) Metaviromes of extracellular soil viruses along a Namib Desert aridity gradient. Genome Announc 5:e01470–e01416. https://doi.org/ 10.1128/genomeA.01470-16 Zeidner G, Preston CM, Delong EF et al (2003) Molecular diversity among marine picophytoplankton as revealed by psbA analyses. Environ Microbiol 5:212–216. https://doi.org/10.1046/ j.1462-2920.2003.00403.x Zhong Y, Chen F, Wilhelm SW et al (2002) Phylogenetic diversity of marine cyanophage isolates and natural virus communities as revealed by sequences of viral capsid assembly protein gene g20. Appl Environ Microbiol 68:1576–1584. https://doi.org/10.1128/AEM.68.4.15761584.2002 Zhong X, Guidoni B, Jacas L, Jacquet S (2015) Structure and diversity of ssDNA Microviridae viruses in two peri-alpine lakes (Annecy and Bourget, France). Res Microbiol 166:644–654. https://doi.org/10.1016/j.resmic.2015.07.003 Zhou Y, Liang Y, Lynch KH et al (2011) PHAST: a fast phage search tool. Nucleic Acids Res 39: W347–W352. https://doi.org/10.1093/nar/gkr485 Zschach H, Joensen KG, Lindhard B et al (2015) What can we learn from a metagenomic analysis of a Georgian bacteriophage cocktail? Viruses 7:6570–6589. https://doi.org/10.3390/v7122958

Novel Approaches for Detection of Bacteriophage Carrie L. Pierce, Jon C. Rees, and John R. Barr

Contents Mass Spectrometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Raman Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

645 650 654 655 655

Abstract

Modern analytical instrumentation combined with increased computational power has dramatically enhanced the capacity to interrogate the physicochemical properties of macromolecular structures. For example, it has become routine to analyze complex protein mixtures by mass spectrometry and quantitatively identify thousands of proteins in the nanomolar to picomolar range in a single analytical run. The relatively small size and simple structure of bacteriophages, generally consisting of a protein coat surrounding a nucleic acid core, make them highly amenable to examination by these new analytical methods, and assays can be developed that exploit inherent bacteriophage functionality. This chapter describes the recent utilization of bacteriophages in conjunction with mass spectrometry and Raman spectroscopy for direct detection of bacteriophages.

C. L. Pierce (*) · J. C. Rees · J. R. Barr Division of Laboratory Sciences, National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, GA, USA e-mail: [email protected]; [email protected]; [email protected]; [email protected] © This is a U.S. Government work and not under copyright protection in the US; foreign copyright protection may apply 2021 D. R. Harper et al. (eds.), Bacteriophages, https://doi.org/10.1007/978-3-319-41986-2_20

645

646

C. L. Pierce et al.

Mass Spectrometry Mass spectrometry (MS) was first used in 1897 by J.J. Thomson to provide information on the chemical composition of a sample based on the mass-to-charge ratio (m/z) of its ionized components. Following the work of Thomson, a century of development ensued wherein mass spectrometry became adept at measuring the mass of small molecules (< m/z 400 amu) and eliciting structural information based upon fragmentation analysis. However, large molecules such as peptides and proteins were difficult to analyze because of the challenges of ionizing these molecules intact into the gas phase, a necessary condition for mass spectrometric analysis. The late 1980s and throughout the 1990s saw a revolution in mass spectrometry, as two new techniques, matrix-assisted laser desorption ionization time of flight (MALDI-TOF) (Karas and Hillenkamp 1988) and electrospray ionization (ESI) (Fenn et al. 1989) mass spectrometry allowed for the mass determination of high molecular weight peptides and intact proteins. These new mass spectrometry techniques were so impactful that the inventors of the respective techniques were awarded the 2002 Nobel Prize in Chemistry, and their advent, along with techniques for database searching (Yates et al. 1993), spawned the proteomic revolution. Currently it is possible to routinely identify and determine the relative abundance of hundreds of proteins in a sample using modern mass spectrometry techniques. Both MALDI-TOF and ESI mass spectrometry have been utilized to analyze bacteriophages. Bacteriophage proteins were initially analyzed with these modern techniques via MALDI-TOF MS, which consists of a simple and straightforward sample preparation followed by rapid mass spectrometry analysis. In MALDI, a sample containing intact bacteriophage is co-crystalized with a matrix containing an aromatic organic acid dissolved in an aqueous/acetonitrile/acidic solution and spotted onto a stainlesssteel plate. After drying on the stainless-steel MALDI plate, the proteins of the intact bacteriophage have presumably disassembled and become co-crystalized in the MALDI matrix. The plate is then introduced into the mass spectrometer where the sample spot is irradiated by a pulsed laser, which desorbs and ionizes the intact neutral bacteriophage proteins off of the surface of the plate. Once ionized and in the gas phase, a high voltage is applied and the instrument optics guide the proteins through a mass analyzer, usually a time-of-flight tube, toward a detector where the ion current can be measured. Fenselau and colleagues presented the first study using MALDI-TOF MS for characterization of intact bacteriophage (Thomas et al. 1998), showing that the major capsid protein at 13.7 kDa of the coliphage MS2 could be readily ionized and detected using MALDI-TOF MS with minimal sample preparation. The MS2 coat protein was detected at 2 fmol in culture medium, permitting rapid, direct, and sensitive bacteriophage classification. Figure 1 shows the MS2 spectrum collected by the Fenselau group. Building on the work of Fenselau, Madonna et al. presented methodology utilizing bacteriophage amplification detection (PAD) combined with MALDITOF MS to imply the presence of a bacterium in culture (Madonna et al. 2003).

