Download Book

Methods in Molecular Biology 1701

Michael Hust Theam Soon Lim Editors Phage Display Methods and Protocols M ETHODS IN M OLECULAR B IOLOGY

Series Editor John M. Walker School of Life and Medical Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK

For further volumes: http://www.springer.com/series/7651 Phage Display

Methods and Protocols

Edited by Michael Hust

Technische Universit€at Braunschweig, Braunschweig, Germany Theam Soon Lim

Institute for Research in Molecular Medicine, Universiti Sains Malaysia, Minden, Penang, Malaysia Editors Michael Hust Theam Soon Lim Technische Universit€at Braunschweig Institute for Research in Molecular Medicine Braunschweig, Germany Universiti Sains Malaysia Minden, Penang, Malaysia

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-4939-7446-7 ISBN 978-1-4939-7447-4 (eBook) DOI 10.1007/978-1-4939-7447-4

Library of Congress Control Number: 2017956713

© Springer Science+Business Media LLC 2018 This work is subject to copyright. 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, express 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.

Printed on acid-free paper

This Humana Press imprint is published by Springer Nature The registered company is Springer Science+Business Media, LLC The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A. Preface

Antibodies have emerged from humble beginnings over a century ago when Emil von Behring and Shibasaburo Kitasato first witnessed the unique ability of antibodies in sera (first called “Anti-toxine”) from immunized rabbits to neutralize toxins. Since then, antibodies have emerged to become the most important biologicals. In the mid-1970s, Ko¨hler and Milstein revolutionized the application of antibodies not only in basic research and diagnostics but also in the biomedical field with their work on hybridoma technology. This was a turning point for many medical approaches applied at that time. The inherent ability to produce monospecific antibodies was a game changer in the pharmaceutical industry, but these antibodies were still mainly murine antibodies which can cause side effects. The evolution of DNA technology coupled with a deeper understanding of molecular biology and immunology allowed for the successive growth of antibody technology. The groundbreaking discovery by Smith on M13 phage display allowed for the presentation of a specific peptide on the surface of bacteriophages. This work was then further enhanced with the presentation of antibody fragments on the surface of bacteriophage which was independently developed around 1990 at the Deutsches Krebsforschungszentrum (DKFZ) in Heidelberg, Germany, at the MRC Laboratory of Molecular Biology in Cambridge, United Kingdom, and at the Scripps Research Institute in La Jolla, USA. The advancement in technologies allowed the selection of fully human antibodies from antibody phage display libraries. The plethora of different types of phage display libraries and applications of phage display highlights the robustness and durability of the method for antibody generation and became a key technology for the generation of therapeutic antibodies. Antibody phage display was also the technology to develop the first human antibody, Adalimumab, and best sold biological. This book provides examples of the generation of different forms of antibody libraries including libraries from different hosts. Many research groups share their expertise and experience in this book. A comprehensive list of different antibody libraries as well as novel approaches for antibody discovery is covered in this book. The chapters in this book can be divided into four sections: the first focuses on the construction of antibody libraries, followed by selection strategies for antibodies, complementary approaches for antibody selection, and finally epitope mapping and biomarker identification. This book provides a comprehensive list of antibody phage display technologies and applications. On a personal note, the ability to carry out intercontinental collaborative efforts was the essence of this book. This book showcases a collection of works from researchers from various countries across the globe. The experience of producing this edition when we were physically located in two different continents, in Europe and Asia, has helped to strengthen our resolve in our quest to contribute further in science. The work on this book has also helped strengthen the relationship and exchange between our laboratories not just on a research level but also more importantly on a personal level. Many new friendships and ideas have been developed over the course of this book that holds well for cross-border collaborative initiatives. Although a comprehensive list of topics has been covered in this book, there are still many more chapters that can still be written considering so much activity in all the antibody laboratories in the world. As the antibody business is a multibillion dollar industry, rapid v vi Preface technological developments as well as the emergence of new laboratories are expected. The new kid on the block in this book is the antibody technology program at the Institute for Research in Molecular Medicine (INFORMM), which is the brainchild of the Malaysian Ministry of Higher Education under the Higher Institution Centre of Excellence (HICoE) program together with Universiti Sains Malaysia (USM). It is our aim that this book can provide technical assistance to new start-ups that are venturing into the field of antibody phage display. We also hope this book will help spur interest and ideas in the field while expanding our growing family of enthusiastic antibody researchers. We would like to thank the authors whose contributions to this volume have allowed it to be a comprehensive guide to the processes involved in antibody phage display. We would also like to thank Prof. John M. Walker for his guidance and assistance throughout the editorial process. Our scientific career would not be possible without our great mentors Erhard Rhiel, Thomas Reinard, and Stefan Dubel€ and Zolta`n Konthur and Jo¨rn Glo¨kler. On a personal note, we would like to thank our families Dagmar, Noah Joris, and Lenja Marie and Poi Hong, Hayley, and Hayden for their patience while preparing this book and all our other projects.

Braunschweig, Germany Michael Hust Penang, Malaysia Theam Soon Lim Contents

Preface ...... v Contributors...... xi

PART ICONSTRUCTION OF ANTIBODY PHAGE DISPLAY LIBRARIES

1 Construction of Human Immune and Naive scFv Libraries ...... 3 Jonas Kugler,€ Florian Tomszak, Andre´ Frenzel, and Michael Hust 2 Construction of Naive and Immune Human Fab Phage-Display Library ...... 25 Noorsharmimi Omar and Theam Soon Lim 3 Construction of Synthetic Antibody Phage-Display Libraries ...... 45 Johan Nilvebrant and Sachdev S. Sidhu 4 Modular Construction of Large Non-Immune Human Antibody Phage-Display Libraries from Variable Heavy and Light Chain Gene Cassettes ...... 61 Nam-Kyung Lee, Scott Bidlingmaier, Yang Su, and Bin Liu 5 Construction of Macaque Immune-Libraries ...... 83 Arnaud Avril, Sebastian Miethe, Michael Hust, and Thibaut Pelat 6 Construction of Bovine Immunoglobulin Libraries in the Single-Chain Fragment Variable (scFv) Format ...... 113 Ulrike S. Diesterbeck 7 Construction of Rabbit Immune Antibody Libraries ...... 133 Thi Thu Ha Nguyen, Jong Seo Lee, and Hyunbo Shim 8 Generation of Semi-Synthetic Shark IgNAR Single-Domain Antibody Libraries ...... 147 Julius Grzeschik, Doreen Ko¨nning, Steffen C. Hinz, Simon Krah, Christian Schro¨ter, Martin Empting, Harald Kolmar, and Stefan Zielonka 9 Construction of High-Quality Camel Immune Antibody Libraries ...... 169 Ema Romaõ, Vianney Poignavent, Cećile Vincke, Christophe Ritzenthaler, Serge Muyldermans, and Baptiste Monsion 10 Construction of Chicken Antibody Libraries ...... 189 Jeanni Fehrsen, Susan Wemmer, and Wouter van Wyngaardt ® 11 Construction and Selection of Affilin Phage Display Libraries ...... 205 Florian Settele, Madlen Zwarg, Sebastian Fiedler, Daniel Koscheinz, and Eva Bosse-Doenecke 12 Construction of a Synthetic Antibody Gene Library for the Selection of Intrabodies and Antibodies ...... 239 De´borah Caucheteur, Gautier Robin, Vincent Parez, and Pierre Martineau

vii viii Contents

13 Targeting Intracellular Antigens with pMHC-Binding Antibodies: A Phage Display Approach ...... 255 Zhihao Wu, Brian H. Santich, Hong Liu, Cheng Liu, and Nai-Kong V. Cheung

PART II SELECTION STRATEGIES FOR ANTIBODIES

14 Parallelized Antibody Selection in Microtiter Plates ...... 273 Giulio Russo, Doris Meier, Saskia Helmsing, Esther Wenzel, Fabian Oberle, Andre´ Frenzel, and Michael Hust 15 Mass Spectrometry Immuno Assay (MSIA™) Streptavidin ® Disposable Automation Research Tips (D.A.R.T’s ) Antibody Phage Display Biopanning ...... 285 Chai Fung Chin, Yee Siew Choong, and Theam Soon Lim 16 Magnetic Nanoparticle-Based Semi-Automated Panning for High-Throughput Antibody Selection...... 301 Angela Chiew Wen Ch’ng, Nurul Hamizah Binti Hamidon, Zoltań Konthur, and Theam Soon Lim 17 Phage Display and Selections on Cells ...... 321 Wieland Fahr and Andre´ Frenzel 18 Combine Phage Antibody Display Library Selection on Patient Tissue Specimens with Laser Capture Microdissection to Identify Novel Human Antibodies Targeting Clinically Relevant Tumor Antigens ...... 331 Yang Su, Scott Bidlingmaier, Nam-Kyung Lee, and Bin Liu 19 Antibody Isolation From a Human Synthetic Combinatorial and Other Libraries of Single-Chain Antibodies ...... 349 Almog Bitton, Limor Nahary, and Itai Benhar 20 Screening Phage-Display Antibody Libraries Using Protein Arrays ...... 365 Ricardo Jara-Acevedo, Paula Dıéz, Marıá Gonza´lez-Gonza´lez, Rosa MarıáDe´gano, Nieves Ibarrola, Rafael Gongora, Alberto Orfao, and Manuel Fuentes 21 Antibody Selection on FFPE Tissue Slides ...... 381 Andre ten Haaf, Stefan Gattenlo¨hner, and Mehmet Kemal Tur 22 Antibody Affinity and Stability Maturation by Error-Prone PCR...... 393 Tobias Unkauf, Michael Hust, and Andre´ Frenzel

PART III COMPLEMENTARY APPROACHES FOR ANTIBODY PHAGE DISPLAY SELECTIONS

23 Upgrading Afﬁnity Screening Experiments by Analysis of Next-Generation Sequencing Data ...... 411 Christian Grohmann and Michael Blank

24 Next-Generation DNA Sequencing of VH/VL Repertoires: A Primer and Guide to Applications in Single-Domain Antibody Discovery ...... 425 Kevin A. Henry Contents ix

25 High-Throughput IgG Reformatting and Expression ...... 447 Chao-Guang Chen, Georgina Sansome, Michael J. Wilson, and Con Panousis 26 Monitoring Phage Biopanning by Next-Generation Sequencing ...... 463 Anna Vaisman-Mentesh and Yariv Wine

PART IV PHAGE DISPLAY FOR EPITOPE MAPPING AND IDENTIFICATION OF BIOMARKERS

27 ORFeome Phage Display ...... 477 Jonas Zantow, Gustavo Marc¸al Schmidt Garcia Moreira, Stefan Dubel,€ and Michael Hust 28 Epitope Mapping by Phage Display ...... 497 Gustavo Marc¸al Schmidt Garcia Moreira, Viola Fuhner,€ and Michael Hust 29 Metasecretome Phage Display ...... 519 Milica Ciric, Filomena Ng, Jasna Rakonjac, and Dragana Gagic 30 Phagekines: Screening Binding Properties and Biological Activity of Functional Cytokines Displayed on Phages ...... 535 Gertrudis Rojas and Tania Carmenate

Index ...... 561 Contributors

ARNAUD AVRIL De´partement des Maladies Infectieuses, unite´ biothe´rapies anti-infectieuses et immunite´, Institut de Recherche Biome´dicale des Armeés, Bre´tigny-sur-Orge, France ITAI BENHAR Department of Molecular Microbiology and Biotechnology, The George S. Wise Faculty of Life Sciences, Tel-Aviv University, Tel-Aviv, Israel SCOTT BIDLINGMAIER Department of Anesthesia, UCSF Helen Diller Family Comprehensive Cancer Center, University of California at San Francisco, San Francisco, CA, USA ALMOG BITTON Department of Molecular Microbiology and Biotechnology, The George S. Wise Faculty of Life Sciences, Tel-Aviv University, Tel-Aviv, Israel MICHAEL BLANK AptaIT GmbH, Planegg-Martinsried, Germany EVA BOSSE-DOENECKE Navigo Proteins GmbH, Halle (Saale), Germany TANIA CARMENATE Center of Molecular Immunology, Atabey, Playa, La Habana, Cuba DE´ BORAH CAUCHETEUR IRCM, Institut de Recherche en Cance´rologie de Montpellier, Montpellier, France; INSERM, U1194, Montpellier, France; Universite´ de Montpellier, Montpellier, France; Institut re´gional du Cancer de Montpellier, Montpellier, France ANGELA CHIEW WEN CH’NG Analytical Biochemistry Research Centre, Institute for Research in Molecular Medicine, Universiti Sains Malaysia, Penang, Malaysia CHAO-GUANG CHEN Research and Development, CSL Limited, Parkville, VIC, Australia NAI-KONG V. CHEUNG Department of Pediatrics, Memorial Sloan Kettering Cancer Center, New York, NY, USA; Gerstner Sloan Kettering Graduate School of Biomedical Sciences, Memorial Sloan Kettering Cancer Center, New York, NY, USA CHAI FUNG CHIN Institute for Research in Molecular Medicine, Universiti Sains Malaysia, Penang, Malaysia YEE SIEW CHOONG Institute for Research in Molecular Medicine, Universiti Sains Malaysia, Penang, Malaysia MILICA CIRIC Institute of Fundamental Sciences, Massey University, Palmerston North, New Zealand ROSA MARIÁ DE´ GANO Proteomics Unit, Cancer Research Center (CSIC/USAL/IBSAL), Salamanca, Spain PAULA DIÉZ Department of Medicine and General Cytometry Service-Nucleus, Cancer Research Center (CSIC/USAL/IBSAL), Salamanca, Spain; Proteomics Unit, Cancer Research Center (CSIC/USAL/IBSAL), Salamanca, Spain STEFAN Du€BEL Institut fur€ Biochemie, Biotechnologie und Bioinformatik, Abteilung Biotechnologie, Technische Universitat€ Braunschweig, Braunschweig, Germany ULRIKE S. DIESTERBECK Laboratory of Viral Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD, USA MARTIN EMPTING Department Drug Design and Optimization, Helmholtz-Institute for Pharmaceutical Research Saarland (HIPS), Saarland University, Saarbrucken,€ Germany VIOLA Fu€HNER Abteilung Biotechnologie, Institut fur€ Biochemie, Biotechnologie und Bioinformatik, Technische Universitat€ Braunschweig, Braunschweig, Germany WIELAND FAHR Abteilung Biotechnologie, Institut fur€ Biochemie, Biotechnologie und Bioinformatik, Technische Universitat€ Braunschweig, Braunschweig, Germany

xi xii Contributors

JEANNI FEHRSEN Immunology Division, New Generation Vaccines Programme, ARC-Onderstepoort Veterinary Research, Onderstepoort, Pretoria, South Africa SEBASTIAN FIEDLER Navigo Proteins GmbH, Halle (Saale), Germany ANDRE´ FRENZEL Abteilung Biotechnologie, Institut fur€ Biochemie, Biotechnologie und Bioinformatik, Technische Universitat€ Braunschweig, Braunschweig, Germany; YUMAB GmbH, Braunschweig, Germany MANUEL FUENTES Department of Medicine and General Cytometry Service-Nucleus, Cancer Research Center (CSIC/USAL/IBSAL), Salamanca, Spain; Proteomics Unit, Cancer Research Center (CSIC/USAL/IBSAL), Salamanca, Spain DRAGANA GAGIC Institute of Fundamental Sciences, Massey University, Palmerston North, New Zealand STEFAN GATTENLO¨ HNER Institute of Pathology, University Hospital Giessen, Justus Liebig University Giessen, Giessen, Germany RAFAEL GO´ NGORA Department of Medicine and General Cytometry Service-Nucleus, Cancer Research Center (CSIC/USAL/IBSAL), Salamanca, Spain MARIÁ GONZA´ LEZ-GONZA´ LEZ Department of Medicine and General Cytometry Service- Nucleus, Cancer Research Center (CSIC/USAL/IBSAL), Salamanca, Spain; Proteomics Unit, Cancer Research Center (CSIC/USAL/IBSAL), Salamanca, Spain CHRISTIAN GROHMANN AptaIT GmbH, Planegg-Martinsried, Germany JULIUS GRZESCHIK Institute for Organic Chemistry and Biochemistry, Technische Universitat€ Darmstadt, Darmstadt, Germany NURUL HAMIZAH BINTI HAMIDON Analytical Biochemistry Research Centre, Institute for Research in Molecular Medicine, Universiti Sains Malaysia, Penang, Malaysia SASKIA HELMSING Abteilung Biotechnologie, Institut fur€ Biochemie, Biotechnologie und Bioinformatik, Technische Universitat€ Braunschweig, Braunschweig, Germany KEVIN A. HENRY Human Health Therapeutics Portfolio, National Research Council Canada, Ottawa, ON, Canada STEFFEN C. HINZ Institute for Organic Chemistry and Biochemistry, Technische Universitat€ Darmstadt, Darmstadt, Germany MICHAEL HUST Abteilung Biotechnologie, Institut fur€ Biochemie, Biotechnologie und Bioinformatik, Technische Universitat€ Braunschweig, Braunschweig, Germany NIEVES IBARROLA Proteomics Unit, Cancer Research Center (CSIC/USAL/IBSAL), Salamanca, Spain RICARDO JARA-ACEVEDO ImmunoStep SL. Edificio Centro de Investigacion del Cańcer. Avda. Coimbra s/n, Salamanca, Spain DOREEN KO¨ NNING Institute for Organic Chemistry and Biochemistry, Technische Universitat€ Darmstadt, Darmstadt, Germany JONAS Ku€GLER YUMAB GmbH, Braunschweig, Germany HARALD KOLMAR Institute for Organic Chemistry and Biochemistry, Technische Universitat€ Darmstadt, Darmstadt, Germany ZOLTA´ N KONTHUR Max Planck Institute of Colloids and Interfaces, Potsdam, Germany DANIEL KOSCHEINZ Navigo Proteins GmbH, Halle (Saale), Germany SIMON KRAH Institute for Organic Chemistry and Biochemistry, Technische Universitat€ Darmstadt, Darmstadt, Germany; Protein Engineering and Antibody Technologies, Merck-Serono, Merck KGaA, Darmstadt, Germany JONG SEO LEE AbClon Inc., Seoul, Republic of Korea NAM-KYUNG LEE Department of Anesthesia, UCSF Helen Diller Family Comprehensive Cancer Center, University of California at San Francisco, San Francisco, CA, USA Contributors xiii

THEAM SOON LIM Institute for Research in Molecular Medicine, Universiti Sains Malaysia, Minden, Penang, Malaysia; Analytical Biochemistry Research Centre, Universiti Sains Malaysia, Minden, Penang, Malaysia BIN LIU Department of Anesthesia, UCSF Helen Diller Family Comprehensive Cancer Center, University of California at San Francisco, San Francisco, CA, USA CHENG LIU Eureka Therapeutics, Emeryville, CA, USA HONG LIU Eureka Therapeutics, Emeryville, CA, USA PIERRE MARTINEAU IRCM, Institut de Recherche en Cance´rologie de Montpellier, Montpellier, France; INSERM, U1194, Montpellier, France; Universite´ de Montpellier, Montpellier, France; Institut re´gional du Cancer de Montpellier, Montpellier, France DORIS MEIER Abteilung Biotechnologie, Institut fur€ Biochemie, Biotechnologie und Bioinformatik, Technische Universitat€ Braunschweig, Braunschweig, Germany SEBASTIAN MIETHE YUMAB GmbH, Braunschweig, Germany BAPTISTE MONSION Institut de Biologie Molećulaire des Plantes du CNRS, Universite´de Strasbourg, Strasbourg Cedex, France GUSTAVO MARÇAL SCHMIDT GARCIA MOREIRA Abteilung Biotechnologie, Institut fur€ Biochemie, Biotechnologie und Bioinformatik, Technische Universitat€ Braunschweig, Braunschweig, Germany SERGE MUYLDERMANS Cellular and Molecular Immunology, Vrije Universiteit Brussel, Brussel, Belgium LIMOR NAHARY Department of Molecular Microbiology and Biotechnology, The George S. Wise Faculty of Life Sciences, Tel-Aviv University, Tel-Aviv, Israel FILOMENA NG Institute of Fundamental Sciences, Massey University, Palmerston North, New Zealand; Grasslands Research Centre, AgResearch Ltd., Palmerston North, New Zealand THI THU HA NGUYEN Department of Life Science, Ewha Womans University, Seoul, Republic of Korea JOHAN NILVEBRANT Division of Protein Technology, School of Biotechnology, Royal Institute of Technology, Stockholm, Sweden; Donnelly Centre for Cellular and Biomolecular Research, Banting and Best Department of Medical Research, University of Toronto, Toronto, ON, Canada FABIAN OBERLE Abteilung Biotechnologie, Institut fur€ Biochemie, Biotechnologie und Bioinformatik, Technische Universitat€ Braunschweig, Braunschweig, Germany NOORSHARMIMI OMAR Institute for Research in Molecular Medicine, Universiti Sains Malaysia, Penang, Malaysia ALBERTO ORFAO Department of Medicine and General Cytometry Service-Nucleus, Cancer Research Center (CSIC/USAL/IBSAL), Salamanca, Spain CON PANOUSIS Research and Development, CSL Limited, Parkville, VIC, Australia VINCENT PAREZ IRCM, Institut de Recherche en Cance´rologie de Montpellier, Montpellier, France; INSERM, U1194, Montpellier, France; Universite´ de Montpellier, Montpellier, France; Institut re´gional du Cancer de Montpellier, Montpellier, France THIBAUT PELAT BIOTEM, Apprieu, France VIANNEY POIGNAVENT Institut de Biologie Molećulaire des Plantes du CNRS, Universite´ de Strasbourg, Strasbourg Cedex, France JASNA RAKONJAC Institute of Fundamental Sciences, Massey University, Palmerston North, New Zealand CHRISTOPHE RITZENTHALER Institut de Biologie Molećulaire des Plantes du CNRS, Universite´ de Strasbourg, Strasbourg Cedex, France xiv Contributors