Novel Approaches for Detection of Bacteriophage

13787

100 80 Relative Intensity

Fig. 1 MALDI mass spectrum from MS2 bacteriophage. Optimized spectra were obtained using R-cyano-4-hydroxycinnamic acid matrix following pretreatment with acetic acid (1:1, v/v) for protonation, fragmentation, and ionization of the analyte (Reprinted (adapted) with permission from Thomas et al. (1998). Copyright 1998 American Chemical Society)

647

60 40 20 0 4000

6000

8000

10000 M/Z

12000

14000

Following immunomagnetic separation (IMS) to isolate the Escherichia coli from a complex mixture, Madonna added MS2 at a concentration below the detection limit of the mass spectrometer and incubated the phage-bacterium mixture at a suitable temperature. After allowing time for the phage amplification process to progress, MALDI-TOF MS analysis unambiguously detected the increase in the major MS2 capsid protein concentration, thereby implicating the presence of E. coli in the sample. While specific E. coli associated protein peaks could be detected by MALDI-TOF MS at 106 CFU mL 1, the use of PAD using MS2 improved the MALDI-TOF MS limits of detection by two orders of magnitude, allowing 104 CFU mL 1 detection in In some instances, the use of acid cleaners is required for surfaces bearing precipitated minerals or with high levels of food residues/minerals. Nonionic wetting agents are used due to their ability to control foaming and their good emulsifying properties. Depending on water hardness, sequestrants can be used to chelate minerals (Chmielewski and Frank 2003). It is very important that these agents, used during the cleaning process, disrupt or dissolve the EPS matrix in order to allow the access of disinfectants to the viable cells (Simões et al. 2010). The type of chemical agents used, the contact time, the temperature, and the application of mechanical forces are variables that together can greatly influence the efficiency of cleaning (Van Houdt and Michiels 2010). Cleaning procedures can remove most of the microorganisms associated with a surface, but are not necessarily able to kill them. Therefore, removed microorganisms can reattach to other surfaces and initiate the formation of novel biofilms (Simões et al. 2010; Srey et al. 2013). Antimicrobial agents are used during disinfection procedures, which are often performed after cleaning, with the aim of killing the microorganisms. They are applied as a liquid spray to kill microorganisms on surfaces or, alternatively, as a fine mist to kill airborne microorganisms (Coughlan et al. 2016). The presence of organic substances (fat, carbohydrates, and protein based materials), water hardness, temperature, pH, time of application, and degree of physical contact with the microorganisms are limiting factors for the effectiveness of the antimicrobial agents (Simões et al. 2010; Srey et al. 2013). A wide range of disinfectants are used in the food industry, including chlorine, quaternary ammonium compounds (QAC), hydrogen peroxide, peracetic acid, ozone, phosphoric acid, sulfamic acid, and acid blends (Chmielewski and Frank 2003). Although cleaning and disinfection procedures can be efficient in removing and then killing bacteria before they irreversibly attach to a surface, elimination of mature biofilms is a more difficult task. Interestingly, the selection of antimicrobial agents to be used tends to be based on studies performed with planktonic cells, though it is common sense that biofilm cells have an increased resistance to their action (Simões et al. 2010; Coughlan et al. 2016). Several studies have reported lower susceptibility of biofilms formed by spoilage and pathogenic bacteria to common disinfectants used in the food industry. For instance, Corcoran et al. assessed the activities of sodium hypochlorite, sodium hydroxide, and benzalkonium chloride against Salmonella enterica biofilms from food contact surface materials, and none of the three antimicrobial agents eradicated the biofilms (Corcoran et al. 2014). Other studies also have reported the resistance of Salmonella biofilm cells to disinfectants commonly used in the food industry (Joseph et al. 2001). Pan et al. (2006) used a hydrogen peroxide-based agent against L. monocytogenes biofilms and found that it was ineffective in their elimination. These findings suggest that there is a need of new strategies to control biofilms, particularly in the food industry

Biofilm Applications of Bacteriophages

799

where biofilms are associated with spoilage food, economic losses, and foodborne illnesses/outbreaks. Current strategies used to manage foodborne pathogenic bacteria can also be done using physical disruption, including heat, steam, and UV-light irradiation to non-selectively reduce the bacterial loads on food contact surfaces (Stanfield 2003; Delaquis and Bach 2012). Another food processing technique is high hydrostatic pressure (HHP) that works between pressure ranges of 100 and 1200 MPa and has great effectiveness in inactivating microorganisms (Chawla et al. 2011). HHP was first reported to have effect on foodborne microorganisms in 1899 when milk was subjected to 650 MPa for 10 min at room temperature (Hite 1899). There are many chemical-, mechanical-, thermal-, and pressure-mediated systems in use, and in addition to being effective, the current products and procedures should be safe, easy to use, and economical. Some of the technologies in use can, however, have significant downsides, such as corrosion of equipment, toxic chemical residues, and damage to the quality of foods. For instance, chlorine-based compounds have to be rinsed off from the surfaces in order to eliminate toxic products that can affect the final properties of the foods (Sampathkumar et al. 2005; Simões et al. 2010).

Bacteriophage Application to Biofilms and Other Surface-Attached Bacteria Formed in the Food Industry Bacterial biofilms can grow in virtually any industrial setting, causing damage to working surfaces, equipment, and food products. In this section, we discuss the application of bacteriophages to the different surfaces depicted in Fig. 1 with the main outcomes of these studies detailed in separate subsections.

Application of Bacteriophages to Equipment Surfaces and to Working Surfaces and to Enhance Equipment Performance Biofilm formation, which starts with an initial reversible bacterial attachment, may cause several equipment operating troubles, including decreased heat transfer, blocking of tubes, and plugging of filters. Biofilm formation may also be related to damage of surfaces (pitting and corrosion), which most likely leads to further microbial growth on those surfaces. The presence of biofilms on improperly disinfected and sanitized working surfaces (reviewed in Giaouris et al. 2014 and Coughlan et al. 2016) can lead to food product contamination and spoilage, which causes great monetary losses for the food industry and represents a risk to consumer health. Therefore, good disinfection procedures must be applied within processing plants, and bacteriophages present themselves as potential disinfection tools since they are effective in the control and eradication of biofilms, are nontoxic to humans, and are also considered to be environmentally sound (“green”) (see chapter ▶ “Bacteriophage as Biocontrol Agents”). Table 1

800

C. Milho et al.

Table 1 Bacteriophage publications against biofilms in working and equipment surfaces Working and equipment surfaces Bacteriophage or bacteriophage Treatment efficacy (reduction of Bacteria; surface cocktail viable cells) Bacteriophage application to control biofilm formation P. fluorescens; stainless Phi-S1 Biomass reduction of 85% using steel MOI of 0.5 P. fluorescens; stainless PhiIBB-PF7A Biomass removal varied between steel 63 and 91% depending on biofilm age and infection conditions P. fluorescens and PhiIBB-PF7A and Significant reduction after 2 h of S. lentus; stainless steel phiIBB-SL58B treatment; bacteriophage ΦIBBPF7A reduced cell counts by 3 log10 in dual species biofilms E. coli O157:H7; Cocktail of Reduction of 4.5 log10 after 2 h of spinach harvester blades 5 bacteriophages incubation E. coli O157:H7; BEC8 cocktail No E. coli O157:H7 cells detected stainless steel and after 10 min of incubation with ceramic tile MOI of 100 at 37 ºC S. Kentucky and S. SalmoFresh™ Decrease on number of cells by Brandenburg; stainless cocktail 2.1–4.3 log10 steel and glass L. monocytogenes; LiMN4L, LiMN4p On both surfaces, single stainless steel coupons and LiMN17 bacteriophages reduced cells by with or without a fish individually or in 3–4.5 log10, and cocktail by 3.8–5.4 log10 broth layer cocktail V. alginolyticus; VP01 Cell counts were reduced by polystyrene 56 and 86%, when using bacteriophage concentrations of 1010 and 1012 L. monocytogenes; H387, H387-A, The efficiency of combined stainless steel and and 2671in agents was higher compared to polypropylene cylinders combination with their use individually QUATAL P. aeruginosa; glass Cocktail of RNA Removal of 97% of P. aeruginosa coupons bacteriophages biofilms after combined combined with treatment. The products alone chlorine reduced biofilm by 89% (cocktail) and 40% (chlorine) P. aeruginosa; water and Cocktail of RNA After 1 h dosing on day wastewater filtration bacteriophages 136, removal increased by 56% systems and 70%, in anthracite and granular activated biofilters D. tsuruhatensis; DTP1 Treatment increased the flux by wastewater filtration 70% of the original system

Reference Sillankorva et al. (2004) Sillankorva et al. (2008) Sillankorva et al. (2010)

Patel et al. (2011) Viazis et al. (2011) Woolston et al. (2013) Arachchi et al. (2013)

Sasikala and Srinivasan (2016) Roy et al. (1993)

Zhang and Hu (2013)

Zhang et al. (2013)

Bhattacharjee et al. (2015)

summarizes all bacteriophage work carried out against biofilms that mimic working and equipment surfaces and bacteriophage application to filtration systems to enhance their performance.

Biofilm Applications of Bacteriophages

801

Outbreaks linked to consumption of spinach have raised awareness that during harvesting, harvester blades can potentially come into contact with fecal matter (Centers for Disease Control and Prevention 2006). Pathogens in fecal matter or improperly composted manure may attach to the harvester blades and potentiate biofilm formation. To study this, bacteriophages specific for E. coli O157:H7 were used on biofilms formed on spinach harvester blades. The blades were immersed in a cocktail of five bacteriophages (Patel et al. 2011) and after 2 h reduced bacterial loads on the blades by 4.5 log10. The action of bacteriophages against biofilms formed on food working surface materials such as stainless steel, ceramic tile, and glass has been studied (see Fig. 5). For instance, two bacteriophages specific for Pseudomonas fluorescens (phi-S1 and phiIBB-PF7A) showed good biomass removal efficiencies at optimal conditions (Sillankorva et al. 2004, 2008). Combining phiIBB-PF7A with a bacteriophage specific for Staphylococcus lentus, phiIBBSL58B, significantly decreased dual-species biofilm populations after 2 h of bacteriophage application (Sillankorva et al. 2010). Furthermore, to test the hypothesis if a single bacteriophage could reach its host and cause destruction in the presence of a non-specific host, the P. fluorescens bacteriophage ΦIBB-PF7A was added to dualspecies biofilms. This approach not only reduced viable cell counts but also caused detachment of the non-susceptible host to the planktonic phase. In another study,