GAUTIER ROBIN IRCM, Institut de Recherche en Cance´rologie de Montpellier, Montpellier, France; INSERM, U1194, Montpellier, France; Universite´ de Montpellier, Montpellier, France; Institut re´gional du Cancer de Montpellier, Montpellier, France GERTRUDIS ROJAS Center of Molecular Immunology, Atabey, Playa, La Habana, Cuba EMA ROMA˜O Cellular and Molecular Immunology, Vrije Universiteit Brussel, Brussel, Belgium GIULIO RUSSO Abteilung Biotechnologie, Institut fur€ Biochemie, Biotechnologie und Bioinformatik, Technische Universitat€ Braunschweig, Braunschweig, Germany GEORGINA SANSOME Research and Development, CSL Limited, Parkville, VIC, Australia BRIAN H. SANTICH Gerstner Sloan Kettering Graduate School of Biomedical Sciences, Memorial Sloan Kettering Cancer Center, New York, NY, USA CHRISTIAN SCHRO¨ TER Institute for Organic Chemistry and Biochemistry, Technische Universitat€ Darmstadt, Darmstadt, Germany; Protein Engineering and Antibody Technologies, Merck-Serono, Merck KGaA, Darmstadt, Germany FLORIAN SETTELE Navigo Proteins GmbH, Halle (Saale), Germany HYUNBO SHIM Department of Life Science, Ewha Womans University, Seoul, Republic of Korea; Department of Bioinspired Science, Ewha Womans Univesity, Seoul, Republic of Korea SACHDEV S. SIDHU Donnelly Centre for Cellular and Biomolecular Research, Banting and Best Department of Medical Research, University of Toronto, Toronto, ON, Canada YANG SU Department of Anesthesia, UCSF Helen Diller Family Comprehensive Cancer Center, University of California at San Francisco, San Francisco, CA, USA ANDRE TEN HAAF Bio-Rad AbD Serotec GmbH, Puchheim, Germany FLORIAN TOMSZAK YUMAB GmbH, Braunschweig, Germany MEHMET KEMAL TUR Institute of Pathology, University Hospital Giessen, Justus Liebig University Giessen, Giessen, Germany TOBIAS UNKAUF Institut fur€ Biochemie, Biotechnologie und Bioinformatik, Abteilung Biotechnologie, Technische Universitat€ Braunschweig, Braunschweig, Germany ANNA VAISMAN-MENTESH Department of Molecular Microbiology and Biotechnology, Faculty of Life Science, Tel Aviv University, Tel Aviv, Israel CE´ CILE VINCKE Cellular and Molecular Immunology, Vrije Universiteit Brussel, Brussel, Belgium SUSAN WEMMER Immunology Division, New Generation Vaccines Programme, ARC-Onderstepoort Veterinary Research, Onderstepoort, Pretoria, South Africa ESTHER WENZEL Abteilung Biotechnologie, Institut fur€ Biochemie, Biotechnologie und Bioinformatik, Technische Universitat€ Braunschweig, Braunschweig, Germany MICHAEL J. WILSON Research and Development, CSL Limited, Parkville, VIC, Australia YARIV WINE Department of Molecular Microbiology and Biotechnology, Faculty of Life Science, Tel Aviv University, Tel Aviv, Israel ZHIHAO WU Department of Pediatrics, Memorial Sloan Kettering Cancer Center, New York, NY, USA WOUTER VAN WYNGAARDT Immunology Division, New Generation Vaccines Programme, ARC-Onderstepoort Veterinary Research, Onderstepoort, Pretoria, South Africa JONAS ZANTOW Institut fur€ Biochemie, Biotechnologie und Bioinformatik, Abteilung Biotechnologie, Technische Universitat€ Braunschweig, Braunschweig, Germany STEFAN ZIELONKA Institute for Organic Chemistry and Biochemistry, Technische Universitat€ Darmstadt, Darmstadt, Germany; Protein Engineering and Antibody Technologies, Merck-Serono, Merck KGaA, Darmstadt, Germany MADLEN ZWARG Navigo Proteins GmbH, Halle (Saale), Germany Part I

Construction of Antibody Phage Display Libraries Chapter 1

Construction of Human Immune and Naive scFv Libraries

Jonas Kugler,€ Florian Tomszak, Andre´ Frenzel, and Michael Hust

Abstract

Antibody phage display is the most commonly used in vitro selection technology for the generation of human recombinant antibodies and has yielded thousands of useful antibodies for research, diagnostics, and therapy. The prerequisite for successful generation of antibodies using phage display is the construction of high-quality antibody gene libraries. Here, we give the detailed methods for the construction of human immune and naive scFv gene libraries.

Key words Phage display, Immune phage-display library, Antibody gene library, Naive phage-display library, scFv, Single-chain fragment variable (scFv), Human antibody phage-display library, V-gene ampliﬁcation, Antibody gene ampliﬁcation, PMBC isolation

1 Introduction

Antibody phage display is a key technology to generate human antibodies for research, diagnostic, and therapy. As of October 2016, more than 50 antibody and antibody conjugates are approved by the US Food and Drug Administration (FDA) or European Medicines Agency (EMEA, since 2009 EMA) with additional candidates pending approval [1]. The business volume of biologics was ~75 billion US$ in 2013 and the predicted sales for 2020 have a volume of ~125 billion US$ [2]. The primary application of therapeutic antibodies is cancer and autoimmune diseases [3]. In 1986, the first antibody of murine ® origin muronomab-CD3 (Orthoclone OKT3 ) was approved for therapy [4]. The next generation of therapeutic antibodies was chimeric, e.g., the anti-tumor necrosis factor (TNF) antibody ® Infliximab (Remicade ) for the treatment of rheumatic arthritis and Crohn’s disease [5] or the anti-epidermal growth factor receptor ® (EGFR) antibody cetuximab (Erbitux ) for cancer therapy [6]. Later, the antibodies were humanized, e.g., trastuzumab (Hercep- ® tin )[7]. In 2002, the first fully human antibody was commercially ® available. Adalimumab (Humira ) was isolated using antibody

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_1, © Springer Science+Business Media LLC 2018 3 4 Jonas Kugler€ et al.

phage display by guided selection [8]. Human antibodies are assumed to be less immunogenic, but also these fully human molecules can lead to adverse events [9, 10]. Currently, six approved antibodies were generated by phage display and a panel is in clinical development [11]. The phage-display technology is based on the groundbreaking work of Georg P. Smith on filamentous phage [12]. Here, genotype and phenotype of oligo-peptides were linked by fusing the corresponding gene fragments to the minor coat protein III gene of the filamentous bacteriophage M13. The resulting peptide::pIII fusion protein is expressed on the surface of phage allowing the affinity purification of the peptide and its corresponding gene. In the same way, antibody fragments fused to pIII can be presented on the surface of M13 phage particles. This technology was invented in parallel in Cambridge, Heidelberg, and La Jolla in 1990/91 [13–18]. Due to limitations of the E. coli folding machinery, only antibody fragments such as scFv (single-chain Fragment variable), Fab (Fragment antigen binding), VHH (camel heavy chain variable domain), or dAbs (human heavy chain variable domain) are used routinely for antibody phage display [19–21]. Two different genetic systems have been developed for the expression of the antibody::pIII fusion proteins for phage display. First, the antibody genes can be directly inserted into the phage genome fused to the wild-type pIII gene [18]. However, most of the successful systems uncouple antibody expression from phage propagation by providing the genes encoding the antibody::pIII fusion proteins on a separate plasmid (called “phagemid”) These phagemids contain a phage morphogenetic signal for packaging of the vector into the assembled phage particles [14]. Despite other in vitro methods such as ribosomal display [22, 23], puromycin display [24], yeast surface display [25], or mammalian cell display [26], antibody phage display has become the most widely used selection method for human antibodies.

2 Materials

2.1 Isolation 1. Phosphate-buffered saline (PBS) pH 7.4: 8.0 g NaCl, 0.2 g of Lymphocytes KCl, 1.44 g Na2HPO4*2H2O, 0.24 g KH2PO4 in 1 L. 2. Lymphoprep. 3. mRNA isolation Kit or TRIzol for total RNA.

2.2 cDNA Synthesis 1. Superscript IV Reverse Transcriptase þ5Â RT buffer þ0.1 m DTT. 2. RNAseOut Recombinant Ribonuclease Inhibitor.

3. Random hexamer oligonucleotide primer (dN6). 4. dNTP mix: 10 mM each. Construction scFv Libraries 5

2.3 First and Second 1. GoTaq2 Polymerase þ5Â buffer. Antibody Gene PCR 2. dNTP mix: 10 mM each. 3. Oligonucleotide primer: see Table 1. 4. Agarose. 5. TAE-buffer 50Â: 2 M Tris–HCl, 1 M acetic acid, 0.05 M EDTA pH 8. 6. NucleoSpin Gel and PCR Clean-up.

2.4 First Cloning 1. MluI-HF restriction enzyme. Step—VL 2. NotI-HF restriction enzyme. 3. Cut Smart Buffer. 4. Calf intestine phosphatase (CIP). 5. T4 ligase. 6. Amicon Ultra Centrifugal Filters (30 K) (Millipore, Schwal- bach, Germany). 7. E. coli XL1-Blue MRF0(Agilent), genotype: Δ(mcrA)183 Δ(mcrCB-hsdSMR-mrr)173 endA1 supE44 thi-1 recA1 gyrA96 relA1 lac [F´ proAB lacIqZΔM15 Tn10 (Tetr)]. 8. Electroporator MicroPulser, 0.1 cm cuvettes. 9. 2 M Glucose, sterile ﬁltered.

10. 2 M Magnesium solution (autoclaved): 1 M MgCl2,1M MgSO4. 11. SOC medium pH 7.0: 2% (w/v) tryptone, 0.5% (w/v) yeast extract, 0.05% (w/v) NaCl, 20 mM Mg solution, 20 mM glucose, sterilize magnesium and glucose separately, add solutions after autoclavation. 12. 2xTY-medium pH 7,0: 1,6% (w/v) tryptone, 1% (w/v) yeast extract, 0,5% (w/v) NaCl). 13. Ampicilline: 100 mg/mL stock. 14. 2xTY-GAT: 2xTY þ 100 mM glucose þ100 μg/mL ampicilline þ20 μg/mL tetracycline. 15. Tetracycline: 10 mg/mL stock. 16. 9 cm Petri dishes. 17. 25 cm square Petri dishes (“pizza plates”). 18. 2xTY-GAT agar plates: 2xTY-GAT, 1.5% (w/v) agar-agar). 19. Nucleobond Extra Midi Kit.

2.5 Second Cloning 1. NcoI-HF restriction enzyme. Step—VH 2. HindIII-HF restriction enzyme. 3. Cut Smart Buffer. 6 Jonas Kugler€ et al.

Table 1 Primers used for first and second PCR of antibody genes for antibody gene library construction using phagemids like pHAL14, pHAL30, or pHAL35

Primer 5’ to 3’ sequence

First antibody gene PCR VH MHVH1_f cag gtb cag ctg gtg cag tct gg MHVH1/7_f car rts cag ctg gtr car tct gg MHVH2_f cag rtc acc ttg aag gag tct gg JokVH3_f1 sag gtg cag ctg gtg gag tct gg JokVH3_f2 gar gtg cag ctg ktg gag tct gg MHVH4_f1 cag gtg car ctg cag gag tcg gg JokVH4_f2 cag gtg cag cta car cag tgg gg JokVH4_f3 cag ctg cag ctg cag gag tcs gg MHVH5_f gar gtg cag ctg gtg cag tct gg MHVH6_f cag gta cag ctg cag cag tca gg MHIgMCH1_r aag ggt tgg ggc gga tgc act MHIgGCH1_r gac cga tgg gcc ctt ggt gga MHIgECH1_r tgg gct ctg tgt gga gg First antibody gene PCR kappa MHVK1_f1 gac atc cag atg acc cag tct cc MHVK1_f2 gmc atc crg wtg acc cag tct cc MHVK2_f gat rtt gtg atg acy cag wct cc MHVK3_f gaa atw gtg wtg acr cag tct cc MHVK4_f gac atc gtg atg acc cag tct cc MHVK5_f gaa acg aca ctc acg cag tct cc MHVK6_f gaw rtt gtg mtg acw cag tct cc MHkappaCL_r aca ctc tcc cct gtt gaa gct ctt First antibody gene PCR lambda MHVL1_f1 cag tct gtg ctg act cag cca cc MHVL1_f2 cag tct gtg ytg acg cag ccg cc MHVL2_f cag tct gcc ctg act cag cct MHVL3_f1 tcc tat gwg ctg acw cag cca cc MHVL3_f2 tct tct gag ctg act cag gac cc MHVL4_f1 ctg cct gtg ctg act cag ccc MHVL4_f2 cag cyt gtg ctg act caa tcr yc (continued) Construction scFv Libraries 7

Table 1 (continued)

Primer 5’ to 3’ sequence

MHVL5_f cag sct gtg ctg act cag cc MHVL6_f aat ttt atg ctg act cag ccc ca MHVL7/8_f cag rct gtg gtg acy cag gag cc MHVL9/10_f cag scw gkg ctg act cag cca cc MHlambdaCL_r tga aca ttc tgt agg ggc cac tg MHlambdaCL_r2 tga aca ttc cgt agg ggc aac tg Second antibody gene PCR VH MHVH1-NcoI_f gtcctcgca cc atg gcc cag gtb cag ctg gtg cag tct gg MHVH1/7-NcoI_f gtcctcgca cc atg gcc car rts cag ctg gtr car tct gg MHVH2-NcoI_f gtcctcgca cc atg gcc cag rtc acc ttg aag gag tct gg JokVH3-NcoI_f1 gtcctcgca cc atg gcc sag gtg cag ctg gtg gag tct gg JokVH3-NcoI_f2 gtcctcgca cc atg gcc gar gtg cag ctg ktg gag tct gg MHVH4-NcoI_f1 gtcctcgca cc atg gcc cag gtg car ctg cag gag tcg gg JokVH4-NcoI_f2 gtcctcgca cc atg gcc cag gtg cag cta car cag tgg gg JokVH4-NcoI_f3 gtcctcgca cc atg gcc cag ctg cag ctg cag gag tcs gg MHVH5-NcoI_f gtcctcgca cc atg gcc gar gtg cag ctg gtg cag tct gg MHVH6-NcoI_f gtcctcgca cc atg gcc cag gta cag ctg cag cag tca gg MHIgMCH1scFv-HindIII_r gtcctcgca aag ctt tgg ggc gga tgc act MHIgGCH1scFv-HindIII_r gtcctcgca aag ctt gac cga tgg gcc ctt ggt gga MHIgECH1scFv-HindIII_r gtcctcgca aag ctt tgg gct ctg tgt gga gg Second antibody gene PCR kappa MHVK1-MluI_f1 accgcctcc a cgc gta gac atc cag atg acc cag tct cc MHVK1-MluI_f2 accgcctcc a cgc gta gmc atc crg wtg acc cag tct cc MHVK2-MluI_f accgcctcc a cgc gta gat rtt gtg atg acy cag wct cc MHVK3-MluI_f accgcctcc a cgc gta gaa atw gtg wtg acr cag tct cc MHVK4-MluI_f accgcctcc a cgc gta gac atc gtg atg acc cag tct cc MHVK5-MluI_f accgcctcc a cgc gta gaa acg aca ctc acg cag tct cc MHVK6-MluI_f accgcctcc a cgc gta gaw rtt gtg mtg acw cag tct cc MHkappaCLscFv-NotI_r accgcctcc gc ggc cgc gaa gac aga tgg tgc agc cac agt (continued) 8 Jonas Kugler€ et al.

Table 1 (continued)

Primer 5’ to 3’ sequence

Second antibody gene PCR lambda MHVL1-MluI_f1 accgcctcc a cgc gta cag tct gtg ctg act cag cca cc MHVL1-MluI_f2 accgcctcc a cgc gta cag tct gtg ytg acg cag ccg cc MHVL2-MluI_f accgcctcc a cgc gta cag tct gcc ctg act cag cct MHVL3-MluI_f1 accgcctcc a cgc gta tcc tat gwg ctg acw cag cca cc MHVL3-MluI_f2 accgcctcc a cgc gta tct tct gag ctg act cag gac cc MHVL4-MluI_f1 accgcctcc a cgc gta ctg cct gtg ctg act cag ccc MHVL4-MluI_f2 accgcctcc a cgc gta cag cyt gtg ctg act caa tcr yc MHVL5-MluI_f accgcctcc a cgc gta cag sct gtg ctg act cag cc MHVL6-MluI_f accgcctcc a cgc gta aat ttt atg ctg act cag ccc ca MHVL7/8-MluI_f accgcctcc a cgc gta cag rct gtg gtg acy cag gag cc MHVL9/10-MluI_f accgcctcc a cgc gta cag scw gkg ctg act cag cca cc MHLambdaCLscFv-NotI_r accgcctcc gc ggc cgc aga gga sgg ygg gaa cag agt gac Primer for colony PCR and sequencing MHLacZ-Pro_f ggctcgtatgttgtgtgg MHgIII_r c taa agt ttt gtc gtc ttt cc Restriction sites are underlined

4. E. coli ER2738 (Lucigen), genotype: [F’proA þ Bþ lacIq Δ(lacZ)M15 zzf::Tn10 (tetr)] fhuA2 glnVΔ(lac-proAB) thi-1Δ(hsdS-mcrB)5. 5. Glycerol of 99.5%.

2.6 Colony PCR 1. Oligonucleotide primer: see Table 1

2.7 Library 1. 2xTY media pH 7.0: 1.6% (w/v) tryptone, 1% (w/v) yeast Packaging and scFv extract, 0.5% (w/v) NaCl. Phage Production 2. 2xTY-GA: 2xTY, 100 mM glucose, 100 μg/mL ampicillin. 3. M13 K07 Helperphage for monovalent display (Thermo Fisher Scientiﬁc, Waltham, USA). 4. Hyperphage for oligovalent display (Progen, Heidelberg, Germany). 5. 2xTY-AK: 2xTY þ 100 μg/mL ampicillin þ50 μg/mL kanamycin. 6. Sorval Centrifuge RC5B Plus, rotor GS3 and SS34. Construction scFv Libraries 9

7. Polyethylenglycol (PEG) solution: 20% (w/v) PEG 6000, 2.5 M NaCl. 8. Phage dilution buffer: 10 mM Tris–HCl pH 7.5, 20 mM NaCl, 2 mM EDTA. 9. Mouse α-pIII monoclonal antibody PSKAN3 (Mobitec, Go¨t- tingen, Germany). 10. Goat α-mouse IgG alkaline phosphatase (AP) conjugate.

2.8 Phage Titration 1. 2xTY-GA agar plates: 2xTY-GA þ 1.5% (w/v) agar-agar.

3 Methods

Depending on the scientific or medical applications, various types of antibody gene libraries can be constructed and used. Immune libraries are constructed from antibody V-genes isolated from IgG secreting plasma cells of immunized donors [15, 27, 28]. Immune libraries are typically generated and used in medical research to obtain antibodies against one particular target antigen, e.g., a cell surface antigen of a pathogen or a tumor marker. Naive, semi- synthetic and synthetic libraries have been subsumed as “single- pot” libraries, as they are designed to isolate antibody fragments binding to nearly every possible antigen. Naive libraries are constructed from rearranged V genes from B cells (IgM) of non-immunized donors. An overview of antibody gene libraries and vectors is given in several reviews [21, 29, 30]. A panel of methods has been employed to clone the genetic diversity of human antibody repertoires. After the isolation of mRNA from B-lymphocytes and the preparation of cDNA, construction of immune libraries is usually done by a two-step cloning or assembly PCR (see below). Very large “single pot” naive antibody gene libraries are generally constructed by two or three separate cloning steps. In the “two step cloning strategy,” the amplified repertoire of light chain genes is cloned into the phage display vector first. In the second step the heavy chain gene repertoire—as the heavy chain contributes more to antibody diversity, due to its highly variable CDRH3 [31]—is cloned into the phagemids containing the light chain gene repertoire ([32–35]. In the “three step cloning strategy,” separate heavy and light chain libraries are engineered. The VH gene repertoire has then to be excised and cloned into the phage-display vector containing the repertoire of VL genes [36]. Assembly PCR is another common method used for the cloning of naive [37, 38]orimmune[15] scFv phage-display libraries. The VH and VL genes including the additional linker sequence are amplified separately and fused by assembly PCR, before the scFv encoding gene fragments are cloned into the vector. Since the CDRH3 is a major source of sequence variety in antibodies [31], 10 Jonas Kugler€ et al.

the assembly PCR can be combined with a randomization of the CDR3 regions, leading to semi-synthetic libraries. Here, oligonucleotide primers encoding various CDR3 and J gene segments were used for the amplification of the V gene segments of human germ- lines [39]. Hoogenboom and Winter [40] as well as Nissim and colleagues [41] used degenerated CDRH3 oligonucleotide primers to generate a semi-synthetic heavy chain repertoire derived from human V gene germline segments. Afterward, this VH repertoire was combined with an anti-BSA light chain. For some libraries a single framework of a well-known/robust antibody was used as a scaffold for the integration of randomly created CDRH3 and CDRL3 [42, 43]. Jirholt and colleagues [44]andSo¨derlind and colleagues [45] amplified all CDR regions derived from B cells before shuffling them into this antibody framework by assembly PCR. The first approach for full synthetic libraries was made by Knappik et al. [46]. They used 49 framework region genes and generated the randomized CDR3s of the heavy and light chain by trinucleotide synthesis. An entirely synthetic library was described by Rothe et al. [47] who utilized seven different VH and VL germline master frameworks combined with all six synthetically created CDR cassettes. Tiller and colleagues [48] generated a fully synthetic human library based on 36 fixed VH and VL pairs that were selected for biophysical properties favorable for antibody developability. Con- struction of large naive and synthetic libraries requires a significant effort to tunnel the genetic diversity through the bottleneck of E. coli transformation, e.g., 287 transformations were performed for the generation of the human antibody libraries (HAL) 9/10 with a combined size of 1.5 Â 1010 independent clones [49]. The following protocols describe the generation of human naive or immune scFv antibody gene libraries by a two-step cloning strategy already approved for naive [32, 49] and immune libraries [27].