Fig. 5 S. Enteritidis adhesion and biofilm formation on stainless steel (a-c) and treatment with bacteriophage phi38 (unpublished data). (a) Cell adhesion, (b) proliferation, (c) microcolonies, and (d-f) biofilms treated with bacteriophage phi38 at a multiplicity of infection of 1; some cells are still present but mainly the surface is covered with cell debris

802

C. Milho et al.

stainless steel, ceramic tile, and high-density polyethylene chips were artificially contaminated with E. coli O157:H7 and then challenged with a bacteriophage mixture called BEC8 at different multiplicities of infection (MOI) and also at different temperatures (Viazis et al. 2011). At temperatures above 12  C and a MOI of 100, no surviving cells were detected after 10 to 60 min of incubation. Another study using bacteriophages against biofilms formed on food contact surfaces (stainless steel and glass surfaces) was carried out with the bacteriophage cocktail SalmoFresh™, consisting of six lytic bacteriophages specific for Salmonella. In this experiment, SalmoFresh™ was capable of significantly reducing the number of Salmonella Kentucky (Salmonella enterica subsp. enterica serovar Kentucky) and S. Brandenburg (Salmonella enterica subsp. enterica serovar Brandenburg) attached to the surfaces (Woolston et al. 2013). The presence of L. monocytogenes biofilms on food working surfaces is also a great concern to the food industry. Hence, many studies have been done with bacteriophages regarding elimination of L. monocytogenes biofilm. The effectiveness of three L. monocytogenes specific bacteriophages (LiMN4L, LiMN4p, and LiMN17) was tested individually or as a cocktail against L. monocytogenes biofilms grown on stainless steel coupons with or without a fish broth layer (Arachchi et al. 2013). Treatment of both surfaces with single bacteriophages reduced bacterial cells, though the use of the bacteriophage cocktail led to higher reductions on both surfaces. Steel present in food production facilities can support reservoirs of pathogenic bacteria. Listex™ P100 bacteriophage was used in the control of L. monocytogenes biofilms formed on steel wafers, and the end result was complete elimination of the biofilms from these surfaces (Iacumin et al. 2016). L. monocytogenes bacteriophages have also been tested in combination with QUATAL (10.5% N-alkyl dimethyl-benzylammonium HCL and 5.5% glutaraldehyde), and the combination of both agents produced higher antimicrobial effect (Roy et al. 1993). Aquaculture steel tanks are vulnerable to the formation of biofilms. In fact, the presence of Vibrio spp. attached to the steel tanks is a frequent and major concern since many species from this genus are pathogens for the animals grown in these tanks and also for the consumer (Haldar 2012). For example, V. alginolyticus is an important opportunistic pathogen that causes disease in humans and frequently contaminates marine animals (Mustapha et al. 2013). Because it can form biofilms, V. alginolyticus becomes even more difficult to eradicate, and bacteriophages can serve as alternative sanitizing agents. Sasikala and Srinivasan (2016) used bacteriophage VP01 to disrupt V. alginolyticus biofilms. The choice of surface material was not steel. Nonetheless, on polystyrene surfaces, a significant reduction in cell count numbers of 56% and 86% was achieved using bacteriophage concentrations of 1010 and 1012, respectively. Little work with bacteriophages has been carried out on biofilm removal from equipment surfaces. To our knowledge, the only reports are related to the use of bacteriophages in water and wastewater filtration systems to control biofouling. These filtration systems are often impaired by biofilm formation. In one study, anthracite and granular activated carbon biofilter systems were artificially

Biofilm Applications of Bacteriophages

803

contaminated with Pseudomonas aeruginosa (Zhang et al. 2013). A mixture of bacteriophages specific for P. aeruginosa was applied only once, on day 136 of this study, and the efficiency of P. aeruginosa removal from filters increased by 56% and 70%, in the anthracite and in the granular activated biofilters, respectively. The bacterium, Delftia tsuruhatensis, isolated from a wastewater treatment plant, was used to form biofilms on a laboratory-scale membrane bioreactor that simulated biofouling on a membrane bioreactor treating wastewater (Bhattacharjee et al. 2015). Biofilm formation was responsible for a water flux on the bioreactor decrease of about 31% of its capacity. In this case, the use of a lytic bacteriophage specific for D. tsuruhatensis led to a flux increase of 70% of the original.

Application of Bacteriophages to Foods Biofilms containing spoilage and pathogenic microorganisms can lead to serious problems, compromising the quality and safety of food products. Foods can become contaminated at any stage of the food chain, and cleaning and disinfection procedures may not be effective in the elimination of microbial contaminants (Sampathkumar et al. 2005; Pereira and Melo 2009; Gutiérrez et al. 2016). The use of bacteriophages is one of the alternatives to the traditionally used methods in the elimination of biofilms in food settings. Several studies have been conducted using bacteriophages directly on food products (see chapter ▶ “Food Safety”). Even though the majority of the studies that are addressed in this section do not imply an application of bacteriophages to control and remove biofilm structures, many studies were included because the presence of surface-attached bacteria may be inferred. The criteria for inclusion of the works were (i) works performed with adhered cells which display no increase of cells throughout time in the untreated control samples since these cells, which probably mimic biofilm detached cells, persist attached to surfaces and are involved in many cross-contamination events and many times can themselves lead to the development of biofilms and (ii) works where the authors have in vitro contaminated their products with bacteria in the absence of any remaining liquid media that could promote their planktonic growth and where the untreated samples (controls) register an increase in the number of cells throughout time (i.e., if there is insufficient liquid within which bacteria can be planktonic, then we assume that the bacteria are not planktonic). We have assumed, although no microscopy studies are presented by the authors, that these might be biofilm-related (Fig. 6).