3.1 Isolation 1. Mix 20 mL fresh blood or EDTA/citric acid treated blood of Lymphocytes (~2 Â 107 cells) of each donor with 20 mL PBS (see Note 1). (Peripher Blood 2. Fill 10 mL Lymphoprep in a 50 mL polypropylen tube. Care- Mononuclear Cells fully cover Lymphoprep with 40 mL of the diluted blood using (PBMC)) a plastic pipette. 3. Centrifuge the blood with 800 Â g for 20 min at RT (without brake!). 4. The lymphocytes form a distinct layer between the Lympho- prep and the medium, whereas the erythrocytes and granulo- cytes will be pelleted. Carefully aspirate the lymphocytes using a plastic pipette and transfer to a new 50 mL polypropylen tube. 5. Fill up with 50 mL PBS and pellet the lymphocytes with 250 Â g for 10 min at RT. Discard the supernatant (be careful, the lymphocyte pellet is not solid). Construction scFv Libraries 11

6. Repeat this washing step to remove most of the thrombocytes. 7. Resuspend the lymphocytes pellet in the supplied extraction buffer of the mRNA isolation kit according to the manufacturer’s instructions or use 0.5 mL TRIzol for total RNA isolation (see Note 2). After resuspension use the mRNA extraction buffer or TRIzol. The RNA pellet can be stored at À80 C.

3.2 cDNA Synthesis 1. Set up mixture for the ﬁrst-strand cDNA synthesis:

Solution or component Volume Final concentration

mRNA or total RNA Up to 50–250 ng (mRNA) or 11 μL 2–20 μg (total RNA) Random hexamer oligonucleotide 1 μL 2.5 μM primer (dN6) (50 μM) dNTP-mix (10 mM each) 1 μL 500 μM DEPC-treated or nuclease-free water To 13 μL

2. Denature the RNA for 5 min at 65 C. Afterward chill on ice for at least 1 min. 3. Add following components:

Volume Final Solution or component (μL) concentration

SSIV buffer (5Â)41Â 0.1 M DTT 1 10 mM Superscript IV reverse transcriptase 1 200 U (200 U/μL) RNAseOut (200 U/μL) 1 –

4. Incubate the 20 μL mixture for 10 min at 23 C for primer annealing. Afterward incubate for 10 min at 50 C for ﬁrst- strand synthesis. 5. Denature the RNA/DNA hybrids and the enzyme for 10 min at 80 C. Store at À20 C.

3.3 First Antibody 1. The cDNA derived from 50–250 ng mRNA or 2–20 μg total Gene PCR RNA will be used as a template to amplify VH and the light chain. Set up the PCR reactions as follows (30Â mastermix for 28 PCR reactions): 12 Jonas Kugler€ et al.

Solution or component Volume (μL) Final concentration

dH2O 1130 – GoTaq Buffer (5Â) 300 1Â dNTPs (10 mM each) 30 200 μM each cDNA 25 Complete ﬁrst-strand synthesis reaction GoTaq2 5 U/μL 7.5 1.25 U

2. Divide the master mix in 500 μL for VH, 350 μL for kappa and 550 μL for lambda. 3. Add to each of the three reactions the corresponding reverse primers (see also Table 1) as follows (use the IgM primer for naive antibody gene libraries or the IgG primer for immune antibody gene libraries. Also IgE libraries are possible with the IgE primer set):

Final Antibody Volume concentration gene Primer (μL) (μM)

VH MHIgMCH1_r or 20 0.4 MHIgGCH1_r or MHIgECH1_r (10 μM) Kappa MHkappaCL_r (10 μM) 14 0.4 Lambda MHlambdaCL_r1/ _r2 mix 22 0.4 (9:1) (10 μM)

4. Divide the mixture into 10 (VH), 7 (Kappa), and 11 (Lambda) PCR reactions each with 48 μL and add 2 μL (10 μM, 0.4 μM ﬁnal concentration) of the subfamily speciﬁc forward primer (see also Table 1):

(1) MHVH1_f, (2) MHVH1/7_f, (3) MHVH2_f, (4) JokVH3_f1 (5) JokVH3_f2, (6) MHVH4_f1, (7) JokVH4_f2, (8) JokVH4_f3, (9) VH: MHVH5_f, (10) MHVH6_f

Vkappa: (11) MHVK1_f1, (12) MHVK1_f2, (13) MHVK2_f, (14) MHVK3_f, (15) MHVK4_f, (16) MHVK5_f, (17) MHVK6_f Vlambda: (18) MHVL1_f1, (19) MHVL1_f2, (20) MHVL2_f, (21) MHVL3_f1, (22) MHVL3_f2, (23) MHVL4_f1, (24) MHVL4_f2, (25) MHVL5_f, (26) MHVL6_f, (27) MHVL7/8_f, (28) MHVL9/10_f Construction scFv Libraries 13

5. Carry out the PCR using the following program:

95 C 1 min

95 C 1 min 30Â 55 C 1 min 72 C 2 min 72 C 10 min

6. Separate PCR products by 1.5% TAE agarose gel electrophoresis, cut out the amplified antibody genes (VH: ~380 bp, kappa/ lambda: ~650 bp) (see Note 3), and purify the PCR products using a gel extraction kit according to the manufacturer’s instructions. Pool all VH, kappa, and lambda subfamilies separately. Determine the DNA concentration. Store the three purified first PCR pools at À20 C.

3.4 Second Antibody 1. In the second PCR the restriction sites for library cloning will Gene PCR be added. Set up the PCR reactions as follows (30Â mastermix for 28 PCR reactions) (see Note 4):

Solution or component Volume (μL) Final concentration

dH2O 2200 – GotTaq Buffer (5Â) 600 1Â dNTPs (10 mM each) 60 200 μM each GoTaq 5 U/μL 15 2.5 U

2. Divide the master mix in 1000 μL for VH, 700 μL for kappa and 1100 μL for lambda. 3. Add to each of the three reactions the corresponding reverse primers (see also Table 1) as follows:

Final Antibody Volume concentration gene Primer (μL) (μM)

VH MHIgMCH1scFv-HindIII_r 20 0.2 or MHIgGCH1scFv- HindIII_r (10 μM) Kappa MHKappaCLscFv-NotI_r 14 0.2 (10 μM) Lambda MHLambdaCLscFv-NotI_r 22 0.2 (10 μM) 14 Jonas Kugler€ et al.

4. Add the corresponding PCR products of the ﬁrst PCR as follows:

VH 1000 ng

Kappa 700 ng Lambda 1100 ng

5. Divide the solutions into 10 (VH), 7 (Kappa) and 11 (Lambda) PCR reactions, each with 98 μL and add 2 μL (10 μM, 0.2 μM ﬁnal concentration) the subfamily-speciﬁc forward primer (see also Table 1):

VH: (1) MHVH1-NcoI_f, (2) MHVH2-NcoI_f, (3) MHVH1/7-NcoI_f,

(4) JokVH3-NcoI_f1, (5) JokVH3-NcoI_f2, (6) MHVH4- NcoI_f1, (7) JokVH4-NcoI_f2, (8) JokVH4-NcoI_f3, (9) MHVH5- NcoI_f, (10) MHVH6-NcoI_f Vkappa: (11) MHVK1-MluI_f1, (12) MHVK1-MluI_f2, (13) MHVK2-MluI_f, (14) MHVK3-MluI_f, (15) MHVK4-MluI_f, (16) MHVK5- MluI_f, (17) MHVK6-MluI_f Vlambda: (18) MHVL1-MluI_f1, (19) MHVL1-MluI_f2, (20) MHVL2- MluI_f, (21) MHVL3-MluI_f1, (22) MHVL3-MluI_f2, (23) MHVL4- MluI_f1, (24) MHVL4-MluI_f2, (25) MHVL5-MluI_f, (26) MHVL6- MluI_f, (27) MHVL7/8-MluI_f, (28) MHVL9/10-MluI_f

6. Carry out the PCR using the following program:

95 C 1 min

95 C 1 min 20Â 57 C 1 min 72 C 1.5 min 72 C 10 min Construction scFv Libraries 15

7. Separate the PCR products by 1.5% TAE agarose gel electrophoresis, cut out the ampliﬁed antibody genes (VH: ~400 bp, kappa/lambda: ~400 bp), and purify the PCR products using a gel extraction kit according to the manufacturer’s instructions. Pool all VH, kappa, and lambda subfamilies separately. Deter- mine the DNA concentration. Store the three puriﬁed second PCR pools at À20 C.

3.5 First Cloning 1. Prepare a plasmid preparation of pHAL30 vector for library Step—VL cloning (see Note 5). 2. Digest the vector and the VL PCR products. Always perform additional single-enzyme digestions of the vector in parallel to check whether the digestion is complete (see also Note 6):

Solution or component Volume Final concentration

dH2O 87–x μL– pHAL30 or VL ÂμL5μgor2μg NEB cut smart buffer (10Â)10μL1Â NEB MluI-HF (20 U/μL) 1.5 μL30U NEB NotI-HF (20 U/μL) 1.5 μL30U

3. Incubate at 37 C for 2 h. Control the digest of the vector by using a 5 μL aliquot on 1% TAE agarose gelelectrophoresis. If the vector is not fully digested, extend the incubation time. 4. Inactivate the enzymes at 65 C for 10 min. 5. Add 0.5 μL CIP (1 U/μL) to the vector digest and incubate at 37 C for 30 min. Repeat this step once. 6. Purify the vector and the PCR product using a PCR puriﬁca- tion Kit according to the manufacturer’s instructions and elute with 50 μL elution buffer or water. The short stuffer fragment containing multiple stop codons between MluI and NotI in pHAL30 will be removed. Determine the DNA concentration. 7. Ligate the vector pHAL30 (4255 bp) and VL (~380 bp) as follows (see Note 4):

Solution or component Volume Final concentration

dH2O 89-x–y μL– pHAL30 ÂμL 1000 ng VL (kappa or lambda) y μL 270 ng T4 ligase buffer (10Â)10μL1Â T4 ligase (3 U/μL) 1 μL3U 16 Jonas Kugler€ et al.

8. Incubate at 16 C overnight. 9. Inactivate the ligation at 70 C for 10 min. 10. Purify the ligation using an Amicon Ultra column. Add the ligation to the column and add water to 500 μL. Centrifuge at 10 min at 14,000 Â g. Discard the ﬂow-through and repeat the washing with 470 μL (about 30 μL will remain in the column) step three times. 11. For elution invert the column and elute the remaining DNA solution in a new cap for 3 min at 1000 Â g. 12. Add water to 35 μL. 13. Thaw 25 μL electrocompetent E. coli XL1-Blue MRF’ on ice and mix with the ligation reaction. 14. Transfer the 60 μL mix to a prechilled 0.1 cm cuvette. Dry the electrode of the cuvette with a tissue paper. 15. Perform a 1.7 kV pulse using an electroporator (see Note 7). Immediately, add 1 mL 37 C pre-warmed SOC medium, transfer the suspension to a 2 mL cap and shake for 1 h at 600 rpm and 37 C. 16. To determine the amount of transformants, use 10 μL À (¼10 2 dilution) of the transformation and perform a dilu- À À tion series down to 10 6 dilution. Plate out a 10 6 dilution on 2xTY-GAT agar plates and incubate overnight at 37 C. 17. Plate out the remaining 990 μL on 2xYT-GAT agar “pizza plate” and incubate overnight at 37 C. 18. Calculate the amount of transformants, which should be 1 Â 106–2 Â 108 cfu. Control colonies for full-size insert by colony PCR (see Subheading 3.7). 19. Float off the colonies on the “pizza plate” with 40 mL 2xTY medium using a drigalsky spatula. Use 5 mL bacteria solution for midi plasmid preparation according to the manufacturer’s instructions. Determine the DNA concentration.

3.6 Second Cloning 1. Digest the pHAL30-VL repertoire and the VH PCR products. Step—VH Always perform additional single enzyme digestions of the vector in parallel (see also Note 6):

Solution or component Volume Final concentration

dH2O 82-x μL– pHAL30-VL or VH ÂμL5μgor2μg NEB cut smart buffer (10Â)10μL1Â NEB NcoI-HF (20 U/μL) 1.5 μL30U NEB HindIII-HF (20 U/μL) 1.5 μL30U Construction scFv Libraries 17

2. Incubate at 37 Cfor2h(see Note 8). Control the digest of the vector by using a 5 μL aliquot on 1% agarose gelelectrophoresis. 3. Inactivate the digestion at 80 C for 20 min. 4. Add 0.5 μL CIP (1 U/μL) to the vector digest and incubate at 37 C for 30 min. Repeat this step once. 5. Purify the vector and the PCR product using a PCR puriﬁca- tion Kit according to the manufacturer’s instructions and elute with 50 μL elution buffer or water. The short stuffer fragment between NcoI and HindII in pHAL30 will be removed. Deter- mine the DNA concentration. See also Note 9. 6. Ligate the vector pHAL30-VL (~4610 bp) and VH (~380 bp) as follows (see Note 4):

Solution or component Volume Final concentration dH2O 89-x–y μL– pHAL30 ÂμL 1000 ng VH y μL 250 ng T4 ligase buffer (10Â)10μL1Â T4 ligase (3 U/μL) 1 μL3U

7. Incubate at 16 C overnight. 8. Inactivate the ligation at 65 C for 10 min. 9. Purify the ligation using an Amicon Ultra column. Add the ligation to the column and add water to 500 μL. Centrifuge at 10 min at 14000 Â g. Discard the ﬂow-through and repeat the washing with 470 μL (about 30 μL will remain in the column) step three times. 10. For elution invert the column and elute the remaining DNA solution in a new cap for 3 min at 1000 Â g. 11. Add water to 35 μL. 12. Thaw 25 μL electrocompetent E. coli ER2738 on ice and mix with the ligation reaction (see Note 10). 13. Transfer the 60 μL mix to a prechilled 0.1 cm cuvette. Dry the electrode of the cuvette with a tissue paper. 14. Perform a 1.7 kV pulse using an electroporator (see Note 7). Immediately, add 1 mL 37 C pre-warmed SOC medium (Lucigen), transfer to a 2 mL cap, and incubate for 1 h at 600 rpm. 15. To determine the amount of transformants, use 10 μL À (¼10 2 dilution) of the transformation and perform a dilu- À À tion series down to 10 6 dilution. Plate out a 10 6 dilution on 2xTY-GAT agar plates and incubate overnight at 37 C. 18 Jonas Kugler€ et al.

16. Plate out the remaining 990 μL on 2xTY-GAT agar “pizza plate” and incubate overnight at 37 C. 17. Calculate the amount of transformants (1 Â 107–2 Â 108 should be reached to be included into the ﬁnal library). Control colonies for full-size insert by colony PCR (see Sub- heading 3.7). 18. Float off the colonies on the “pizza plate” with 25 mL 2xTY medium using a drigalsky spatula (~O.D. 20–25 ¼ ~2 Â 1010 cells/mL). Use 800 μL bacteria solution (~1 Â 1010 bacteria) and 200 μL glycerol for glycerol stocks. Make 5–20 glycerol stocks per sublibrary and store at À80 C. 19. When all transformations are done, thaw one aliquot of each sublibrary on ice, mix all sublibraries and make new aliquots for storage at À80 C(see also Note 11).

3.7 Colony PCR 1. Choose 10–20 single colonies per transformation. Set up the 10 μL PCR reaction per colony as follows (see Table 1 for primer sequences):

Solution or component Volume Final concentration

dH2O 7.5 μL GoTaq buffer (5Â)2μL1Â dNTPs (10 mM each) 0,2 μL 200 μM each MHLacZPro_f 10 μM 0,1 μL 0,1 μM MHgIII_r 10 μM 0,1 μL 0,1 μM GoTag2 (5 U/ μL) 0,1 μL 0.5 U Template Picked colonies from dilution plate

2. Control the PCR by 1.5% TAE agarose gelelectrophoresis. 3. The PCR products should be ~1100 bp when including VH and VL, ~750 bp when including only VL or VH and 375 bp if the vector contains no insert. Each used sublibrary should have more than 80% full-size inserts to be included into the ﬁnal library.

3.8 Library 1. To package the library, inoculate 400 mL 2xTY-GA in a 1 L Packaging and scFv Erlenmeyer ﬂask with 1 mL antibody gene library stock. Grow Phage Production at 250 rpm at 37 C up to an O.D.600 nm ~ 0.5. 2. Infect 25 mL bacteria culture (~1.25 Â 1010 cells) with 2.5 Â 1011 colony forming units (cfu) of the helper phage M13 K07 or Hyperphage according to a multiplicity of Construction scFv Libraries 19

infection (moi) ¼ 1:20 (see Note 12). Incubate for 30 min without shaking and the following 30 min with 250 rpm at 37 C. 3. To remove the glucose that represses the lac promoter of pHAL30 and therefore the scFv::pIII fusion protein expression, harvest the cells by centrifugation for 10 min at 3200 Â g in 50 mL polypropylene tubes. 4. Resuspend the pellet in 400 mL 2xTY-AK in a 1 L Erlenmeyer ﬂask. Produce scFv-phage overnight at 250 rpm and 30 C. 5. Pellet the bacteria by centrifugation for 10 min at 10,000 Â g in two GS3 centrifuge tubes. If the supernatant is not clear, centrifuge again to remove remaining bacteria. 6. Precipitate the phage from the supernatant by adding 1/5 volume PEG solution in two GS3 tubes. Incubate for 1 h at 4 C with gentle shaking, followed by centrifugation for 1 h at 10,000 Â g. 7. Discard the supernatant, resolve each pellet in 10 mL phage dilution buffer in SS34 centrifuge tubes, and add 1/5 volume PEG solution. 8. Incubate on ice for 20 min and pellet the phage by centrifugation for 30 min at 10,000 Â g. 9. Discard the supernatant and put the open tubes upside down on tissue paper. Let the viscous PEG solution move out completely. Resuspend the phage pellet in 1 mL phage dilution buffer. Titer the phage preparation (see Subheading 3.9). Store the packaged antibody phage library at 4 C. 10. The library packaging should be controlled by 10% SDS-PAGE, Western-Blot and anti-pIII immunostain (mouse anti-pIII 1:2000, goat anti-mouse IgG AP conjugate 1:10,000). Wild-type pIII has a calculated molecular mass of 42.5 kDa, but it runs at an apparent molecular mass of 65 kDa in SDS-PAGE. Accordingly, the scFv::pIII fusion protein runs at about 95 kDa.

3.9 Phage Titration 1. Inoculate 5 mL 2xTY-T in a 100 mL Erlenmeyer ﬂask with E. coli XL1-Blue MRF’ and grow overnight at 37 C and 250 rpm. 2. Inoculate 50 mL 2xTY-T with 500 μL overnight culture and grow at 250 rpm at 37 C up to O.D.600 ~ 0.5 (see Note 13). 3. Make serial dilutions of the phage suspension in PBS. The package library phage preparation should have a titer of about 1011–1013 phage/mL. 4. Infect 50 μL bacteria with 10 μL phage dilution and incubate for 30 min at 37 C. 20 Jonas Kugler€ et al.

5. You can perform titrations in two different ways: (a) Plate the 60 μL infected bacteria on 2xTY-GA agar plates (9 cm petri dishes). (b) Pipet 10 μL (in triplicate) on 2xTY-GA agar plates. Here, about 20 titering spots can be placed on one 9 cm petri dish. Dry drops on work bench. 6. Incubate the plates overnight at 37 C. 7. Count the colonies and calculate the cfu or cfu/mL titer according to the dilution.

4 Notes

1. Be careful with human blood samples since it is potentially infectious (HIV, hepatitis, etc.). 2. Both methods, mRNA or total RNA isolation, work well. 3. The VH amplifications of VH subfamilies sometimes result also in longer PCR products. Cut out only the ~380 bp fragment. The amplifications of kappa subfamilies should always give a clear ~650 bp fragment (complete light chain). When amplifying lambda subfamilies often other PCR products are generated, especially the amplification of the lambda 2 subfamily results often in slushy bands. If some subfamilies are bad amplified and no clear ~650 bp fragment is detectable, use only the ~650 bp fragments from the well-amplified subfamilies. Additional comment: since the first PCR amplifies the full LC, it can be used also to construct Fab or scFab [50] libraries from this material. 4. For a very large naive antibody gene library perform as many PCRs as sufficient to perform 20 light chains ligations/transformations and about 100 VH ligations. For an immune library four light chains ligations/transformations and eight VH ligations are usually sufficient. Prepare and digest also adequate amounts of pHAL30 and VL for the first cloning step and pHAL30-VL library and VH for the second cloning step. Keep kappa and lambda libraries in all steps (cloning, packaging) separately and mix only after phage production before panning. 5. The vector pHAL30 is a modified version of pHAL14. In pHAL30 the orientation of the tags is Myc-His instead of His-Myc resulting in a higher scFv production rate [49]. 6. Always perform single digests using only one enzyme in parallel, to control the success of the restriction reaction. Analyze the digestion by TAE agarose gelelectophoresis by comparing with the undigested plasmid. Use only material where both Construction scFv Libraries 21

single digests are successful and where no degradation is visible in the double digest. 7. The pulse time should be between 4–5 ms for optimal electroporation efficiency. 8. Often the HindIII digestion is incomplete after 2 h. Then, inactivate the enzymes by heating up to 65 C for 10 min, add additional 5 μL of HindIII and incubate overnight. You can use also higher concentrated HindIII. Alternative: perform the NcoI digest first for 2 h, inactive the digest, and afterward perform the HindIII digest. This problem only occurs when HindIII is used and not if HindIII-HF. 9. Keep aliquots of the light chain repertoire as plasmid, but also—more convenient—the MluI/NotI digested VL chains for future light chains shuffling for affinity and/or stability maturation. For affinity maturation, use the VH and clone it into an empty pHAL30 vector, subsequently clone the new light chain repertoire to combine it with the selected. This light chains shuffling library can be used for the panning under harsher conditions or competition to select improved antibodies [51]. 10. The E. coli ER2738 cells have a higher transformations efficiency compared to E. coli XL1-Blue MRF’. These ER2738 cells are used only for the second VH cloning step, because the quality of isolated plasmids from these cells is lower compared to XL1-Blue MRF’. The XL1-Blue MRF’ cells are used for the first cloning VL cloning step to get high-quality plasmids for the second digestion and VH cloning step. In the VL cloning step, the library size can be lower (1 Â 107–1 Â 108) because this repertoire will be combined with the VH repertoire in the second cloning step. 11. To minimize loss of diversity, avoid too many freeze and thaw steps, e.g., when constructing an immune library make eight transformations in parallel and directly package the immune library. When making a big immune library, combine only a glycerinstock of each sublibrary that corresponds to max. 2 Â 109 independent clones to ensure that the library diversity can be kept when packaging 1 mL of mixed library glycerinstock. When the library size is bigger than 2 Â 109 independent clones, do not package the library as complete library, package “blocks” of sublibraries. Combine the phage particles of each “block” before panning to get the final complete library. 12. The use of Hyperphage as helperphage instead of M13 K07 offers oligovalent phage display, facilitates the selection of specific binders in the first and most critical panning round by an avidity effect [52–55]. The Hyperphage should be only used 22 Jonas Kugler€ et al.

for library packaging. For the following panning rounds use M13 K07 to enhance the stringency of the panning process.

13. If the bacteria have reached O.D.600 ~ 0,5 before they are needed, you can store the culture immediately on ice to maintain the F pili on the E. coli cells for several hours. M13 K07 helperphage (kan+) or other scFv-phage (amp+) can be used as positive control to check the infectibility of the E. coli cells.

Acknowledgments

This review is an updated and revised version of Ref. 56.

References

1. Reichert JM (2016) Therapeutic monoclonal 10. Harding FA, Stickler MM, Razo J, DuBridge antibodies approved or in review in the RB (2010) The immunogenicity of humanized European Union or the United States. In: and fully human antibodies: residual immuno- The Antibody Society. Accessed 24 Jun 2016 genicity resides in the CDR regions. MAbs http://www.antibodysociety.org/news/ 2:256–265 approved-antibodies/ 11. Frenzel A, Schirrmann T, Hust M (2016) 2. Ecker DM, Jones SD, Levine HL (2015) The Phage display-derived human antibodies in therapeutic monoclonal antibody market. clinical development and therapy. MAbs MAbs 7:9–14. https://doi.org/10.4161/ 8:1177–1194. https://doi.org/10.1080/ 19420862.2015.989042 19420862.2016.1212149 3. Dubel€ S (2007) Recombinant therapeutic anti- 12. Smith GP (1985) Filamentous fusion phage: bodies. Appl Microbiol Biotechnol novel expression vectors that display cloned 74:723–729. https://doi.org/10.1007/ antigens on the virion surface. Science s00253-006-0810-y 228:1315–1317 4. Chatenoud L, Bluestone JA (2007) 13. Barbas CF, Kang AS, Lerner RA, Benkovic SJ CD3-specific antibodies: a portal to the treat- (1991) Assembly of combinatorial antibody ment of autoimmunity. Nat rev Immunol libraries on phage surfaces: the gene III site. 7:622–632. Doi: nri2134 Proc Natl Acad Sci U S A 88:7978–7982. Doi: 5. Harriman G, Harper LK, Schaible TF (1999) 1896445 Summary of clinical trials in rheumatoid arthri- 14. Breitling F, Dubel€ S, Seehaus T, Klewinghaus I, tis using infliximab, an anti-TNFalpha treat- Little M (1991) A surface expression vector for ment. Ann Rheum Dis 58(Suppl 1):I61–I64 antibody screening. Gene 104:147–153 6. Dalle S, Thieblemont C, Thomas L, Dumontet 15. Clackson T, Hoogenboom HR, Griffiths AD, C (2008) Monoclonal antibodies in clinical Winter G (1991) Making antibody fragments oncology. Anti-cancer agents in medicinal using phage display libraries. Nature chemistry 8:523–32. Doi: 18537534 352:624–628. Doi: 1907718 7. Jones SE (2008) Metastatic breast cancer: the 16. Hoogenboom HR, Griffiths AD, Johnson KS, treatment challenge. Clin Breast Cancer Chiswell DJ, Hudson P, Winter G (1991) 8:224–233. doi: R8180138H6520884 Multi-subunit proteins on the surface of fila- 8. Osbourn J, Groves M, Vaughan T (2005) mentous phage: methodologies for displaying From rodent reagents to human therapeutics antibody (fab) heavy and light chains. Nucleic using antibody guided selection. Methods Acids Res 19:4133–4137. Doi: 1908075 36:61–68. doi: S1046-2023(05)00016-2 17. Marks JD, Hoogenboom HR, Bonnert TP, 9. Getts DR, Getts MT, McCarthy DP, Chastain McCafferty J, Griffiths AD, Winter G (1991) EML, Miller SD (2010) Have we overesti- By-passing immunization. Human antibodies mated the benefit of human(ized) antibodies? from V-gene libraries displayed on phage. J MAbs 2:682–694 Mol Biol 222:581–597. Doi: 1748994 Construction scFv Libraries 23

18. McCafferty J, Griffiths AD, Winter G, Chiswell neutralizing potency against Japanese encepha- DJ (1990) Phage antibodies: filamentous litis virus. Microbiol Immunol 51:617–625 phage displaying antibody variable domains. 29. Hust M, Dubel€ S (2004) Mating antibody Nature 348:552–554 phage display with proteomics. Trends Bio- 19. Hoet RM, Cohen EH, Kent RB, Rookey K, technol 22:8–14 Schoonbroodt S, Hogan S, Rem L, Frans N, 30. Qi H, Lu H, Qiu H-J, Petrenko V, Liu A Daukandt M, Pieters H, van Hegelsom R, Neer (2012) Phagemid vectors for phage display: NC, Nastri HG, Rondon IJ, Leeds JA, Hufton properties, characteristics and construction. J SE, Huang L, Kashin I, Devlin M, Kuang G, Mol Biol 417:129–143. https://doi.org/10. Steukers M, Viswanathan M, Nixon AE, Sex- 1016/j.jmb.2012.01.038 ton DJ, Hoogenboom HR, Ladner RC (2005) 31. Shirai H, Kidera A, Nakamura H (1999) Generation of high-affinity human antibodies H3-rules: identification of CDR-H3 structures by combining donor-derived and synthetic in antibodies. FEBS Lett 455:188–197. Doi: complementarity-determining-region diversity. 10428499 Nat Biotechnol 23:344–348. Doi: nbt1067 32. Hust M, Meyer T, Voedisch B, Rulker€ T, 20. Holt LJ, Herring C, Jespers LS, Woolven BP, Thie H, El-Ghezal A, Kirsch MI, Schutte€ M, Tomlinson IM (2003) Domain antibodies: Helmsing S, Meier D, Schirrmann T, Dubel€ S proteins for therapy. Trends Biotechnol (2011) A human scFv antibody generation 21:484–490 pipeline for proteome research. J Biotechnol 21. Hust M, Dubel€ S (2005) Phage display vectors 152:159–170. https://doi.org/10.1016/j. for the in vitro generation of human antibody jbiotec.2010.09.945 fragments. Methods Mol Biol 295:71–96 33. Johansen LK, Albrechtsen B, Andersen HW, 22. Hanes J, Pluckthun€ A (1997) In vitro selection Engberg J (1995) pFab60: a new, efficient vec- and evolution of functional proteins by using tor for expression of antibody fab fragments ribosome display. Proc Natl Acad Sci U S A displayed on phage. Protein Eng 94:4937–4942. Doi: 9144168 8:1063–1067. Doi: 8771188 23. He M, Taussig MJ (1997) Antibody-ribosome- 34. Little M, Welschof M, Braunagel M, Hermes I, mRNA (ARM) complexes as efficient selection Christ C, Keller A, Rohrbach P, Kurschner€ T, particles for in vitro display and evolution of Schmidt S, Kleist C, Terness P (1999) Genera- antibody combining sites. Nucleic Acids Res tion of a large complex antibody library from 25:5132–5134. Doi: 9396828 multiple donors. J Immunol Methods 24. Roberts RW, Szostak JW (1997) RNA-peptide 231:3–9. Doi: 10648923 fusions for the in vitro selection of peptides and 35. Welschof M, Terness P, Kipriyanov SM, proteins. Proc Natl Acad Sci U S A Stanescu D, Breitling F, Do¨rsam H, Dubel€ S, 94:12297–12302. Doi: 9356443 Little M, Opelz G (1997) The antigen-binding 25. Boder ET, Wittrup KD (1997) Yeast surface domain of a human IgG-anti-F(ab’)2 autoanti- display for screening combinatorial polypeptide body. Proc Natl Acad Sci U S A libraries. Nat Biotechnol 15:553–557. Doi: 94:1902–1907. Doi: 9050877 9181578 36. de Haard HJ, van Neer N, Reurs A, Hufton SE, 26. King DJ, Bowers PM, Kehry MR, Horlick RA Roovers RC, Henderikx P, de Bruı¨ne AP, (2014) Mammalian cell display and somatic Arends JW, Hoogenboom HR (1999) A large hypermutation in vitro for human antibody non-immunized human fab fragment phage discovery. Curr Drug Discov Technol library that permits rapid isolation and kinetic 11:56–64 analysis of high affinity antibodies. J Biol Chem 27. Trott M, Weiβ S, Antoni S, Koch J, von 274:18218–18230. Doi: 10373423 Briesen H, Hust M, Dietrich U (2014) Func- 37. McCafferty J, Fitzgerald KJ, Earnshaw J, Chis- tional characterization of two scFv-fc antibo- well DJ, Link J, Smith R, Kenten J (1994) dies from an HIV controller selected on soluble Selection and rapid purification of murine anti- HIV-1 Env complexes: a neutralizing V3- and a body fragments that bind a transition-state ana- trimer-specific gp41 antibody. PLoS One 9: log by phage display. Appl Biochem Biotechnol e97478. https://doi.org/10.1371/journal. 47:157–171–173. Doi: 7944335 pone.0097478 38. Vaughan TJ, Williams AJ, Pritchard K, 28. Arakawa M, Yamashiro T, Uechi G, Tadano M, Osbourn JK, Pope AR, Earnshaw JC, Nishizono A (2007) Construction of human McCafferty J, Hodits RA, Wilton J, Johnson fab (gamma1/kappa) library and identification KS (1996) Human antibodies with of human monoclonal fab possessing sub-nanomolar affinities isolated from a large 24 Jonas Kugler€ et al.

non-immunized phage display library. Nat Bio- antibodies. J Mol Biol 376:1182–1200. technol 14:309–314. Doi: 9630891 https://doi.org/10.1016/j.jmb.2007.12.018 39. Akamatsu Y, Cole MS, Tso JY, Tsurushita N 48. Tiller T, Schuster I, Deppe D, Siegers K, (1993) Construction of a human Ig combina- Strohner R, Herrmann T, Berenguer M, torial library from genomic V segments and Poujol D, Stehle J, Stark Y, Heßling M, synthetic CDR3 fragments. J Immunol Daubert D, Felderer K, Kaden S, Ko¨lln J, 151:4651–4659. Doi: 8409426 Enzelberger M, Urlinger S (2013) A fully syn- 40. Hoogenboom HR, Winter G (1992) thetic human fab antibody library based on By-passing immunisation. Human antibodies fixed VH/VL framework pairings with favor- from synthetic repertoires of germline VH able biophysical properties. MAbs 5:445–470 gene segments rearranged in vitro. J Mol Biol 49. Kugler€ J, Wilke S, Meier D, Tomszak F, 227:381–388. Doi: 1404359 Frenzel A, Schirrmann T, Dubel€ S, 41. Nissim A, Hoogenboom HR, Tomlinson IM, Garritsen H, Hock B, Toleikis L, Schutte€ M, Flynn G, Midgley C, Lane D, Winter G (1994) Hust M (2015) Generation and analysis of the Antibody fragments from a “single pot” phage improved human HAL9/10 antibody phage display library as immunochemical reagents. display libraries. BMC Biotechnol 15:10. EMBO J 13:692–698. Doi: 7508862 https://doi.org/10.1186/s12896-015-0125- 42. Barbas CF, Bain JD, Hoekstra DM, Lerner RA 0 (1992) Semisynthetic combinatorial antibody 50. Hust M, Jostock T, Menzel C, Voedisch B, libraries: a chemical solution to the diversity Mohr A, Brenneis M, Kirsch MI, Meier D, problem. Proc Natl Acad Sci U S A Dubel€ S (2007) Single chain fab (scFab) frag- 89:4457–4461. Doi: 1584777 ment. BMC Biotechnol 7:14 43. Desiderio A, Franconi R, Lopez M, Villani ME, 51. Steinwand M, Droste P, Frenzel A, Hust M, Viti F, Chiaraluce R, Consalvi V, Neri D, Ben- Dubel€ S, Schirrmann T (2014) The influence venuto E (2001) A semi-synthetic repertoire of of antibody fragment format on phage display intrinsically stable antibody fragments derived based affinity maturation of IgG. MAbs from a single-framework scaffold. J Mol Biol 6:204–218. https://doi.org/10.4161/mabs. 310:603–615. Doi: 11439027 27227 44. Jirholt P, Ohlin M, Borrebaeck CA, So¨derlind 52. Rondot S, Koch J, Breitling F, Dubel€ S (2001) E (1998) Exploiting sequence space: shuffling A helper phage to improve single-chain anti- in vivo formed complementarity determining body presentation in phage display. Nat Bio- regions into a master framework. Gene technol 19:75–78 215:471–476. Doi: 9714846 53. Soltes G, Hust M, Ng KKY, Bansal A, Field J, 45. So¨derlind E, Strandberg L, Jirholt P, Stewart DIH, Dubel€ S, Cha S, Wiersma EJ Kobayashi N, Alexeiva V, Aberg AM, (2007) On the influence of vector design on Nilsson A, Jansson B, Ohlin M, Wingren C, antibody phage display. J Biotechnol Danielsson L, Carlsson R, Borrebaeck CA 127:626–637 (2000) Recombining germline-derived CDR 54. Pelat T, Hust M, Laffly E, Condemine F, sequences for creating diverse single- Bottex C, Vidal D, Lefranc M-P, Dubel€ S, framework antibody libraries. Nat Biotechnol Thullier P (2007) High-affinity, human 18:852–856. Doi: 10932154 antibody-like antibody fragment (single-chain 46. Knappik A, Ge L, Honegger A, Pack P, variable fragment) neutralizing the lethal factor Fischer M, Wellnhofer G, Hoess A, Wo¨lle J, (LF) of bacillus anthracis by inhibiting protec- Pluckthun€ A, Virnek€as B (2000) Fully synthetic tive antigen-LF complex formation. Antimi- human combinatorial antibody libraries crob Agents Chemother 51:2758–2764 (HuCAL) based on modular consensus frame- 55. Kirsch M, Hulseweh€ B, Nacke C, Rulker€ T, works and CDRs randomized with trinucleo- Schirrmann T, Marschall H-J, Hust M, Dubel€ tides. J Mol Biol 296:57–86. Doi: 10656818 S (2008) Development of human antibody 47. Rothe C, Urlinger S, Lo¨hning C, Prassler J, fragments using antibody phage display for Stark Y, J€ager U, Hubner B, Bardroff M, the detection and diagnosis of Venezuelan Pradel I, Boss M, Bittlingmaier R, Bataa T, equine encephalitis virus (VEEV). BMC Bio- Frisch C, Brocks B, Honegger A, Urban M technol 8:66. Doi: 1472-6750-8-66 (2008) The human combinatorial antibody 56. Frenzel A, Kugler€ J, Wilke S, Schirrmann T, library HuCAL GOLD combines diversifica- Hust M (2014) Construction of human anti- tion of all six CDRs according to the natural body gene libraries and selection of antibodies immune system with a novel display method by phage display. Methods Mol Biol for efficient selection of high-affinity 1060:215–243. https://doi.org/10.1007/ 978-1-62703-586-6_12 Chapter 2

Construction of Naive and Immune Human Fab Phage-Display Library

Noorsharmimi Omar and Theam Soon Lim

Abstract

This protocol describes the processes involved in the generation of human antibody libraries in Fab format. The antibody repertoire is derived from peripheral blood mononucleocytes focusing on different immunoglobulin isotypes. A two-step cloning process was used to generate a diverse human Fab library for subsequent selection by phage display. The method can be applied for the generation of both naive and immune antibody libraries. The naive repertoire allows for the library to be applied for the generation of human monoclonal antibodies against a broad range of target antigens making it a useful resource for antibody generation. However, the immune repertoire will be focused against target antigens from a particular disease. The protocol will focus on the generation of the library including the panning process.

Key words Naive antibody repertoires, Antibody libraries, Fab, Human, Phage display, Phagemid, Monoclonal antibodies

1 Introduction

The evolution of human monoclonal antibody technology took precedence with the introduction of phage-display technology by George Smith in 1985 [1]. Since the early work by Smith on the presentation of peptides on the surface of phage particles, the evolution of phage display gained further momentum. Phage display is a versatile and robust method that allows the presentation of peptides or proteins as a fusion protein on the surface of bacteriophages [2]. The application of phage display has transcended beyond peptides to also include enzymes [3, 4], scaffolds [5, 6], and more importantly antibodies [7, 8]. The introduction of antibody phage display has revolutionized the way monoclonal antibodies are being generated in the twenty- ﬁrst century. It has since played a pivotal role in the rise of antibody- based therapies in the pharmaceutical industry with the ﬁrst phage ® display derived therapeutic antibody, Adalimumab (Humira ) [9]. Since then, several other therapeutic antibodies have made it

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_2, © Springer Science+Business Media LLC 2018 25 26 Noorsharmimi Omar and Theam Soon Lim

to the market and there are constantly new antibodies entering clinical trials. Antibodies derived from phage display have also been applied successfully for diagnostic applications with antibodies targeting disease-specific biomarkers [8]. Another important application of antibody phage display is the generation of antibodies for large-scale proteome studies [10]. The method allows for rapid and continuous generation of antibodies against a wide array of target proteins [11, 12]. The ability to automate the selection process also adds value to the generation process in terms of cost and time efficiency [13]. There are several forms of antibody libraries that are classified mainly by the cDNA source encoding the antibody genes. Anti- body libraries are commonly divided into either naive, immune, or synthetic libraries [11]. Naive libraries are unique as the antibody gene repertoire used is generally from healthy individuals. In this way, the antibody repertoire would not have been exposed to any disease-specific antiges resulting in a naive unbiased repertoire. Naive libraries are generally generated through the isolation of IgM repertoire instead of other isotypes. The main concern when dealing with naive libraries is the true unbiased nature of the repertoire. This is because it is difficult to obtain truly unexposed repertoires, as all individuals would have encountered some form of infection in their lifetime. This also takes into account the formation of memory immune responses that is typically associated with IgM [14, 15]. Naive libraries although are somewhat unfocused in its target antigen population, it has been shown to be efficient in the generation of antibodies against disease-specific antigens [16, 17]. Immune antibody libraries are different from naive libraries as the former contains distinctive antibodies against particular antigens in response to an infection. Generally, the immune response would react towards an infection by generating IgG isotype antibodies against the target antigens [14, 15]. This is the main reason why IgG is the preferred isotype used for the immune antibody library generation. The immune library is normally produced to isolate a distinct antibody against a defined disease antigen [18, 19]. As IgG repertories usually undergo maturation processes, the antibodies derived from immune libraries will generally possess higher affinities if compared to those obtained from naive libraries. There are several different formats that can be applied for phage display. The most common formats used are the domain antibodies, single-chain fragment variable (scFv), and fragment antigen binding (Fab) [20]. Domain antibodies are the smallest in size with just the variable domains of either the heavy or light chain being presented [21]. However, both the scFv and Fab formats carry the major characteristic of a collective binding effect from the heavy and light chain domains [22]. These formats are smaller in size when compared with the full Y-shaped antibody structure. This size characteristic makes these formats easier and more efficiently Human Fab Library Generation 27

presented on phage surfaces due to the size tolerance of the phage packaging mechanism. The Fab is structurally formed by the heavy chain (HC) and light chain (LC) variable regions and constant regions juxtaposed to each other, whereby the HC and LC are connected by inter- chain disulphide bridges [23]. The size of Fab antibodies is double the size of scFv making it challenging for phage presentation. Nevertheless, the size of the Fab is still under the display tolerance of the phage particle to be presented at the pIII position. Although the generation of the Fab phage-display library is challenging in terms of cloning and phage packaging, Fab is still a preferred antibody format together with scFv for recombinant antibody production. There are several different approaches employed for the cloning of the genetic diversity of the antibody repertoires. After the isolation of the mRNA from B-lymphocytes, cDNA templates are prepared by RT-PCR. There are several different approaches used for antibody library cloning like PCR assembly [17], intermediate TOPO cloning [18], and isothermal rolling circle amplification [24]. However, the method commonly used is the two-step cloning. The two-step cloning strategy is typically done by the insertion of the light chain into the phagemid for an intermediate mini light chain library followed by the cloning of the heavy chain repertoire to yield the full antibody repertoire [25, 26]. The main challenge when developing such libraries is the diversity and size of the library at the final point. Therefore, great care is taken into consideration during the downstream cloning process to ensure the antibody repertoire is maintained as diverse as possible to reduce sequence redundancy and cloning bias. Here, we describe the application of the two-step cloning strategy for the generation of human Fab antibody phage-display libraries. The choice of the library being a naive or immune library would mainly depend on the isotype of the heavy chain repertoire that will ultimately be used for library cloning. The construction of a large diverse antibody library would require balancing the diversity with the limitations of E. coli transformation efficiencies. The protocol described here is optimized for Fab library cloning to generate a diverse library for phage display.