Fresh-Cut Fruits and Vegetables Fresh-cut fruit and vegetable products are widely available packaged items and can be already trimmed, peeled, and/or cut offering consumers a high nutrition product that keeps its flavor and freshness. Major focus has been given to E. coli O157:H7 (Table 2), a major foodborne pathogen which is commonly found contaminating fresh produce. Many of these studies were carried out on in vitro contaminated foods where the bacteria remained adhered and did not increase in concentration

804

C. Milho et al.

Fig. 6 Target bacterial cell types present in foods and surfaces that have been challenged with bacteriophages

throughout the experiments. Some commercially available bacteriophage products have been tested in fresh-cut fruits and vegetables. This is the case of EcoShield™ and ListShield™, two bacteriophage mixtures available from Intralytix, Inc., designed for treating foods that are at risk of E. coli O157:H7 and L. monocytogenes contamination. The latter can also be used to reduce L. monocytogenes levels from non-food contact equipment, surfaces, etc. in the food processing plants and other food establishments. In a study conducted by Carter et al. (2012), EcoShield™ was used for reduction of E.coli on lettuce, and it was observed that this product significantly reduced bacterial loads by 87% after only 5 min of contact. In another study, EcoShield™ was applied on fresh-cut leafy greens, under atmospheric air or modified atmosphere ((5% O2, 35%, CO2, and 60% N2) (Boyacioglu et al. 2013). Although this bacteriophage cocktail was effective in reducing E. coli O157:H7 levels on fresh-cut lettuce and spinach under atmospheric air, this reduction was even greater when EcoShield™ was applied under modified atmosphere conditions. Hong et al. (2014) used a mixture of three bacteriophages to reduce E. coli O157:H7 on the surface of spinach leaves. They observed a reduction of 3.28, 2.88, and 2.77 log10

Biofilm Applications of Bacteriophages

805

Table 2 Bacteriophage publications for inhibiting cell adhesion and consequent biofilm formation in fresh-cut fruits and vegetables Fresh-cut fruits and vegetables Bacteriophage or bacteriophage Bacteria; surface cocktail Bacteriophage application to adhered cells E. coli O157:H7; vB_EcoS_FFH_1, spinach leaves 2 and 3 cocktail

Treatment efficacy (reduction of viable cells)

Reduction of 3.28, 2.88, and 2.77 log10 at RT for 24, 48, and 72 h, respectively E. coli O157:H7; OSY-SP Decrease of 2.4–3.0 log10, on cut green pepper, and 3.4–3.5 cut green pepper log10, on spinach leaves, during and spinach leaves 72 h storage E. coli O157:H7; EcoShield™ cocktail Reduction of 87% after 5 min of lettuce contact E. coli O157:H7; EcoShield™ cocktail At 4 and 10 ºC reductions in the leafy greens range of 1.19-4.34 log10 under atmospheric air and modified atmosphere L. monocytogenes; ListShield™ cocktail Reduction of 1 log10 on both produces lettuce and apple Bacteriophage application to prevent biofilm formation L. monocytogenes; LM-103 and Reduction by 2–4.6 log10 using honeydew melon LMP-102 cocktails bacteriophage; combination of alone or in bacteriophage with nisin combination with reduced counts up to 5.7 log10 on honeydew melon slices nisin L. monocytogenes; A511 and Listex™ Decrease of 2.3 log10 and approx. 5 log10 in cabbage and cabbage, lettuce P100 lettuce leaves E. coli O157:H7; HY01 Reduction >2 log10, after 2 h of treatment cabbage leaves S. Enteritidis; SCPLX-1 cocktail Reduction of 3.5 log10 at honeydew melon temperatures ranging from 5 to 20 ºC

Reference Hong et al. (2014) Snyder et al. (2016)

Carter et al. (2012) Boyacioglu et al. (2013)

Perera et al. (2015) Leverentz et al. (2003)

Guenther et al. (2009) Lee et al. (2016) Leverentz et al. (2001)

CFU/ml when spinach was stored at room temperature for 24, 48, and 72 h, respectively. Cut green pepper and spinach leaves were artificially contaminated with E. coli O157:H7 and challenged with bacteriophage OSY-SP, which lead to a decrease of the bacterial load by 2.4–3.0 log10 CFU/g on cut green pepper and 3.4–3.5 log10 CFU/g on spinach leaves (Snyder et al. 2016). Also, reduction of adhered L. monocytogenes on lettuce and apples was attempted with ListShield™, a commercially available bacteriophage cocktail (Intralytix, Inc.) (Perera et al. 2015). The antibacterial effect with ListShield™, reduced L. monocytogenes loads on both products by 1 log10. A few studies that initially start with bacteria spiked on foods register a high increase of bacterial numbers in the samples untreated with bacteriophage. Leverentz