2 Materials

2.1 Isolation 1. Ficoll-Paque™ PLUS solution (GE Healthcare, Life Sciences). of B-Cells 2. 1Â Phosphate-Buffered Saline (PBS), pH 7.4: Add 8.0 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4.2H2O, 0.24 g KH2PO4 in 1 L dH2O, autoclave and store at room temperature. 3. QIAamp RNA Blood Mini Kit (Qiagen, Germany). 28 Noorsharmimi Omar and Theam Soon Lim

2.2 First-Strand 1. 300–500 ng RNA sample used as template. cDNA Synthesis 2. SuperscriptTM II Reverse Trancriptase (Invitrogen™, USA). 3. Oligo dT 12–18, 500 μg/mL (Invitrogen™, USA). 4. Random hexamers, 3 μg/mL (Invitrogen™, USA). 5. 10 mM dNTP mixture (Invitrogen™, USA).

® ® 2.3 Amplification 1. Vent DNA polymerase (NEB Inc.,USA) and Pfu DNA poly- of HC and LC Fab Gene merase (Thermo Fisher) together with theirs’ corresponding Repertoire buffer. 2. Forward and Reverse Forward and Reverse primers for first amplification of HC and LC (see Table 1) primers for second amplification of HC and LC with restriction endonuclease sites (see Table 2). 3. 10 mM dNTP mixture (Thermo Scientific). 4. Agarose powder (Invitrogen™,USA). 5. 10Â TBE buffer pH 8.0: Add 108 g Tris base, 55 g Boric acid and 7.4 g EDTA in 1 L dH2O, autoclave and kept in room temperature. 6. QIAquick Gel extraction kit (Qiagen, Germany).

® 2.4 Two-Step 1. NcoI and corresponding buffer (NEB Inc., USA). ® Cloning Approach 2. MluI and corresponding buffer (NEB Inc., USA). ® 3. 100Â BSA (NEB Inc., USA). 2.4.1 First Step Cloning ® (Fab HC) 4. Antarctic phosphatase and corresponding buffer (NEB Inc., USA). ® 5. T4 DNA Ligase and corresponding buffer (NEB Inc., USA). 6. 3 M sodium acetate pH 5.2. 7. ElectroMAX Escherichia coli, E. coli DH10β cells (Invitro- gen™, USA) À :F mcrA Δ(mrr-hsdRMS-mcrBC) Φ80lacZΔM15 À ΔlacX74 recA1 endA1araD139Δ(ara, leu)7697 galU galK λ rpsL nupG. 8. Micropulser™ electroporator (BIO-Rad, USA). 9. 50 mg/mL ampicillin stock solution: Add 0.5 g ampicillin in 10 mL of 50% (v/v) ethanol, ﬁlter-sterilize and store at À20 C. 10. 2YT media pH 7.2: Add 16 g tryptone, 10 g yeast extract and 5 g NaCl in 1 L dH2O, autoclave and store at room temperature. 11. 2YT media containing 100 μg/mL ampicillin and 2% glucose. Human Fab Library Generation 29

Table 1 Primers used for first amplification Fab HC and LC gene repertoire

Primer name Primer sequence

Fab HC first amplification primers Human Fab IgM CH1 Rv 50- TGGAAGAGGCACGTTCTTTTCTTT -30 Human Fab IgG CH1 Rv 50- TCTTGTCCACCTTGGTGTTG -30 VH1 Fw 50- CAG GTC CAG CTK GTR CAG TCT GG -30 VH157 Fw 50 - CAG GTG CAG CTG GTG SAR TCT GG -30 VH2 Fw 50 – CAG RTC ACC TTG AAG GAG TCT G -30 VH3 Fw 50- GAGGTGCAGCTGKTGGAGWCY- 30 VH4 Fw 50- CAGGTGCAGCTGCAGGAGTCSG -30 VH4 DP63 Fw 50- CAGGTGCAGCTACAGCAGTGGG -30 VH6 Fw 50- CAGGTACAGCTGCAGCAGTCA -30 Fab LC first amplification primers scFv Fab Lambda CL1Rv 50- TGAACATTCTGTAGGGGCCACTG -30 scFv Fab Lambda CL2 Rv 50- TGAACATTCCGTAGGGGCAACTG -30 scFv Fab kappa CL Rv 50- ACACTCTCCCCTGTTGAAGCTCTT -30 Vλ 1Fw 50- CAGTCTGTSBTGACGCAGCCGCC -30 Vλ 1459 Fw 50- CAGCCWGKGCTGACTCAGCCMCC -30 Vλ 15,910 Fw 50- CAGTCTGYYCTGAYTCAGCCT -30 Vλ 2Fw 50- TCCTATGWGCTGACWCAGCCAA -30 Vλ 3Fw 50- TCCTCTGAGCTGASTCAGGASCC -30 Vλ 3DPL16 Fw 50- TCCTATGAGCTGAYRCAGCYACC -30 Vλ 338 Fw 50- AATTTTATGCTGACTCAGCCCC -30 Vλ 6Fw Vκ 1Fw 50- GACATCCRGDTGACCCAGTCTCC -30 Vκ 246 Fw 50- GGATATTGTGMTGACBCAGWCTCC-30 Vκ 3Fw 50- GGAAATTGTRWTGACRCAGTCTCC-30 Vκ 5Fw 50- GGAAACGACACTCACGCAGTCTC-30

12. 2YT agar plate containing100 μg/mL ampicillin and 2% glucose. 13. Sterile Petri dish, 25 cm (Brandon). 14. Sterile Bio-assay dish, 25 cm Â 25 cm (Nunc). 15. Plasmid puriﬁcation Maxi-prep kit (Qiagen, Germany). 30 Noorsharmimi Omar and Theam Soon Lim

Table 2 Primers used for second amplification for restriction endonucleases site

Primer name Primer sequence (50 –30)

RE sites introduction for HC genes VH Fab IgM CH1 MluI Rv 50-ATGACGCGTTGGAAGAGGCACGTTCTTTTCTTT-30 VH Fab IgG CH1 MluI Rv 50-ATGACGCGTTCTTGTCCACCTTGGTGTTG-30 VH1 NcoI Fw 50- CCCAGCCGGCCATGGCCCAGGTCCAGCTKGTRCAGTC TGG-30 VH157 NcoI Fw 50- CCCAGCCGGCCATGGCCCAGGTGCAGCTGGTGSART CTG-30 VH2 NcoI Fw 50- CCCAGCCGGCCATGGCCCAGRTCACCTTGAAGGAGT CTG-30 VH3 NcoI Fw 50- CCCAGCCGGCCATGGCCGAGGTGCAGCTGKTGGAG WCY-30 VH4 NcoI 50- CCCAGCCGGCCATGGCCCAGGTGCAGCTGCAGGAGT CSG-30 VH4 DP 63 NcoI Fw 50- CCCAGCCGGCCATGGCCCAGGTGCAGCTACAGCAGT GGG-30 VH6 NcoI Fw 50- CCCAGCCGGCCATGGCCCAGGTACAGCTGCAGCAG TCA-30 RE sites introduction for LC genes Human Lambda 1NheI Rv 50- CTTGCTAGCTTATGAACATTCTGTAGGGGCCACTG-30 Human Lambda 2 NheI Rv 50- CTTGCTAGCTTATGAACATTCCGTAGGGGCAACTG-30 Human kappa CL NheI Rv 50- CTTGCTAGCTTAACACTCTCCCCTGTTGAAGCTCTT-30 Vλ 1 SalI Fw 50- TGTGACAAAGTCGACGCAGTCTGTSBTGACGCAGCC GCC-30 Vλ 1459 SalI Fw 50- TGTGACAAAGTCGACGCAGCCTGTGCTGACTCARYC-30 Vλ 15910 SalI Fw 50- TGTGACAAAGTCGACGCAGCCWGKGCTGACTCAGCC MCC-30 Vλ 2 Sal I Fw 50- TGTGACAAAGTCGACGCAGTCTGYYCTGAYTCAGCCT- 30 Vλ 3 Sal I Fw 50- TGTGACAAAGTCGACGTCCTATGWGCTGACWCAGC CAA-30 Vλ3DPL16 Fw 50- TGTGACAAAGTCGACGTCCTCTGAGCTGASTCAGGA SCC-30 Vλ338 SalI Fw 50- TGTGACAAAGTCGACGTCCTATGAGCTGAYRCAGCY ACC-30 Vλ 6 SalI Fw 50- TGTGACAAAGTCGACGAATTTTATGCTGACTCAGC CCC-30 (continued) Human Fab Library Generation 31

Table 2 (continued)

Primer name Primer sequence (50 –30)

Vκ 1 SalI Fw 50- TGTGACAAAGTCGACGGACATCCRGDTGACCCAGTC TCC-30 Vκ 246 SalI Fw 50- TGTGACAAAGTCGACGGATATTGTGMTGACBCAGWC TCC-30 Vκ 3 SalI Fw 50- TGTGACAAAGTCGACGGAAATTGTRWTGACRCAGTC TCC-30 Vκ 5 SalI Fw 50- TGTGACAAAGTCGACGGAAACGACACTCACGCAGT CTC-30

16. 80% Glycerol: Add 80 mL glycerol in 20 mL of dH2O, autoclave and store at room temperature. 17. pLABEL-Fab phagemid vector.

® 2.4.2 Second Step 1. SalI HF and corresponding buffer (NEB Inc., USA). ® Cloning (Fab LC) 2. NheI HF corresponding buffer. (NEB Inc., USA). 3. E. coli XL1 Blue cells (Agilent Technologies, USA): recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac [F´ proAB lacIq ZΔM15 Tn10 (Tetr). 4. Other materials are the same as ﬁrst step cloning.

2.5 Colony PCR Oligonucleotide primers used: 1. LMB3 Fw: 50 – CAG GAA ACA GCT ATG AC – 30. 2. pIII Rv: 50 – TTA GAT CGT TAC GCT AAC - 30.

2.6 Fab Phage- 1. 2YT media pH 7.2: Add 16 g Tryptone, 10 g Yeast extract and Library Packaging 5 g NaCl, autoclave and store at room temperature. 2. 2YT media containing 100 μg/mL ampicillin and 2% glucose. 3. 2YT media containing 100 μg/mL ampicillin and 60 μg/mL kanamycin. ® 4. M13 K07 helper phage (NEB Inc.,USA). 5. 20% Polyethylene glycol 6000/2.5 M NaCl (PEG/NaCl) solution: Add 200 g PEG and 146 g NaCl in 1 L of dH2O, autoclave and store at room temperature. 6. 1 M Isopropyl-β-D-thiogalactopyranoside, IPTG dioxane free: Add 2.38 g IPTG in 10 mL dH2O and store in 1 mL aliquots. 7. 1Â PBS buffer. 32 Noorsharmimi Omar and Theam Soon Lim

® ® 2.7 Fab Antibody 1. Corning Costar 96 Microtiter strip well, ﬂat-bottom plate Panning (Sigma Aldrich, Inc). 2. Carbonate bi-carbonate coating buffer, pH 9.6: Add 3.03 g 2.7.1 Antigen Coating Na2CO3, 6.0 g NaHCO3 in 1 L dH2O, autoclave and keep in on Microtiter Wells 4 C. 3. 0.5% PBS-T: Add 5 mL Tween 20 into 1 L PBS. 4. 2% (w/v) PTM blocking buffer: Add 2 g skimmed milk in 100 mL 0.1% PBS-T (Need to be prepared freshly).

2.7.2 Fab Phage Library 1. 2YT media containing 100 μg/mL ampicillin and 60 μg/mL Packaging kanamycin. 2. 2YT media containing 100 μg/mL ampicillin and 2% glucose). 3. 20% Polyethylene glycol 6000/2.5 M NaCl (PEG/NaCl) solution: Add 200 g PEG and 146 g NaCl in 1 L of dH2O, autoclave and store at room temperature. 4. 1Â PBS (for phage library titration).

2.7.3 Fab Selection 1. 0.5% PBS-T: Add 5 mL Tween-20 into 1 L PBS.

2. 10 μg/mL Trypsin: Add 10 μg Trypsin in 1 mL dH2O. 3. E. coli XL1 Blue F0 (Agilent Technologies, USA): recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac [F´ proAB lacIq ZΔM15 Tn10 (Tetr). 4. 2YT media, pH 7.2: Add 16 g Tryptone, 10 g Yeast extract and 5 g NaCl in 1 L dH2O, autoclave and store at room temperature. 5. 2YT media containing kanamycin 60 μg/mL. 6. 2YT media containing100 μg/mL ampicillin and 2% glucose). 7. Sterile 10 cm disposable Petri dishes. 8. 1 M Isopropyl-β-D-thiogalactopyranoside, IPTG dioxane free: Add 2.38 g IPTG in 10 mL dH2O and store in 1 mL aliquots.

2.8 Phage Library, 1. 2YT amp/glu agar plates: Add 31 g premixed 2 YT and 15 g Rescued and Ampliﬁed agar in 1 L dH2O, autoclave, add 2% glucose and 100 μg/mL Phage Titration ampicillin. 2. 2YT kan agar plates: Add 31 g premixed 2 YT and 15 g agar in 1LdH2O, autoclave, add 2% glucose and 100 μg/mL kanamycin.

2.9 Polyclonal Fab 1. 2% (w/v) PTM blocking buffer: Add 2 g skimmed milk in Antibody Phage ELISA 100 mL 0.1% PBS-T (Need to be prepared freshly). 2. 0.5% PBS-T: Add 5 mL Tween 20 into 1 L PBS. 3. Anti-M13 Horseradish peroxidase: Prepare 1:5000 dilution in 2% PTM. Human Fab Library Generation 33

4. 2,20-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) substrate, ABTS developing solution: add 10 mg tablet ABTS in 5 mL of 50 mM citric acid, 5 mL of 50 mM trisodium citrate and 10 μLH2O2. Store in the dark.

5. 2.0 M Sulphuric Acid, H2SO4. (to Stop the ABTS Reaction).

3 Methods

3.1 Blood Collection Blood sample can be collected from healthy or disease related donors for human naive and immune library generation respectively (see Note 1 (a, b)). 1. Extract B-cells from 10 mL whole blood using Ficoll-Paque™ Plus (see Note 2). 2. The isolated lymphocytes were diluted with 1Â PBS, pH 7.4 and centrifuged at 12,000 Â g, 15 min, 18 C. 3. Remove the plasma cells using a clean ﬁltered pippete tip. 4. Gently transfer the lymphocytes layer into a new 50 mL conical tube and ﬁll with 1Â PBS until 50 mL. 5. Centrifuge the mixture at 12,000 Â g for 10 min at 18 C.

3.2 Total RNA 1. Extract total RNA from the B-cells using a total RNA extrac- Preparation tion kit based on the manufacturer’s protocols (see Note 3). 2. Keep total RNA at À80 C until use.

3.3 cDNA Synthesis 1. Synthesize the cDNA using SuperscriptTM II Reverse Tran- scriptase according to the manufacturer’s procedure. 2. A reaction mix will contain 500 ng RNA of each donor, 1 μLof oligo dT 12–18 (500 μg/mL), 0.1 μL random hexamer (3 μg/ mL), and 1 μL of 10 mM dNTP and filled up to a final volume of 12 μL with dH2O. 3. Denature the RNA at 65 C for 5 min and immediately chill down on ice. 4. Add 4 μL5Â first-strand buffer (250 mM Tris-HCl pH 8.3, 375 mM KCl and 5 mM MgCl2), 2 μL of 0.1 M DTT and 40 units of RNAseOUT™ into the first reaction mixture. 5. Incubate the mixture at 42 C for 2 min. 6. Add 200 units of the SuperscriptTM II Reverse Transcriptase and mix it gently. 7. Next, incubate the mixture at 42 C for 50 min and heat inactivate it at 70 C for 15 min. 8. Add 2 units of RNAse H and incubate at 37 C for 20 min. 9. Store the cDNA at À20 C until use. 34 Noorsharmimi Omar and Theam Soon Lim

3.4 V-Gene (Heavy 1. Different reverse primers are used for the different libraries. and Light Chain) The naive library uses the IgM-derived heavy chain V-genes Amplification by PCR reverse primer and 19 different VH forward primers are used for full Fab amplification. The immune library requires the IgG-derived heavy chain V-genes reverse primer is use instead with the same VH forward primers. (All the primers are shown in Table 1)(see Note 4). 2. For light chain amplification, the kappa (κ) and lambda (λ) reverse primers are used with six different λ forward primers and 4 different κ primers. (All the primers are shown in Table 1). 3. A total of 200 ng cDNA is used as a template for V-gene amplification. 4. In a total of 20 μL, add 200 μM dNTPs, 0.2 μM forward and reverse primers, 2 μLof10Â Pfu buffer (MgSO4), and lastly 0.2 μL of 2.5 U/μL Pfu DNA Polymerase. Add dH2O until 20 μL. The PCR program is as follows: 95 C for 2 min, annealing at 55 Cor62C for 1 min 30 sec, elongation at 72 C for 2 min and a final elongation at 72 C for 5 min. Perform the PCR with 30 cycles of amplification (see Note 5). 5. Purify the amplified V-gene (VH—750 bp, VL kappa/lambda À750 bp) products separated on the 1% agarose gel using the QiaQuick Gel extraction kit according to the manufacturer’s protocol. 6. Determine concentrations of each donor sample and store the products at À20 C until used.

3.5 Restriction Sites 1. A total of 20 μg of first amplification of V-genes products is Amplification used as DNA template for the amplification of the restriction of the V-Gene by PCR site. Perform the PCR individually. 2. In a total of 20 μL, add 10 μM of dNTPs, 0.2 μM forward and ® reverse primers, 2 μLof10Â Thermo pol buffer, and lastly 0.2 μL of 2000 U/mL Vent DNA polymerase. Add dH2O until 20 μL. The PCR program is as follows: 95 C for 1 min 30 s, annealing at 55 Cor62C for 30 s, elongation at 72 C for 2 min, and a final elongation at 72 C for 5 min. Perform the PCR for 20 cycles of amplification. 3. Pool all VH and VL (κ and λ) into individual and subfamilies separately.

3.6 Two-Step The cloning of the V-genes was carried out with a two-step cloning Cloning strategy. The ﬁrst step is to generate a mini library containing the Fab VH repertoire in pLABEL-Fab using DH10β E. coli cells. The second cloning step is to introduce the VL (κ and λ) repertoire to the VH-mini library in XL1 Blue E. coli cells. Human Fab Library Generation 35

3.6.1 VH Cloning (First 1. Digest 10 μg of pLABEL-Fab and 2 μg VH pools with 50 U ® Step) and 10 U of NcoI and MluI (NEB Inc., USA) respectively at 37 C for 16 hours (h). All the subfamilies are pooled according to five subsequently smaller groups of five donors. Heat inactivate the digestion mixture at 65 C for 10 min (see Note 6). 2. Add 1 U of alkaline phosphatase and incubate at 37 C for 1 h. 3. Purify the treated products using QIAquick PCR Purification Kit according to the manufacturer’s protocol. 4. Ligate 500 ng of digested pLABEL-Fab and 250 ng digested VH pools using 1 U of T4 DNA Ligase (see Notes 7 and 8). 5. Incubate the ligation mixture for 16 h at 8 C. 6. Heat inactivate the mixture at 65 C for 10 min. 7. Precipitate the ligated DNA using 1 μL of 3 M sodium acetate and 100 μL of 100% ethanol, incubate in À80 C for 1 h and centrifuge for 10 min at 12,000 Â g at 4 C(see Note 9). 8. Wash the DNA pellet with 70% Ethanol.

9. Resuspend the DNA pellet with 4 μLdH2O. 10. Mix 2 μL of DNA with 50 μL of thawed DH10β E. coli cells on ice for 2 min. 11. Transfer the 52 μL mixture to a prechilled 0.1 cm eletropora- tion cuvette. 12. Electroporate the mixture using the bacteria program using an electroporator (see Note 8). 13. Add 1 mL of pre-warmed 2YT media into the mixture and transfer the suspension to a 1.5 mL microcentrifuge tube. Shake the culture for 1 h at 37 C and 700 rpm. 14. Plate out the transformations on 30 bioassay dishes (25 cm Â 25 cm) of 2YT agar containing 100 μg/mL ampicillin and 2% glucose. 15. Incubate the plates at 37 C for 16 h. 16. Scrape the colonies with 300 mL 2YT containing 100 μg/mL ampicillin and 2% glucose. 17. Centrifuge the cells at 8000 Â g for 5 min and resuspend the cell pellet with 50 mL 2YT containing 100 μg/mL ampicillin and 2% glucose. 18. A glycerol stock of the library is prepared (see Subheading 3.7). 19. Estimate the library diversity by titrating the library stock (see Subheading 3.8). 20. Perform a colony PCR to conﬁrm cloning of the heavy chain repertoire insert (see Subheading 3.9) and calculate the library size (see Notes 8 and 10). 36 Noorsharmimi Omar and Theam Soon Lim

3.6.2 VL Cloning (Second 1. Culture a tube of glycerol stock containing the VH-mini library Step) at 37 C for 16 h with 800 rpm shaking in 500 mL of 2YT media containing 100 μg/mL ampicillin and 2% glucose for plasmid extraction using the maxi-prep kit. 2. Digest the VH-mini library and VL (κ and λ) with SalI HF and NheI HF with its compatible buffer. 3. Perform the digestion and ligation as ﬁrst step cloning (see Subheading 3.6.1). 4. Electroporate the Fab library in XL1Blue MRF’ E. coli cells for phage packaging purposes. 5. Plate out the culture into 40 bioassay dish plates. 6. Perform the library size estimation as described (see Subhead- ing 3.8, Note 11). 7. Colony PCR is carried out to determine the cloning efﬁciency of the library (see Subheading 3.9).

3.7 Preparation 1. The collected colonies are then scraped into 300 mL 2YT of Bacteria Library containing100 μg/mL ampicillin and 2% glucose. Stock 2. Centrifuge the collected 2YT suspension for 10 min at 8000 Â g. 3. Resuspend the library pellet with 80 mL 2YT containing 100 μg/mL ampicillin and 2% glucose. 4. Add 100% glycerol to the suspension to make 20% library stock to preserve the library. 5. Aliquot the ﬁnal library suspension into several 2.0 mL cryogenic tubes and keep in À80 C until needed (see Note 12).

3.8 Library Size 1. Dilute a volume of 10 μL from 80 mL bacteria library suspen- Estimation sion with 90 μL 2YT containing 100 μg/mL ampicillin and 2% glucose. À 2. Perform a serial dilution by 1:10 ratio up to 10 13. 3. Spot the diluted cultures on the 2YT agar plate containing 100 μg/mL ampicillin and 2% glucose. 4. Incubate all the plates o/n at 37 C. 5. Count the number of colonies observed from the plate for single colony growth. 6. Colony PCR can be performed using randomly picked colonies to conﬁrm Fab inserts (see Subheading 3.9).

3.9 Colony PCR 1. Pick 20–50 colonies randomly and resuspend in 10 μLdH2O. for Fab Insert 2. Perform a PCR for each colony using the LMB3 Fw and pIII Determination Rv primers. Human Fab Library Generation 37

3. The PCR reaction mix contains 2 μL template, 2 μL10Â Dream Taq Buffer, 1 μLof10μM forward and reverse primers, 0.4 μL of 10 mM dNTPs and 0.2 μL of Taq polymerase and made up to a final volume of 20 μL with dH2O. 4. The PCR program is as follows; 95 C for 1 min 30 s, annealing at 55 Cor62C for 30 s, elongation at 72 C for 2 min, and a final elongation at 72 C for 5 min. Perform the PCR for 20 cycles amplification. For all VH and Vκ genes amplification the annealing temperature is 55 C while for the Vλ Vκ genes the annealing temperature is at 62 C. 5. The Fab size of ~1.6 kb should be observed on 0.9% agarose gel electrophoresis (see Notes 8 and 13).

3.10 Sequencing 1. Pick 20–50 colonies randomly and resuspend in 10 μLdH2O. Analysis 2. The quality of the library is further conﬁrmed by DNA sequencing. Send the positive colony PCR clones with the correct size for full Fab insertion for sequencing. 3. DNA sequencing data can be analyzed using IMGT/V-quest software (www.imgt.org/IMGT_vquest/vquest) using the human reference directory [27]. 4. The IMGT/V-quest analysis will provide V-D-J gene segment and CDR analysis of the clones.

3.11 Fab Phage 1. To package the Fab library, inoculate a glycerol stock of the Library Packaging gene library stock into 500 mL 2YT containing 100 μg/mL ampicillin and 2% glucose in a 1 L flask. Grow at 37 C, 250 rpm up to an OD600 ~ 0.5. 2. Infect a volume of 250 mL bacteria culture with ~1012 M13KO7 helper phage (see Note 14). 3. Incubate the bacteria culture at 37 C for 30 min static. 4. Transfer the cells into several 50 mL conical tubes and centrifuge at 2400 Â g for 30 min at 12 C to remove the excess glucose that suppresses the expression of the HC: pIII protein as well as the free LC protein. 5. Resuspend the pellet with 300 mL 2YT containing 100 μg/mL ampicillin, 60 μg/mL kanamycin, and 100 μM IPTG in a 1 L flask. Grow at 30 C, 200 rpm for o/n (see Note 15). 6. Next day, spin down the bacteria cells at 2400 Â g for 30 min at 12 C. 7. Transfer the supernatant to a new 50 mL conical tube filling it up to 4/5 volume of the tube and add the remaining 1/5 volume with PEG/NaCl and mix well. 8. Incubate the mixture on ice for 1 h (see Note 16). 9. Centrifuge the mixture at 2400 Â g for 30 min at 12 C. 38 Noorsharmimi Omar and Theam Soon Lim

10. Gently pour the supernatant away and resuspend the phage pellet in 8 mL 1Â PBS and immediately ﬁll with 2 mL PEG/NaCl. 11. Continue to incubate the mixture on ice for 1 h. 12. Centrifuge the mixture at 4800 Â g for 15 min at 12 C. 13. Again, gently pour the supernatant away and resuspend the phage pellet in 4 mL 1Â PBS. 14. Next, centrifuge the mixture at 4800 Â g for 15 min at 12 C. 15. To remove any remaining bacterial cells from the phage pre- cipiatation, repeat steps 13 and 14 for three times. 16. Finally, resuspend the phage pellet in 2 mL freshly prepared 1Â PBS and store at 4 C until use (see Note 17).

3.12 Fab Phage 1. Dilute 10 μL of Fab phage library in 90 μL1Â PBS. The serial À Library Titration dilution is performed until a range of 10 14. 2. Infect the Fab phage dilutions with E. coli TG1 or XL1Blue cells at OD600 ~ 0.5. 3. Spot 10 μL of the infected phage on a 2YT agar plate containing 100 μg/mL ampicillin and 2% glucose and on a second 2 YT agar plate containing 60 μg/mL kanamycin (see Note 14). 4. Dry the agar plates under laminar ﬂow. 5. Incubate the agar plates at 37 C for o/n.

3.13 Antigen Prior to panning, evaluate the antigens using SDS-PAGE and Analysis Western-blot analysis to conﬁrm antigen purity. The purity of the antigen used will ensure a good and speciﬁc isolation of Fab antibody binders.

3.14 Phage Selection 1. Coat a microtiter well with a minimum amount of antigen (Panning) (10 μg) o/n at 4 C using the Carbonate bi-carbonate coating buffer. The wells are then washed 3Â with 300 μL PBS-T. Block the antigen-coated wells with 300 μL of 2% PTM for 1 h at RT and wash the wells 3Â with 300 μL PBS-T. 2. Concurrently, pre-block 1011 phage particles of the Fab phage library (see Note 18) with 300 μL 2% PTM for 1 h on a microtiter well (see Note 19). 3. Transfer the pre-blocked Fab phage library into the antigen- coated wells and incubate for 2 h at RT with constant shaking at 700 rpm (see Note 20). 4. Wash the wells with PBS-T for 3Â in the ﬁrst round; gradually increase the wash to 5Â and 10Â for each subsequent panning round. Human Fab Library Generation 39

5. Elute the bound phages with 50 μL of 100 μg/mL Trypsin and incubate at 37 C for 30 min. 6. Remove the eluted phages to a new 1.5 mL microcentrifuge tube and infect with 150 μL XL1 Blue E. coli cells OD600 ~ 0.5 static. 7. Incubate the cells for 30 min at 30 C and continue incubation at 37 C, shake at 700 rpm for 30 min. 8. Take out 10 μL of infected phage to determine the rescued phage by titration (see Subheading 3.12). 9. Add 20 μLof10Â ampicillin and glucose into the remaining mixture. 10. Incubate the mixture at 37 C, 250 rpm o/n. 11. Remove 10 μL of o/n cultured cells and continue culturing in a new cell culture plate containing 190 μL of 2YT ampicillin and 2% glucose for 2 ½ h at 37 C and shake at 700 rpm. The remaining o/n culture is added with 20 μL of glycerol and mixed well for storage at À80 C. 12. Infect the cultured cells with 1011 M13KO7 helper phage. 13. Incubate the cells at 37 C for 30 min static. 14. Transfer the infected cells onto the ﬁlter plate and centrifuge the plate at 2400 Â g for 5 min (see Note 21). 15. Resuspend the cell pellet with 2YT containing 100 μg/mL ampicillin, 60 μg/mL kanamycin, and 100 μM IPTG. 16. Culture the cells at 30 C o/n shaking at 700 rpm. 17. Finally, transfer the packaged Fab phage from round 1 to a new ﬁlter plate to separate phage and cells by centrifugation at 2400 Â g for 5 min. 18. Apply 100 μL of the phage for subsequent round of panning. The remaining phage is kept at 4 C for polyclonal evaluation. The phage titer from each round of panning is determined for the recovery and enrichment analysis (see Note 22). The schematic representations of Fab antibody phage library packaging and Fab phage selection on microtiter well is shown in Fig. 1.

3.15 Polyclonal Fab 1. Coat an appropriate number of microtiter wells with a mini- Antibody Phage ELISA mum amount of antigen (10 μg) using the carbonate bi-carbonate coating buffer for 1 h at RT and wash 3Â with 300 μL PBS-T. Block the antigen-coated wells with 300 μLof 2% PTM for 1 h at RT and wash the wells 3Â with 300 μL PBS-T. Concurrently, coat an equal number of empty wells with 300 μL of 2% PTM as negative controls to observe for nonspeciﬁc binders. 40 Noorsharmimi Omar and Theam Soon Lim

Fig. 1 (a) A schematic review of a Fab antibody phage library packaging performed using M13KO7 helper phage with 100 μM IPTG induction to improve Fab antibody phage packaging efﬁciency; (b) A schematic presentation of Fab antibody phage selection (panning) against target antigen on microtiter wells

2. Pre-block the phage obtained from each round of panning with 300 μL of 2% PTM for 1 h at RT to reduce nonspeciﬁc binding during antigen recognition. 3. Incubate the pre-cleanup phage of each panning round in the antigen-coated wells as well as the control plates for 2 h at RT and shake at 700 rpm. 4. Wash the wells 3Â with 300 μL PBS-T (see Note 23). 5. Add 200 μL of anti-M13-HRP (1: 5000 dilution in 2% PTM) into samples and control wells. 6. Incubate the wells for 1 h at RT and shake 700 rpm. Human Fab Library Generation 41

7. Wash the wells 3Â with 300 μL PBS-T. 8. Lastly, develop the wells with 100 μL ABTS solution for 30 min in the dark. 9. Measure the readings at 405 nm using a microtiter plate spectrophotometer for absorbance values of the ABTS color change.

3.16 Polyclonal Fab For a successful panning process, an obvious enrichment is Antibody Phage ELISA observed as the panning round increases. Choose the best round Assesment of panning to isolate monoclonal antibodies with good speciﬁcity and diversity as this is critical to isolate different monoclonal antibodies with varying afﬁnities. 1. Infect the phage of that particular round from the storage plate with XL1 Blue E. coli cells of OD600 ~ 0.5 or plate out the glycerol stock and continue with monoclonal Fab antibody phage ELISA.

3.17 Monoconal Fab 1. For a 96 microtiter plate, pick 47 monoclonal antibodies from Antibody Phage ELISA the respective round of panning including a positive control of any in house antibody clone as positive control. 2. Inoculate randomly selected phage monoclonal antibodies in 200 μL 2YT containing 100 μg/mL ampicillin and 2% glucose. 3. Grow the clones at 37 C o/n with constant shaking at 700 rpm. 4. Follow steps 11 until 17 of the panning method to package the phage monoclonal Fab antibodies (see Subheading 3.14). 5. For ELISA analysis, follow steps 1 until 9 of the polyclonal phage ELISA (see Subheding 3.15). 6. DNA sequencing of the positive monoclonal Fab antibodies with a satisfactory signal-to-noise ratio is performed with the appropriate primers. The primer sets are scFv Fab (CL1 þ CL2) Rv with LMB3 Fw, scFv Fab Kappa Rv with LMB3 Fw and Human Fab IgM/IgG CH1 Rv with Fab Spacer NheI Fw.

4 Notes

1. (a) Donors with family backgrounds of severe illnesses, suffering from autoimmune disorder, had fever, on any medi- cation including antibiotics and immunosuppressors within a month are not suitable for naive library sampling. (b) Donor history including treatment and state of illness should be recorded for immune library sampling. The type of treatment regimen is important to standardize the library generation process to ensure the donors are capable of eliciting an 42 Noorsharmimi Omar and Theam Soon Lim

immune response toward the disease. Donors that have under- gone immunosuppresors are not suitable as the immune response is ineffective. Conditions may vary depending on the mechanism of disease. 2. The blood layer must not mix with the Ficoll-Paque™ PLUS as it leads to aggregation of erhythrocytes resulting in lower yields. 3. The total RNA obtained can be analyzed by Agilent Bioanaly- zer for RNA integrity and total concentration. 4. The amplification of the VH Fab repertoire for naive library generation will use the IgM CH1 Rv-specific primer, whereas the immune library will use the IgG CH1 Rv-specific primer. This is to amplify isotype specific repertoire of the heavy chain. A band size of ~750 bp Fab fragment is expected; however, the presence of other bands is expected due to the rearranged V-genes. The excision of the targeted band size must be carried out with care. 5. The V-gene repertoire from each donor is amplified independently to avoid bias and loss of repertoire by sample pooling. This is to ensure higher repertoire diversity. 6. Repeat digestion until the amount of DNA required is sufficient for the corresponding cloning experiment. A sequential digestion may be required if the reaction buffer compatibility affects the digestion efficiency. 7. Multiple ligation reactions may be required to obtain sufficient amount of DNA material for a highly diverse antibody. 8. The library size for the heavy chain repertoire should be as high as possible as the variable heavy chain is the predominant region for antigen binding. The working range should be within107–109 or higher for naive libraries and 105–107 for immune libraries. The highest possible diversity should be achieved for a good quality library. 9. Alternatively, liquid nitrogen may be used for snap freeze to allow a rapid purification process. After snap freezing, centrifuge for 10 min at 12,000 Â g at 4 C. 10. The expected band size of the VH repertoire cloning is approximately 800 bp. 11. The targeted range of the final library size should be 109–1012 or higher for naive libraries and 106–108 for immune libraries. These are acceptable working sizes although larger library sizes are preferred. 12. The bacteria library stocks are normally prepared to allow a starting OD value of OD600 ~ 0.1. A tube of immune or naive Fab bacteria library stock is enough to be cultured for phage Fab library preparation. Human Fab Library Generation 43

13. The cloning efficiency may vary between users. A cloning efficiency of more than 80% inserts of VH-mini library and full Fab library is preferred. 14. A normal ratio of helper phage to be used when packaging is to ensure 102 more helper phage from the library size. This is to ensure at least 100 copies of each clone are represented in the packaging process. 15. The addition of 100 μM IPTG during Fab phage packaging is to aid the packaging process by inducing the expression of the HC: pIII fused protein and LC monomers during phage library packaging. This will increase the natural formation of Fab antibody molecules on the phage particle. 16. For a higher phage library packaging efficiency, the ice-cold incubation can be pro-longed to 2 h to increase the amount of phage particles to be precipitated. 17. Avoid storing the phage preparation for more than 2–4 weeks at 4 C. 18. For better panning results, fresh phage preparations are used for each panning round to avoid loss from proteolysis of the displayed antibodies through time. 19. To check the quality of the packaged Fab phage library, the colony growth on ampicillin agar plate should be at least two-fold higher in comparison with the kanamycin plate. This is to make sure less helper phage is being produced concurrently with the phage library. In addition, the colony growth can be used to quantify the phage library size prior for panning. 20. As a guide, the amount of phage titer to be used for the first panning round should be twofolds higher than the library size. 21. Occasionally, clogging may occur during cell separation using the filter plates. If clogging occurs, remove the remaining cells to another well and proceed. 22. The phage recovery (Input/Output) can be calculated from the titer of the input and output phage used for each panning round. 23. Care must be taken to ensure no bubbles are present during washing steps by agitation. Tap the microtiter wells on a pile of clean dry paper towel to remove any remaining PBS-T in the microtiter wells.

Acknowledgments

The authors would like to acknowledge the support of the Malaysian Ministry of Education through the Higher Institution Centre of Excellence (HICoE) Grant (Grant No. 311/CIPPM/4401005). 44 Noorsharmimi Omar and Theam Soon Lim

References

1. Smith GP (1985) Filamentous fusion phage: 16. Lim BN, Chin CF, Choong YS, Ismail A, Lim novel expression vectors that display cloned TS (2016) Generation of a naı¨ve human single antigens on the virion surface. Science 228 chain variable fragment (scFv) library for the (4705):1315–1317 identification of monoclonal scFv against sal- 2. Barbas CF, Burton DR, Scott JK, Sivrmann GJ monella Typhi Hemolysin E antigen. Toxicon (2004) Phage display. Cold Spring Harbor 117:94–101 Laboratory Press, Cold Spring Harbor, NY 17. Zhu ZD, Dimitrov S (2009) Construction of a 3. Soumillion P, Jespers L, Bouchet M, Marchand- large naive human phage-displayed fab library Brynaert J (1994) Phage display of enzymes and through one-step cloning. Therapeutic Anti- in vitro selection for catalytic activity. Appl Bio- bodies 6:129–142 chem Biotechnol 47(2–3):175–189 18. Rahumatullah A, Ahmad A, Rahmah N, Lim 4. Fujita S, Taki T, Taira K (2005) Selection of an TS (2015) Delineation of BmSXP antibody active enzyme by phage display on the basis of V-gene usage from a lymphatic filariasis based the enzyme’s catalytic activity in vivo. Chem- immune scFv antibody library. Mol Immunol biochem 6(2):315–321 67:512–523 5. Zoller F, Haberkorn U, Mier W (2011) Mini- 19. Shen Y, Yang X, Dong N, Xie X et al (2007) proteins as phage display-scaffolds for clinical Generation and selection of immunized fab applications. Molecules 16(3):2467–2485 phage display library against human B cell lym- 6. Uchiyama F, Tanaka Y, Minari Y, Tokui N (2005) phoma. Cell Res 17(7):650–660 Designing scaffolds of peptides for phage display 20. Nelson AL (2010) Antibody fragments: hope libraries. J Biosci Bioeng 99(5):448–456 and hype. MAbs 2(1):77–83 7. Schirrmann T, Hust M (2010) Construction of 21. Ward ES, Gussow D, Griffiths AD, Jones PT, human antibody gene libraries and selection of Winter G (1989) Binding activities of a reper- antibodies by phage display. Methods Mol Biol toire of single immunoglobulin variable 651:177–209 domains secreted from Escherichia coli. Nature 8. Lim BN, Tye GJ, Choong YS, Ong EB, 341(6242):544–546 Ismail A, Lim TS (2014) Principles and appli- 22. Ro¨thlisberger D, Honegger A, Pluckthun€ A cation of antibody libraries for infectious dis- (2005) Domain interactions in the fab frag- eases. Biotechnol Lett 36(12):2381–2392 ment: a comparative evaluation of the single- 9. Bain B, Brazil M (2003) Adalimumab. Nat Rev chain Fv and fab format engineered with vari- Drug Discov 2(9):693–694 able domains of different stability. J Mol Biol 347(4):773–789 10. Bradbury A, Velappan N, Verzillo V, Ovecka M € et al (2003) Antibodies in proteomics I: gen- 23. Hust M, Dubel S (2004) Mating antibody erating antibodies. Trends Biotechnol 21 phage display with proteomics. Trends Bio- (6):275–281 technol 22(1):8–14 11. Ponsel D, Neugebauer J, Ladetzki-Baehs K, 24. Shahsavarian MA, Le Minoux D, Matti KM, Tissot K (2011) High affinity, developability Kaveri S et al (2014) Exploitation of rolling and functional size: the holy grail of combina- circle amplification for the construction of torial antibody library generation. Molecules large phage-display antibody libraries. J Immu- 16(5):3675–3700 nol Methods 407:26–34 12. Krebs B, Rauchenberger R, Reiffert S, Rothe C 25. Kirsch M, Zaman M, Meier D, Dubel S, Hust (2001) High-throughput generation and engi- M (2005) Parameters affecting the display of neering of recombinant human antibodies. J antibodies on phage. J Immunol Methods 301 Immunol Methods 254(1):67–84 (1):173–185 13. Konthur Z, Walter G (2002) Automation of 26. Kipriyanov SM, Moldenhauer G, Little M phage display for high-throughput antibody (1997) High level production of soluble single development. Targets 1(1):30–36 chain antibodies in small-scale Escherichia coli cultures. J Immunol Methods 200(1–2):69–77 14. Janeway CA, Travels P, Walport M, Capra JD (1999) Immunobiology: the immune system in 27. Giudicelli V, Chaume D, Lefranc MP (2004) health and disease, vol 157. Current Biology IMGT/V-QUEST, an integrated software pro- Publications, New York gram for immunoglobulin and T cell receptor V-J and V-D-J rearrangement analysis. Nucleic 15. Kindt T, Goldsby R, Osborne B (2007) Kuby Acids Res 32:435–440 immunology, 6th edn. WH Freeman, New York Chapter 3

Construction of Synthetic Antibody Phage-Display Libraries

Johan Nilvebrant and Sachdev S. Sidhu

Abstract

Synthetic antibody libraries provide a vast resource of renewable antibody reagents that can rival or exceed those of natural antibodies and can be rapidly isolated through controlled in vitro selections. Use of highly optimized human frameworks enables the incorporation of defined diversity at positions that are most likely to contribute to antigen recognition. This protocol describes the construction of synthetic antibody libraries based on a single engineered human autonomous variable heavy domain scaffold with diversity in all three complementarity-determining regions. The resulting libraries can be used to generate recombinant domain antibodies for a wide range of protein antigens using phage display. Furthermore, analogous methods can be used to construct antibody libraries based on larger antibody fragments or second- generation libraries aimed to fine-tune antibody characteristics including affinity, specificity, and manufac- turability. The procedures rely on standard reagents and equipment available in most molecular biology laboratories.