806

C. Milho et al.

et al. (2001) used a mixture of four bacteriophages (SCPLX-1, Intralytix, Inc.) to reduce Salmonella Enteritidis population on the surface of several fresh-cut melons and apples, at different temperatures (5, 10, and 20  C). Although they observed a significant reduction of the bacterial load on melon slices, the same was not observed for apple slices, probably due the low pH of this fruit. For the elimination of L. monocytogenes also from the surface of melon and apple slices, the bacteriophage mixtures LM-103 and LMP-102, provided by Intralytix, Inc., were applied (Leverentz et al. 2003). Similar results to the study with SCPLX-1 cocktail were obtained, with the bacteriophage mixtures able to reduce the number of L. monocytogenes bacteria on the surface of melon but unable to do so on apple slices. Also of L. monocytogenes, broad host-range bacteriophages A511 and Listex™ regarding the elimination P100 were used on the surface of cabbage and lettuce (Guenther et al. 2009), and both bacteriophages caused greater than 2 log10 cell reductions. Bacteriophage HY01, capable of lysing E. coli O157:H7 and S. flexneri, was applied on cabbage leaves and was able to inhibit the food isolate strain E. coli O157: H7 ATCC 43895 for up 2 h. Bacterial recovery, however, was observed after 2 h of incubation (Lee et al. 2016).

Red and Poultry Meats Another important source of food-related outbreaks is meat (poultry, pig, and bovine) (European Food Safety Authority 2015). Campylobacter, E. coli, and Salmonella are among the most frequently recovered pathogens from poultry skin and meat and therefore a focus of many bacteriophage studies (Table 3). Bacteriophages have been applied on artificially C. jejuni contaminated chicken pieces (frozen and unfrozen) and reduced adhered bacterial counts by approximately 1 log10 (Atterbury et al. 2003). In another study, bacteriophage 12,673 was applied on chicken skin portions, and C. jejuni counts were reduced by approximately 95% compared to controls, which continued to multiply on the skin (Goode et al. 2003) and possibly formed 3D biofilm structures. Bacteriophages have been used to control/eradicate adhered Salmonella enterica serovars Typhimurium and Enteritidis from different meat products. A cocktail of bacteriophages (Felix 01, ФSH17, ФSH18, and ФSH19) applied on pig skin samples at multiplicities of infection (MOI) of 10 or greater reduced S. Typhimurium (Salmonella enterica subsp. enterica serovar Typhimurium) present on pig skin samples up to 2 log10 during a period of 96 h (Hooton et al. 2011). In another S. Typhimurium bacteriophage study, hot dogs and sliced turkey samples were artificially contaminated and challenged with bacteriophage FO1-E2 (Guenther et al. 2012), and although counts dropped to undetectable levels after 2 days of incubation with FO1-E2, bacteriophage-resistant bacteria appeared after 6 days. A bacteriophage cocktail composed of UAB_Phi20, AB_Phi78, and UAB_Phi87 bacteriophages had its effectiveness tested in pig skin and chicken breasts artificially contaminated with S. enterica serovars Typhimurium and Enteritidis (Spricigo et al.

Biofilm Applications of Bacteriophages

807

Table 3 Bacteriophage works targeting adhered cells and biofilm prevention in red meat and poultry Red and poultry meats Bacteriophage or Bacteria; surface bacteriophage cocktail Bacteriophage application to adhered cells C. jejuni; chicken skin φ2

Salmonella Typhimurium; pig skin S. Typhimurium; hot dogs and sliced Turkey

S. Typhimurium and S. Enteritidis; chicken breasts S. Typhimurium, S. Heidelberg and S. Enteritidis; chicken meat and skin

Felix 01, ФSH17, ФSH18, and ФSH19 cocktail FO1-E2

UAB_Phix (x = 20, 78, 87) cocktail

Treatment efficacy (reduction of viable cells) Reduction by approximately 1 log10; greater reductions on frozen chicken skin pieces MOIs equal or greater than 10 reduced up to 2 log10, during a period of 96 h Reduction to undetectable levels after 2 days of incubation. Bacteriophage resistant bacteria appeared at day 6 Reduction of 2.2 and 0.9 log10, respectively, on chicken breasts

Reference Atterbury et al. (2003) Hooton et al. (2011) Guenther et al. (2012)

Spricigo et al. (2013)

Combined treatment reduced cell counts up to 5 log10, on chicken skin, and up to 1.3 log10, on chicken meat

Sukumaran et al. (2015)

Reduction of 3.04 log10, after 24 h storage at 8 ºC EcoShield™ cocktail Levels reduced by more than 94% E. coli O157:H7; beef T5, T1, T4 and O1 Alone or in cocktail caused individually or in significant reductions at all cocktail temperatures and MOIs tested L. monocytogenes; Listex™ P100 Treatment of L. monocytogenes roast beef and cooked combined with contaminated meat with Turkey potassium lactate and bacteriophage improved the sodium diacetate efficiency of the used chemical compounds L. monocytogenes; hot A511 and Listex™ Decrease of 2.2 log10 and 1.5 log10 of Scott A strain present in dogs, sliced cooked P100 hot dogs and sliced Turkey, Turkey respectively Bacteriophage application to prevent biofilm formation S. Typhimurium and S. UAB_Phix (x = 20, Reduction of 4 and 2 log10 of S. Enteritidis; pig skin 78, 87) cocktail Typhimurium and Enteritidis, and chicken breasts respectively, on pig skin C. jejuni; chicken skin NCTC 12673 Bacterial loads reduced by nearly 95% S. Typhimurium; BPSx (x = 2H1, 7T1, Undetectable numbers after 1 day chicken breasts 8H2, 11Q3, 11T1, of incubation 11T2, and 15Q2) cocktail