Key words Human antibody, Antibody fragment, Domain antibody, Phage display, Protein engineering, Degenerate oligonucleotide

1 Introduction

Antibodies are well established as affinity reagents and therapeutic drugs. They can be generated with high affinity and specificity, are long lived in serum and generally well tolerated as drugs. They are by far the most widely used group of specific protein detection reagents and a growing class of biopharmaceuticals. In the post- genomic era demand for robust and reliable antibody reagents is ever increasing [1]. Monoclonal antibodies have traditionally been generated by hybridoma technology where splenocytes derived from immunized animals are harvested and fused with an immortalized myeloma cell line [2]. Although effective, this approach is laborious and expensive, and results in nonhuman and potentially immunogenic antibodies. Moreover, the natural immune system imposes restrictions that make it difficult to raise antibodies against certain antigens, including self-antigens and highly conserved proteins across

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_3, © Springer Science+Business Media LLC 2018 45 46 Johan Nilvebrant and Sachdev S. Sidhu

species, toxic antigens, and antigens that are unstable in serum. Some of these limitations are also shared by more recent strategies, including generation of human antibodies in transgenic mice engineered to produce a human immunoglobulin repertoire followed by hybridoma [3]. Antibody repertoires displayed on phage provide an attractive alternative to circumvent animal immunization and other caveats of hybridoma technology. In vitro selection methods can be used to target almost any protein and offer more precise control, higher throughput, and adaptability to automation [4]. Phage display, originally introduced by George Smith in 1985 [5], has become the most widely used selection method for human antibodies. Antibody phage-display technology [6, 7] has been used to generate and improve a large number of antibodies for research and medical use [8, 9]. Pools of billions of unique antibodies displayed on phage are subjected to selective pressure for antigen recognition. Binding clones are amplified by infection of an Escherichia coli host and used for additional rounds of selection. Eventually, the population is dominated by antigen-binding clones, which can be screened and subjected to DNA sequencing to decode the sequences of the displayed antibodies. Antibody phage display is now an established method for robust generation of reliable antibody reagents in vitro. A key advantage is that the unique sequence of each antibody is encoded in the encapsulated phage DNA, which allows for facile downstream manipulation or reformatting to optimize antibody properties. Furthermore, use of well- characterized antibody reagents defined by their sequence is critical to help battle the current ántibody reproducibility crisis´ [10, 11]. Phage-displayed antibody libraries commonly integrate diversity from immune- or non-immune natural B-cell sources. How- ever, this approach is limited by the diversity provided by the natural adaptive immune system of the host. Nowadays, molecular details of antibody structure and function are so well understood that defined diversity can be encoded in degenerate synthetic DNA and introduced in regions most likely to contribute to antigen binding in a defined framework. Use of optimized human frameworks can minimize the risk of immunogenicity and thus the need for humanization while ensuring high stability and protein production [12]. Use of synthetic libraries allows control over both library design and selection conditions and facilitates the generation of antibodies with precisely engineered binding specificities. Immunoglobulin G is the most common antibody class in humans. It is composed of two heavy chains and two light chains. The antigen-binding site is formed by three hypervariable loops on each variable domain. Since the structure of the variable domains is only slightly influenced by the diversity in the complementarity determining regions (CDRs) [13], a single framework can accommodate an array of CDR sequences and binding properties. Our Synthetic Antibody Libraries 47

work has shown that a single framework based on the highly stable therapeutic human 4D5 scaffold can support remarkably diverse antibody functions [12]. Furthermore, highly simpliﬁed library designs, which engage fewer CDRs and may encode as little as two possible residues per randomized position, have been shown to be capable of producing speciﬁc synthetic antibodies against a multitude of protein targets [12, 14, 15]. Complex architecture, requirement of expression in mammalian cells, relatively poor tissue penetration, and, sometimes, undesired Fc-mediated effector functions have inspired progressive reduction of the size of the antibody molecule (Fig. 1a). Smaller fragments such as fragment antigen binding (Fab) and single-chain fragment variable (scFv) can retain the binding properties of the parental antibody while enabling high-yield production in prokary- otic expression systems. Domain antibodies consisting of a single- variable domain represent the smallest antibody fragments capable of mediating antigen recognition [16, 17]. Highly functional single-framework libraries with various degrees of CDR diversity can be constructed using any of these smaller antibody fragments [18–21]. This protocol describes the construction of highly diverse synthetic domain antibody libraries built on a single-human VH domain (Fig. 1b)[22, 23] cloned into a phagemid vector. Follow- ing bacterial transformation and infection with helper phage, a phage-displayed library containing billions of individual clones can be used for the rapid isolation of recombinant domain antibodies targeting virtually any protein antigen.

2 Materials

Prepare all the solutions using MilliQ water and analytical grade reagents. 1. 0.2 cm gap electroporation cuvette (BTX Harvard Apparatus, Holliston, MA). 2. 10 mM ATP.

3. 10Â TM buffer: 0.1 M MgCl2, 0.5 M Tris, pH 7.5. 4. 100 mM dithiothreitol (DTT). 5. 14 ml round-bottomed tube (Falcon 352,059). 6. 2YT medium: 10 g bacto-yeast extract, 16 g bacto-tryptone, 5 g NaCl. Add water to 1.0 l; adjust pH to 7.0 with NaOH, and autoclave. 7. 2YT/carb/cmp medium: 2YT, 100 μg/ml carbenicillin, 10 μg/ml chloramphenicol. 48 Johan Nilvebrant and Sachdev S. Sidhu

Fab a V L CDRH1

VH CH1

CH2 Fc

CH3 b CDRH3 CDRH2 MKKNIAFLLASMFVFSIATNAYASEISEVQLVESGGGLVQPGRSLRLSCAASGFNIKDTYIGWVRRAPGKGEELVARIYPTNG YTRYADSVKGRFTISADTSKNTAYLQMNSLRAEDTAVYYCARWGGDGFYAMDYWGQGTLVTVSSADKTHTCGRPSGSGDFDYE KMANANKGAMTENADENALQSDAKGKLDSVATDYGAAIDGFIGDVSGLANGNGATGDFAGSNSQMAQVGDGDNSPLMNNFRQY LPSLPQSVECRPFVFSAGKPYEFSIDCDKINLFRGVFAFLLYVATFMYVFSTFANILRNKES* c CDRH1 CDRH2: Template - FNIKDTYI G - Template - R I Y P T N G Y T R - Oligoes - F A I SY SY SY SY I G - Oligoes - R I SY PS SY SY GS SY T SY -

CDRH3: Template - R W G G D G F Y ------A M D - Oligoes - R X X X X X X X X X X X X X X X X X AG FILM D - 1-17 aa in length (X = Y, S, G, A, F, W, H, P and V in ratio 5:4:4:2:1:1:1:1:1)

Fig. 1 Autonomous domain antibody scaffold and diversification. (a) Structure of IgG1 (PDB 1IGT) with fragment antigen binding (Fab) and fragment crystallizable (Fc) indicated. IgG is a homotetramer consisting of two heavy chains each built from three constant- (CH1, CH2, and CH3) and one variable (VH) domain and two light chains each comprising a constant (CL) and a variable (VL) domain. The backbone is shown as green or gray tubes for the VH or other domains, respectively. An engineered autonomous human VH domain (PDB 3B9V) is used as a scaffold to introduce diversity in three complementarity-determining regions shown in purple (CDRH1), yellow (CDRH2), and red (CDRH3). Spheres represent paratope positions that are diversified. The figure was generated using PyMOL (http://www.pymol.org/). (b) Sequence of the fusion protein designed to enable phage display of an autonomous VH domain (bold) fused to the truncated protein III (gray). The signal sequence stII (underlined) directs the VH domain-pIII fusion to the periplasm and the dimerization domain (dashed box) is used to achieve bivalent display. (c) Amino acids encoded in three mutagenic oligonucleotide sets used to introduce synthetic antibody diversity. All oligonucleotides can be used simultaneously and are designed with at least 15 nucleotides complementary to the template sequence on either side of the region to be randomized. The distance between annealing oligonucleotides used in the same mutagenesis reaction should be at least 15 base pairs. Binary diversity is used in CDRH1 and CDRH2. In CDRH3, “X” denotes nine amino acids (Y, S, G, A, F, W, H, P, V in a ratio of 5:4:4:2:1:1:1:1:1) encoded by a custom trimer phosphoramidite mixture containing the indicated ratio of nine trimers. A mixture of oligonucleotides is used to introduce length diversity in CDRH3 Synthetic Antibody Libraries 49

8. 2YT/carb/kan medium: 2YT, 100 μg/ml carbenicillin, 25 μg/ ml kanamycin. 9. 2YT/carb/kan/uridine medium: 2YT, 100 μg/ml carbenicillin, 25 μg/ml kanamycin, 0.25 μg/ml uridine. 10. 2YT/tet medium: 2YT, 10 μg/ml tetracycline. 11. 3 M sodium acetate, pH 5.0. 12. 96-microwell round-bottom plate (Corning). 13. Baffled E-flasks (250 and 2000 ml). 14. Carbenicillin: 100 mg/ml in water, filter-sterilize. 15. Chloramphenicol: 10 mg/ml in ethanol, filter-sterilize. 16. dNTP mix: solution containing 10 mM each of dATP, dCTP, dGTP, and dTTP. 17. ECM-630 electroporation system (BTX). 18. E. coli CJ236 (New England Biolabs, Ipswich, MA). 19. E. coli OmniMax 2 T1R (Invitrogen, Grand Island, NY). 20. E. coli SS320 (Lucigen, Middleton, WI). 21. Kanamycin: 25 mg/ml in water, filter-sterilize. 22. LB/carb plates: LB agar, 100 μg/ml carbenicillin. 23. LB/kan plates: LB agar, 25 μg/ml kanamycin. 24. LB/tet plates: LB agar, 20 μg/ml tetracycline. 25. M13 K07 helper phage (New England Biolabs). 26. MLB buffer: 1 M sodium perchlorate, 30% (v/v) isopropanol. 27. MP buffer: dissolve 3.3 g citric acid monohydrate in 3 ml USP water at room temperature. Filter through a 0.2 μm syringe filter to give ca. 6 ml buffer. 28. Phosphate-buffered saline (PBS): 137 mM NaCl, 3 mM KCl, 8mMNa2HPO4, 1.5 mM KH2PO4. Adjust pH to 7.2 with HCl, autoclave. 29. PBST: PBS, 0.05% (v/v) Tween 20. 30. PEG/NaCl: 20% PEG-8000 (w/v), 2.5 M NaCl. Mix and filter-sterilize. 31. Phenylmethane sulfonyl fluoride (PMSF): 100 mM in 96% ethanol. 32. QIAquick Gel Extraction Kit (Qiagen, Valencia, CA). 33. QIAprep Spin Miniprep Kit (Qiagen). 34. SOC medium: 5 g bacto-yeast extract, 20 g bacto-tryptone, 0.5 g NaCl, 0.2 g KCl. Add water to 1.0 l and adjust pH to 7.0 with NaOH, autoclave. Add 5.0 ml of autoclaved 2.0 M MgCl2 and 20 ml of filter-sterilized 1.0 M glucose. 35. Spectrophotometer. 50 Johan Nilvebrant and Sachdev S. Sidhu

36. SYBR Safe DNA gel stain (Invitrogen). 37. T4 DNA ligase (New England Biolabs). 38. T7 DNA polymerase (New England Biolabs). 39. T4 polynucleotide kinase (New England Biolabs). 40. TAE buffer: 40 mM Tris–acetate, 1.0 mM EDTA; adjust pH to 8.0; autoclave. 41. TAE/agarose gel: TAE buffer, 1.0% (w/v) agarose, 1:10,000 (v/v) SYBR Safe DNA gel stain. 42. TE buffer: 10 mM Tris-HCl, 1 mM EDTA, pH 8.0. 43. Tetracycline: 10 mg/ml in 70% ethanol, ﬁlter-sterilize. 44. Ultrapure glycerol. 45. Ultrapure irrigation U.S. Pharmacopeia (USP) water (B. Braun medical Inc., Bethlehem, PA). 46. Uridine: 0.25 mg/ml in water, ﬁlter-sterilize.

3 Methods

The following sections describe optimized protocols for the construction of phage-displayed libraries containing in excess of 1010 variants. This synthetic antibody diversity can rival or exceed that of the human periphery. A parental antibody framework is cloned into a phagemid vector to enable display on filamentous phage particles. The framework sequence in the phagemid is modified to introduce appropriate genetic diversity into the CDRs. Passing the genetic library through an E. coli host generates a phage-displayed antibody library that can be used for selections to isolate antigen-binding clones. This protocol uses an engineered human autonomous variable heavy domain (dAb) as a library framework and incorporates diversity in three CDRs (Fig. 1b, c). This scaffold was generated by systematic mutagenesis of a VH3 domain derived from an approved therapeutic antibody to yield a variant with structurally compatible hydrophilic substitutions at the former light chain interface that promote autonomous behavior [22]. These libraries are routinely used to select human dAbs against a variety of targets with affinities in the nanomolar range. A protocol to engineer new alternative autonomous VH domain scaffolds is found in Tonikian and Sidhu [24].

3.1 Phagemid Design This protocol describes the construction of a synthetic dAb library in a phagemid vector. A phagemid (Fig. 2a) is a specialized vector with a double-stranded DNA origin of replication (dsDNA ori), which allows replication in E. coli, and a ﬁlamentous phage ori (f1 ori) to enable packaging of single-stranded DNA (ssDNA) into phage particles. Rather than inserting the antibody genes Synthetic Antibody Libraries 51 ab Template DNA

stII signal peptide dut-/ung-) Promoter dU-ssDNA

Annealing of mutagenic V domain * H oligonucleotides

CDRH1, CDRH2 and CDRH3 *V

Truncated gIII AmpR Synthesis of

* heteroduplex dsDNA *V f1 ori CCC-dsDNA dsDNA ori dut+/ung+) * * * * *!! *

* * Wild-type * * * Mutants *

Fig. 2 Phagemid design and library construction workﬂow. (a) Phagemid vector designed for the display of autonomous VH domains. The vector contains origins of single-stranded f1- (f1 ori) and a double-stranded DNA (dsDNA ori) replication and a selectable marker that confers resistance to carbenicillin (AmpR). An N-terminal stII signal sequence directs the VH-pIII fusion protein into the periplasm. (b) dU-ssDNA template is prepared from phage particles produced by CJ236 E. coli cells harboring the phagemid and superinfected with M13 K07 helper phage. Mutations (asterisk) are introduced through annealing of phosphorylated oligonucleotides on the dU-ssDNA template. Using T7 DNA polymerase and T4 DNA ligase, the oligonucleotides are extended and ligated to form heteroduplex covalently closed ds-DNA (CCC-dsDNA). Following transformation into the dut+/ ung+ E. coli SS320, the mutated strand is preferentially replicated while the uracil-containing parental strand is degraded

directly into the phage genome fused to a coat protein encoding gene, antibody expression can thus be separated from phage propagation by providing the passenger antibody:coat protein fusion on a separate plasmid. Our phagemid has been used for the display of Fabs [25, 26], scFvs [27], VH domains [13], peptides, and other polypeptides [28, 29]. Monomeric scFvs or VH domains are displayed by direct fusion to the N-terminus of the C-terminal domain of the truncated M13 bacteriophage minor coat protein 3 (pIII). Secretion signals direct the pIII-fusion to the periplasm. The isopropyl β-D-1-thiogalactopyranoside (IPTG)-inducible promoter (Ptac) is used in the phagemid for the display of VH domains. 52 Johan Nilvebrant and Sachdev S. Sidhu

Incomplete repression of the promoter during phage production allows low expression of the VH-pIII fusion in the absence of limiting IPTG [18]. Display of heterodimeric Fabs requires bicis- tronic expression. Co-infection with helper phage such as M13 K07 is used to provide additional components necessary to assemble new virions, which also contain phagemid-encoded pIII-fusions.

3.2 Library This protocol is a scaled-up and optimized version based on the Construction mutagenesis method of Kunkel et al. [30]. Mutagenic oligonucleotides are incorporated into heteroduplex covalently closed, circular, double-stranded DNA (CCC-dsDNA) by a three-step procedure (Fig. 2b). dU-ssDNA is used as a template for annealing of phosphorylated mutagenic oligonucleotides to prime the original strand for extension by T7 DNA polymerase followed by ligation by T4 DNA ligase. Upon transformation of the heteroduplex CCC-dsDNA into a dutþ/ungþ host, the uracil-containing parental DNA strand is degraded whereas the mutated strand is preferentially replicated and propagated as a double-stranded plasmid. This procedure results in the formation of ca. 20 μg of highly pure product that can be electroporated into an E. coli host containing an F0episome to enable M13 bacteriophage infection and propagation. This is sufﬁcient to construct a library containing more than 1010 unique members. One of the major advantages of using this method is the ability to simultaneously mutate multiple CDRs in a single reaction without any need for restriction sites. Precise control over library design can be achieved by using mutagenic oligonucleotides that contain degenerate codons to introduce deﬁned diversity at desired positions. Moreover, CDR length-diversity can easily be introduced by using pools of degenerate oligonucleotides of varying lengths. Using a template with stop codons introduced in the CDRs intended for randomization can ensure display of only mutated antibodies. Templates with non-mutated CDRs will contain one or several stop codons that prevent expression of functional pIII- fusions and are thereby eliminated from the pool during binding selections.