Kang et al. (2013) Carter et al. (2012) Liu et al. (2015)

S. Enteritidis; chicken skin E. coli O157:H7; beef

SalmoFresh™ combined with lauric arginate or cetylpyridinum chloride wksl13

Chibeu et al. (2013)

Guenther et al. (2009)

Spricigo et al. (2013) Goode et al. (2003) Han et al. (2017)

808

C. Milho et al.

2013). Significant bacterial reductions were obtained for both serovars on the two food products tested; however, the results are presented separately in Table 3 according to the behavior of the untreated samples. Bacteriophage wksl13, a broad-spectrum bacteriophage infecting S. Enteritidis (Salmonella enterica subsp. enterica serovar Enteritidis) and S. Typhimurium, was used for the control of S. Enteritidis on the surface of chicken skin samples (Kang et al. 2013). With a singledose application of this bacteriophage, bacterial counts were reduced in 3.04 log10 after 24 h of storage at 8  C, with no significant regrowth during the next 7 days. Unlike the other experiments described where cells only remained adhered and did not increase in number, more recently a study evaluated the antimicrobial activity of a bacteriophage cocktail, composed of seven bacteriophages against S. Typhimurium on chicken breasts (Han et al. 2017). At all tested temperatures, bacterial loads in untreated samples increased significantly while bacterial loads were reduced to undetectable, at temperatures above 8  C after 1 day in contact with the bacteriophage cocktail. EcoShield™ was successfully used to reduce the levels of E. coli O157:H7 in experimentally contaminated beef by more than 94% (Carter et al. 2012). Also, the efficacy of four E. coli O157:H7-specific bacteriophages, T5-like (T5, T1-like (T1), T4-like (T4), and O1-like (O1) was tested, individually or as a cocktail in beef samples (Liu et al. 2015). T5-like bacteriophage was the most efficient bacteriophage. Nevertheless, all bacteriophages were able to significantly reduce E. coli levels at all temperatures and multiplicities of infection tested.

Dairy Products The dairy industry is commonly affected by microbial contamination that results from inappropriate disinfection of equipment, with L. monocytogenes being one of the most common foodborne pathogens responsible for these contaminations (European Food Safety Authority 2015). ListShield™ and Listex™ P100 and A511 have been tested on soft and hard cheeses as well as mozzarella cheese brine, to reduce the counts of adhered bacteria. ListShield™ applied on experimentally contaminated hard cheese was capable of reducing bacterial contamination by 0.7 log after 5 min of treatment (Perera et al. 2015). Listex™ P100 was tested on the surface of Minas Frescal and Coalho hard cheeses and was responsible for a bacterial population reduction of 2.3 log10, in Minas Frescal cheese, and 2.1 log10, in Coalho cheese (Silva et al. 2014). Listex™ P100 and A511 were tested in mozzarella cheese brine individually, and both bacteriophages decreased L. monocytogenes more than 5 log10. L. monocytogenes targeted by Listex™ P100, applied on the surface of contaminated soft cheeses, reduced Listeria by at least 3.5 log10 to complete eradication (Carlton et al. 2005). In another study, different types of soft-ripened cheeses were artificially contaminated with L. monocytogenes and challenged with broad-host bacteriophage A551 (Guenther and Loessner 2011). This bacteriophage was able to eradicate bacterial cells from the cheese surfaces to levels under 1 CFU/cm2 (detection limit) when the initial contamination rate was of 102 CFU/cm2 or less, and when higher contamination rates were applied (103 CFU/cm2), a significant reduction of almost 3 log10 was observed.

Biofilm Applications of Bacteriophages

809

Single Bacteriophage Versus Cocktail Approach and Other Approaches Since industrial environments have a range of species distressing the food sector and since multispecies biofilm consortia are present (Flemming et al. 2016), in order to target a higher number of species, the use of bacteriophage cocktails should be considered. A major advantage of using a bacteriophage cocktail is that it can act on a broader range of hosts. Also, the use of bacteriophage cocktails decreases the probability of emergence of bacteriophage-resistant bacteria. Several studies have been performed using bacteriophages as biosanitizers, as presented above, but only a few describe the antibacterial effect of the individual bacteriophages in a cocktail versus the cocktail bacteriophage result. As previously mentioned, elimination of E. coli O157:H7 from beef has been tested with four bacteriophages individually and in cocktail (Liu et al. 2015). The T5-like bacteriophage had better efficacy than the cocktail in all experiments with a MOI of 1000 and 4, 22, and 37  C temperatures. Contrary to what was predicted, the cocktail had a lower efficacy which the authors suggested was due to competitive interference among bacteriophages. This result clearly shows that designing effective cocktails is challenging and if the bacteriophages are not tested individually, then the overall efficacy of the cocktail can be lower than the effect of one of the single bacteriophages present in the cocktail. In another work published by Arachchi et al. (2013), three bacteriophages, LiMN4L, LiMN4p, and LiMN17, were applied individually or in cocktail for the elimination of L. monocytogenes on stainless steel coupons. Contrary to the results from Liu et al. (2015), single bacteriophages reduced bacterial cells by 3–4.5 log10, and the cocktail was responsible for a much higher antibacterial effect (3.8–5.4 log10), which clearly reflects the differences between the bacteriophages used in the aforementioned work. Some studies have also combined bacteriophages with other agents. For instance, bacteriophage activity combined with quaternary ammonium compounds (QACs), widely used disinfectants in the food industry, was tested against Listeria by Roy et al. (1993) who used QUATAL (a product containing 10.5% N-alkyldimethylbenzylammonium HCL and 5.5% glutaraldehyde as active ingredients, Ecochimie Ltée, Quebec, Canada). Disinfection of stainless steel and polypropylene surfaces contaminated with L. monocytogenes biofilms was tested with bacteriophages H387, H387-A, and 2671 and their combination with QUATAL. The efficiency of these agents in combination was higher compared to the use of either one individually. Another combinatorial work tested bacteriophages and chlorine, a chemical compound extensively applied in the disinfection of industrial work surfaces, to remove of Pseudomonas biofilms. The Pseudomonas bacteriophage-chlorine mixture worked synergistically in the control and removal of P. aeruginosa biofilms (Zhang and Hu 2013). The effectiveness of combining bacteriophages with chemical antimicrobials has also been verified on the surface of several foods. Chibeu et al. (2013) tested the efficacy of anti-Listeria bacteriophage preparation Listex™ P100 in reducing L. monocytogenes on the surface of RTE roast beef and cooked turkey, combined with chemical compounds potassium lactate and sodium diacetate. This

810

C. Milho et al.

study showed that the use of Listex™ P100 on L. monocytogenes adhered cells improved the efficiency of the used chemical compounds. The reduction of different Salmonella strains on chicken meat and chicken skin by combined application of the bacteriophage preparation SalmoFresh™ with lauric arginate or cetylpyridinium chloride was evaluated (Sukumaran et al. 2015). This treatment led to significant bacterial load reductions when compared to the application of each product alone. In this way, the examples mentioned above demonstrate that, although bacteriophages have been proven to be efficient in the elimination of biofilms in different industrial settings, their combined use with different chemical antimicrobials can potentiate both products’ efficiency, since bacteriophages very likely disrupt the biofilm’s structure, leaving it more susceptible to the action of chemical compounds.

Challenges that Bacteriophages Face in Industrial Environments Bacteriophage application in industrial environments is challenging since many conditions can lead to their inactivation. For instance, bacteriophages encounter sanitizers and disinfectants that can impair their viability. Several studies have assessed Lactococcus lactis bacteriophage viability in commercial sanitizers (foodgrade chemicals included oxidizing agents, halogenated agents, alcohols, quaternary ammonium compounds, anionic acids, iodine-based acids, and an amphoteric chemical) and disinfectants. L. lactis bacteriophages are a problem in the dairy industry (see chapter ▶ “Industrial Processes Involving Bacteriophages”) because L. lactis strains are particularly vulnerable to members of the so-called 936 group of phages, and therefore their removal is desirable which can only be accomplished using strict sanitization and disinfection with appropriate biocidal solutions. Several bacteriophages possess resistance to specific biocides or biocide types, and the ones showing resistance also tend to possess a broad tolerance to multiple classes of antimicrobial compounds. Many chemical compounds, such as benzalkonium chloride (BAC), a quaternary ammonium compound (QAC), completely eliminated bacteriophages at concentrations of just 0.1% w/v after 30 min (Campagna et al. 2014; Hayes et al. 2017). Although not as efficient as BAC, hydrogen peroxide also is able to cause complete bacteriophage inactivation, but for this it requires high concentrations (20% (v/v)). A QAC-based disinfectant designed for surface cleaning in the food and dairy industry was tested in L. lactis bacteriophages, and the product was highly effective in destroying bacteriophages (Hayes et al. 2017). The several types of commercial sanitizers and disinfectants containing various chemical agents differ in pH according to their active agent. For instance, acid anionic and carboxylic acid products have pH values as low as 2, and the pH range of QACs and phenolic compounds varies in the range of 3–10.5 and 3–9.5, respectively (Richter and Cords 2001). The acidic pH conditions of many of these products are the main cause of bacteriophage inactivation. For instance, many bacteriophages do not survive pH lower than 4.5, such as T4 and PM2, among others (Ly-Chatain 2014). The EPA has performed an extensive study on the role of UV radiation on the inactivation of viruses and bacteriophages with different size and genomic

Biofilm Applications of Bacteriophages

811

composition (Shin et al. 2005). Some viruses rapidly became inactivated by low levels of UV. Nevertheless, although bacteriophage MS2 also suffered inactivation to some extent (2-log inactivation), the UV dose required for such reduction was high (30 mJ/cm2). Large DNA viruses, such as adenovirus 2, also resist inactivation and for a 2 log10 reduction they require an UV dose of 60 mJ/cm2. This behavior cannot however be transposed to bacteriophages since large DNA bacteriophages were rapidly inactivated (more than 5 log10) with a small dose of 10 mJ/cm2. Inactivation by UV is therefore a challenge for bacteriophage application in industrial environments where UV is used as a disinfecting agent, and as the EPA’s report demonstrates, bacteriophage inactivation by UV is not strictly virus size nor type dependent. High-pressure processing (HPP) is a nonthermal method applied to foods that inactivates harmful pathogens and vegetative spoilage microorganisms using intensive pressure (400–600 MPa). In the dairy industry, HPP is usually used on milk, though it is known to have great influence in the physicochemical and technological properties of milk (e.g., casein micelles are disintegrated, whey protein hydrophobicity is modified, color is affected, etc.) (Chawla et al. 2011). To warrant minimal effects on taste, texture, appearance, or nutritional value, the process is usually run at