3.2.1 Purification of dU- The use of highly pure dU-ssDNA is critical for the successful ssDNA Template library construction since mutagenesis efficiency depends on template purity. Template is prepared by using a modified Qiagen QIAprep spin M13 kit protocol. 20 μg is recommended for the construction of one library (see Note 1). 1. Transform the phagemid vector carrying the template sequence to be diversified into competent E. coli CJ236 (or analogous dut-/ung- strain). Plate on LB agar plate supplemented with appropriate antibiotic to select for the vector and grow overnight at 37 C. Synthetic Antibody Libraries 53

2. Pick a single colony of E. coli CJ236 containing the phagemid vector and inoculate 1 ml 2YT medium supplemented with appropriate antibiotics and M13 K07 (1010 pfu/ml) in a 14 ml round-bottomed tube. For example, 2YT/carb/cmp medium contains carbenicillin to select for a phagemid carrying a β-lactamase gene and chloramphenicol to select for the F0episome of E. coli CJ236. 3. Incubate at 37 C with 200 rpm for 2 h before the addition of kanamycin (25 μg/ml) to select for bacteria co-infected with helper phage M13 K07. 4. Shake for 6 h at 37 C. 5. Transfer the culture to a baffled 250 ml E-flask containing 30 ml 2YT/carb/kan/uridine medium and incubate at 37 C and 200 rpm for 20 h. 6. Transfer cultures to 50 ml Falcon tubes and pellet bacteria by centrifuging at 27,000 Â g for 10 min at 4 C. 7. Transfer the phage-containing supernatant to a new tube containing 1/5 final volume of PEG/NaCl and incubate for 5 min on ice to precipitate phage. 8. Pellet precipitated phage by centrifugation at 27,000 Â g for 20 min at 4 C. Decant the supernatant. Centrifuge at 2000 Â g for 2 min and carefully aspirate the remaining supernatant. Always use filter tips when handling phage to avoid pipette contamination. 9. Use a pipette to resuspend the phage pellet in 0.5 ml PBS and transfer to a 1.5 ml microcentrifuge tube. 10. Remove residual cell debris by centrifugation for 5 min at 15,800 Â g in a bench top microcentrifuge at room temperature (RT). All the microcentrifuge steps are performed at RT. Transfer the supernatant to a fresh microcentrifuge tube. 11. Add 7 μl MP buffer and mix. Incubate at RT for at least 2 min. The solution should become cloudy when phages are precipitated in this step. 12. Apply the sample to a QIAprep spin column (Qiagen) in a 2 ml collection tube. Spin for 30 s at 6000 Â g in a microcentrifuge and discard the flow-through. The phage particles remain bound to the column matrix. 13. Add 700 μl MLB buffer to the column. Spin for 30 s at 6000 Â g and discard the flow-through. 14. Add another 700 μl MLB buffer and incubate at RT for at least 1 min. 15. Spin for 30 s at 6000 Â g and discard the flow-through. The DNA is separated from the protein coat and remains adsorbed to the matrix. 54 Johan Nilvebrant and Sachdev S. Sidhu

16. Add 700 μl buffer PE (Qiagen), centrifuge for 30 s at 6000 Â g, and discard the ﬂow-through. 17. Add an additional 700 μl buffer PE (Qiagen) and centrifuge for 30 s at 6000 Â g to remove residual proteins and salt. 18. Centrifuge the empty column for 30 s at 6000 Â g to remove residual PE buffer. 19. Discard the collection tube and transfer the QIAprep column to a fresh 1.5 ml microcentrifuge tube. Add 100 μl buffer EB (Qiagen; 10 mM Tris–HCl, pH 8.5) to the center of the column membrane. Incubate at RT for 10 min. 20. Spin for 30 s at 6000 Â g and save the eluted puriﬁed dU-ssDNA. 21. Analyze the dU-ssDNA by electrophoresing 1 μl on a TAE/a- garose gel. The DNA should appear as a predominant single band (Fig. 3). However, faint bands with lower electrophoretic mobility are often visible, which likely represent secondary structure in the dU-ssDNA. 22. Use absorbance at 260 nm to measure DNA concentration (A260 ¼ 1.0 for 33 ng/μl of dU-ssDNA). Typical dU-ssDNA concentrations range from 200 to 500 ng/μl. 23. Aliquot 20 μg for each library to be prepared and store frozen.

A B C

Fig. 3 Electrophoretic analysis of in vitro synthesis of heteroduplex covalently closed circular, double-stranded DNA (CCC-dsDNA). Lane 1: DNA markers; Lane 2: uracil-containing single-stranded DNA template (dU-ssDNA); Lane 3: Product from the heteroduplex CCC-dsDNA synthesis reaction. The lower band (C)is correctly extended and ligated CCC-dsDNA, the middle band (B) is knicked dsDNA and the upper band (A) is strand-displaced dsDNA Synthetic Antibody Libraries 55

3.2.2 In Vitro Synthesis 1. For each mutagenic oligonucleotide, mix 0.6 μg oligonucleo- of Heteroduplex tide with 2 μl10Â TM buffer, 2 μl 10 mM ATP, and 1 μl CCC-dsDNA: 100 mM DTT in a 1.5 ml microcentrifuge tube. Add ultrapure Oligonucleotide irrigation water (USP water) to a total volume of 20 μl(see Phosphorylation Note 2). 2. Add 20 U T4 polynucleotide kinase to each oligonucleotide and incubate for 1 h at 37 C. Transfer reactions to ice and use as soon as possible for annealing (see Note 3).

3.2.3 In Vitro Synthesis 1. To 20 μg dU-ssDNA template in a microcentrifuge tube, add of Heteroduplex 25 μl10Â TM buffer, 20 μl of each phosphorylated oligonu- CCC-dsDNA: Annealing cleotide (or oligonucleotide pool, see Note 4), and USP water of Phosphorylated to a ﬁnal volume of 250 μl. Assuming a oligonucleotide to a Oligonucleotides to the dU- template length ratio of 1:100, these DNA quantities provide ssDNA Template an oligonucleotide:template molar ratio of 3:1. 2. Incubate at 90 C for 3 min, 50 C for 3 min, and RT for 5 min using a thermo cycler or dry block heaters.

3.2.4 In Vitro Synthesis 1. To the annealed oligonucleotide/template mixture, add 10 μl of Heteroduplex 10 mM ATP, 15 μl 100 mM DTT, 25 μl 10 mM dNTP mix, CCC-dsDNA: Enzymatic 30 Weiss units T4 DNA ligase (5 μl 400 U/μl), and 30 U (3 μl Synthesis of CCC-dsDNA 10 U/μl) T7 DNA polymerase. 2. Mix and incubate at 20 C overnight. 3. Optional: Analyze 1 μl of the reaction mixture alongside the dU-ssDNA template on a TAE/agarose gel (see step 11). 4. Purify and desalt the DNA using the QIAquick Gel Extraction kit (Qiagen). Add 1 ml buffer QG (Qiagen) and mix (see Note 5). 5. Apply half of the sample to each of two QIAquick spin columns placed in 2 ml collection tubes. 6. Spin at 15,800 Â g for 1 min in a microcentrifuge and discard the flow-through. 7. Add 750 μl buffer PE (Qiagen) to each column. Incubate 2–5 min at RT (see Note 6), centrifuge for 1 min at 15,800 Â g, and discard the flow-through. 8. Spin the empty columns at 15,800 Â g for 1 min to remove excess buffer PE. 9. Transfer the columns to fresh 1.5 ml microcentrifuge tubes and add 35 μl USP water to the center of each membrane. Incubate for 10 min at RT. 10. Spin at 15,800 Â g for 1 min to elute the purified DNA. Combine the eluates from the two columns and determine DNA concentration by measuring absorbance at 260 nm (A260 ¼ 1.0 for 50 ng/μl dsDNA). The total recovery should 56 Johan Nilvebrant and Sachdev S. Sidhu

be at least 20 μg. The DNA can be used immediately for E. coli electroporation or stored frozen for later use. 11. Analyze 1 μl of the CCC-dsDNA by electrophoresis alongside the ssDNA template (Fig. 3). A successful reaction will result in near complete conversion of ssDNA to dsDNA, which has lower electrophoretic mobility. Usually, two product bands are visible and no ssDNA should remain. The lower band with higher electrophoretic mobility represents the desired product: correctly extended and ligated CCC-dsDNA with a high mutation frequency (ca. 80%) and high E. coli transformation efficiency. The band with lower mobility is a strand- displaced product, which results from undesirable activity of T7 DNA polymerase [31]. It provides a low (ca. 20%) mutation frequency and at least 30-fold lower transformation efficiency than CCC-dsDNA. A band with intermediate mobility between the other two product bands is sometimes visible. It represents correctly extended product but contains unligated dsDNA and may result from incomplete oligonucleotide phosphorylation or insufficient T4 DNA ligase activity.

3.3 Conversion The final step of library preparation requires transformation of the of CCC-dsDNA into heteroduplex CCC-dsDNA into an E. coli host containing the F´ a Phage-Displayed episome to enable M13 bacteriophage infection and propagation. Antibody Library We use an E. coli strain (SS320) that is ideal for both high-efficiency electroporation and phage production [32]. Protocols to prepare M13 K07 helper phage and electrocompetent E. coli SS320 pre-infected with M13 K07 can be found in [24]. Once transformed with a phagemid, each cell will be able to produce phage particles without the need for further helper phage infection. 1. Chill the purified, desalted CCC-dsDNA (20 μg in a maximum volume of 100 μl) and a 0.2 cm gap electroporation cuvette on ice. 2. Pre-warm SOC medium in a water bath at 37 C(2Â 1mlin 1.5 ml microcentrifuge tubes and 25 ml in a 250 ml baffled E-flask). 3. Thaw a 350 μl aliquot of electrocompetent E. coli SS320 on ice. Add the cells to the DNA and mix gently by pipetting several times (avoid introducing air bubbles). 4. Transfer the mixture to the cuvette, wipe the outside with paper tissue, and electroporate according to the manufacturer’s instructions. We use a BTX ECM-630 electroporation system with the following settings: 2.5 kV field strength, 125 Ω resistance, and 50 μF capacitance. 5. Immediately rescue the electroporated cells by adding 1 ml pre-warmed SOC medium with a 1 ml sterile stripette and Synthetic Antibody Libraries 57

transfer to 25 ml pre-warmed SOC medium in a 250 ml baffled E-flask. Rinse the cuvette with 1 ml SOC medium. 6. Incubate at 37 C for 30 min with shaking at 200 rpm. 7. Prepare serial dilutions and plate on LB/carb plates to determine library diversity. Transfer 10 μl from the culture flask and make 8 tenfold serial dilutions in 90 μl 2YT medium in a round-bottomed 96-microwell plate. Plate 5 μl of each dilution using a multi-pipette. Optional: Plate on LB/tet and/or LB/kan plates to determine total cell concentrations and titer of M13 K07 infected cells, respectively (see Note 7). Incubate the plates at 37 C overnight. 8. Transfer the culture to a 2 l baffled E-flask containing 500 ml 2YT/carb/kan medium for the selection for phagemid and M13 K07 helper phage, respectively. 9. Incubate at 37 C and 200 rpm overnight. 10. After overnight incubation (ca. 18 h) transfer to two 1 L centrifuge bottles and pellet bacteria by centrifugation for 10 min at 16,000 Â g at 4 C. 11. Inoculate 25 ml 2YT/tet medium with a single colony of E. coli OmniMax 2 T1R from a fresh LB/tet plate. Grow at 37 C and 200 rpm to mid-log phase (OD600 ¼ 0.6–0.8) and use for phage titration (see step 19). 12. Transfer the supernatants to fresh centrifuge bottles containing 1/5 total volume of PEG/NaCl solution to precipitate phage. Incubate for 20 min on ice. 13. Spin for 20 min at 16,000 Â g at 4 C to pellet precipitated phage. Decant the supernatant. Spin briefly (2 min 4000 Â g) and remove the remaining supernatant with a pipette. 14. Resuspend each phage pellet in 20 ml of prechilled TE buffer supplemented with 0.5 mM PMSF by gentle pipetting. 15. Combine the resuspended phage pellets and transfer to a clean 50 ml Falcon tube. 16. Pellet insoluble matter by centrifuging at 16,000 Â g for 10 min at 4 C. 17. Transfer the supernatant to a clean tube containing 1/5 volume of PEG/NaCl. Incubate on ice for 20 min to precipitate phage. 18. Spin at 16,000 Â g for 20 min at 4 C and resuspend phage pellet in 4 ml PBST. 19. Determine phage concentration by infecting log-phase E. coli OmniMax 2 T1R cells with serial dilutions of phage: Dilute 10 μlin90μl 2YT and prepare 12 tenfold dilutions. Transfer 10 μl of each dilution to a 96-well round bottom plate and add 90 μl of log-phase E. coli OmniMax 2 T1R cells. Incubate still 58 Johan Nilvebrant and Sachdev S. Sidhu

for 30 min at 37 C and plate 5 μl of each dilution on LB/carb plates. Optional: plate samples on LB/tet and LB/kan plates to determine cell number and helper phage concentration (see Note 8). 20. We recommend that the phage-displayed antibody library be used directly for selection experiments. Alternatively, it can be stored frozen at À80 C following the addition of glycerol to a ﬁnal concentration of 10% and EDTA to a ﬁnal concentration of 2 mM. Several protocols describing phage-display selection strategies, screening, and expression of synthetic antibodies have been published [18, 19, 21, 24].

4 Notes

1. This protocol is based on the recently discontinued Qiagen QIAprep Spin M13 Kit for dU-ssDNA purification. The QIA- prep Spin Miniprep Kit can be used with MP buffer prepared as described in the Subheading 2. Moreover, the MLB buffer originally provided in the QIAprep Spin M13 Kit was replaced with PB buffer, which resulted in lower yield of dU-ssDNA. We recommend using MLB buffer for lysis of phage particles to achieve comparable yield and quality. 2. If length variation is desired, pools of oligonucleotides can be prepared and phosphorylated. It is recommended to test all oligonucleotides in small-scale annealing and enzymatic reactions using 1/20 of the volumes described above (Subheading 3.2.2) followed by gel electrophoresis to confirm efficient oligonucleotide-mediated conversion of ssDNA to dsDNA. 3. It is recommended to use phosphorylated oligonucleotides immediately for the synthesis of CCC-dsDNA. However, they can be stored at À20 C for up to a month without a significantly reduced performance. 4. If many different oligonucleotides are used to introduce length diversity in one CDR, they may be split into sub-pools during the mutagenesis reactions and pooled prior to electroporation (see Ref. 19 for an example). 5. DNA adsorption to the QIAquick column is only efficient at pH below 7.5, under which the pH indicator in buffer QG is yellow. If the solution turns orange or violet upon addition of reaction mixture, adjust the pH by adding 10 μl 3 M sodium acetate (pH 5.0). Synthetic Antibody Libraries 59

6. Incubation after the addition of buffer PE helps remove salt in the DNA solution, which prevents potential electrical discharge during electroporation. 7. The titers from the LB/tet and LB/kan plates should be approximately the same and carb ca. tenfold lower. Approxi- mately 50% of cells survive after electroporation. 8. The expected phage concentration is 1012–1013 cfu/ml.

Acknowledgments

Members of the Sidhu lab are acknowledged for input, particularly Alia Pavlenco and Wei Ye. This work was supported by funds from the Swedish Research Council (637-2013-468 to J.N.). We thank Frederic Fellouse for assistance with Fig. 3.

References

1. Sidhu SS (2012) Antibodies for all: The case 8. Bradbury AR, Sidhu S, Dubel S, McCafferty J for genome-wide affinity reagents. FEBS Lett (2011) Beyond natural antibodies: the power 586(17):2778–2779. https://doi.org/10. of in vitro display technologies. Nat Biotechnol 1016/j.febslet.2012.05.044 29(3):245–254. https://doi.org/10.1038/ 2. Kohler G, Milstein C (1975) Continuous cul- nbt.1791 tures of fused cells secreting antibody of pre- 9. Frenzel A, Schirrmann T, Hust M (2016) defined specificity. Nature 256 Phage display-derived human antibodies in (5517):495–497 clinical development and therapy. MAbs 8 3. Tomszak F, Weber S, Zantow J, Schirrmann T, (7):1177–1194. https://doi.org/10.1080/ Hust M, Frenzel A (2016) Selection of recom- 19420862.2016.1212149 binant human antibodies. Adv Exp Med Biol 10. Baker M (2015) Reproducibility crisis: blame it 917:23–54. https://doi.org/10.1007/978-3- on the antibodies. Nature 521 319-32805-8_3 (7552):274–276. https://doi.org/10.1038/ 4. Miersch S, Li Z, Hanna R, McLaughlin ME, 521274a Hornsby M, Matsuguchi T, Paduch M, Saaf A, 11. Bradbury A, Pluckthun A (2015) Reproduc- Wells J, Koide S, Kossiakoff A, Sidhu SS (2015) ibility: standardize antibodies used in research. Scalable high throughput selection from Nature 518(7537):27–29. https://doi.org/ phage-displayed synthetic antibody libraries. J 10.1038/518027a Vis Exp 95:51492. https://doi.org/10.3791/ 12. Sidhu SS, Fellouse FA (2006) Synthetic thera- 51492 peutic antibodies. Nat Chem Biol 2 5. Smith GP (1985) Filamentous fusion phage: (12):682–688. https://doi.org/10.1038/ novel expression vectors that display cloned nchembio843 antigens on the virion surface. Science 228 13. Bond CJ, Wiesmann C, Marsters JC Jr, Sidhu (4705):1315–1317 SS (2005) A structure-based database of anti- 6. Huse WD, Sastry L, Iverson SA, Kang AS, body variable domain diversity. J Mol Biol 348 Alting-Mees M, Burton DR, Benkovic SJ, Ler- (3):699–709. https://doi.org/10.1016/j. ner RA (1989) Generation of a large combina- jmb.2005.02.063 torial library of the immunoglobulin repertoire 14. Fellouse FA, Li B, Compaan DM, Peden AA, in phage lambda. Science 246 Hymowitz SG, Sidhu SS (2005) Molecular rec- (4935):1275–1281 ognition by a binary code. J Mol Biol 348 7. McCafferty J, Griffiths AD, Winter G, Chiswell (5):1153–1162. https://doi.org/10.1016/j. DJ (1990) Phage antibodies: filamentous jmb.2005.03.041 phage displaying antibody variable domains. 15. Fellouse FA, Wiesmann C, Sidhu SS (2004) Nature 348(6301):552–554. https://doi. Synthetic antibodies from a four-amino-acid org/10.1038/348552a0 code: a dominant role for tyrosine in antigen 60 Johan Nilvebrant and Sachdev S. Sidhu

recognition. Proc Natl Acad Sci U S A 101 25. Lee CV, Liang WC, Dennis MS, Eigenbrot C, (34):12467–12472. https://doi.org/10. Sidhu SS, Fuh G (2004) High-affinity human 1073/pnas.0401786101 antibodies from phage-displayed synthetic Fab 16. Holt LJ, Herring C, Jespers LS, Woolven BP, libraries with a single framework scaffold. J Mol Tomlinson IM (2003) Domain antibodies: Biol 340(5):1073–1093. https://doi.org/10. proteins for therapy. Trends Biotechnol 21 1016/j.jmb.2004.05.051 (11):484–490. https://doi.org/10.1016/j. 26. Persson H, Ye W, Wernimont A, Adams JJ, tibtech.2003.08.007 Koide A, Koide S, Lam R, Sidhu SS (2013) 17. Nilvebrant J, Tessier PM, Sidhu SS (2016) CDR-H3 diversity is not required for antigen Engineered autonomous human variable recognition by synthetic antibodies. J Mol Biol domains. Curr Pharm Des 22(43): 425(4):803–811. https://doi.org/10.1016/j. 6527–6537. jmb.2012.11.037 18. Adams JJ, Nelson B, Sidhu SS (2014) Recom- 27. Sidhu SS, Li B, Chen Y, Fellouse FA, binant genetic libraries and human monoclonal Eigenbrot C, Fuh G (2004) Phage-displayed antibodies. Methods Mol Biol 1060:149–170. antibody libraries of synthetic heavy chain com- https://doi.org/10.1007/978-1-62703-586- plementarity determining regions. J Mol Biol 6_9 338(2):299–310. https://doi.org/10.1016/j. 19. Chen G, Sidhu SS (2014) Design and genera- jmb.2004.02.050 tion of synthetic antibody libraries for phage 28. Ernst A, Avvakumov G, Tong J, Fan Y, Zhao Y, display. Methods Mol Biol 1131:113–131. Alberts P, Persaud A, Walker JR, Neculai AM, https://doi.org/10.1007/978-1-62703-992- Neculai D, Vorobyov A, Garg P, Beatty L, 5_8 Chan PK, Juang YC, Landry MC, Yeh C, 20. Nelson B, Sidhu SS (2012) Synthetic antibody Zeqiraj E, Karamboulas K, Allali-Hassani A, libraries. Methods Mol Biol 899:27–41. Vedadi M, Tyers M, Moffat J, Sicheri F, https://doi.org/10.1007/978-1-61779-921- Pelletier L, Durocher D, Raught B, Rotin D, 1_2 Yang J, Moran MF, Dhe-Paganon S, Sidhu SS (2013) A strategy for modulation of enzymes 21. Fellouse FA, Sidhu S (2013) Making antibo- in the ubiquitin system. Science 339 dies in bacteria. In: Howard GC, Kase MR (6119):590–595. https://doi.org/10.1126/ (eds) Making and using antibodies: a practical science.1230161 handbook. CRC Press, Boca Raton, FL, pp 151–17222 29. McLaughlin ME, Sidhu SS (2013) Engineer- ing and analysis of peptide-recognition domain 22. Barthelemy PA, Raab H, Appleton BA, Bond specificities by phage display and deep sequenc- CJ, Wu P, Wiesmann C, Sidhu SS (2008) Com- ing. Methods Enzymol 523:327–349. https:// prehensive analysis of the factors contributing doi.org/10.1016/B978-0-12-394292-0. to the stability and solubility of autonomous 00015-1 human VH domains. J Biol Chem 283 (6):3639–3654. https://doi.org/10.1074/ 30. Kunkel TA, Roberts JD, Zakour RA (1987) jbc.M708536200 Rapid and efficient site-specific mutagenesis without phenotypic selection. Methods Enzy- 23. Ma X, Barthelemy PA, Rouge L, Wiesmann C, mol 154:367–382 Sidhu SS (2013) Design of synthetic autonomous VH domain libraries and structural anal- 31. Lechner RL, Engler MJ, Richardson CC ysis of a VH domain bound to vascular (1983) Characterization of strand displace- endothelial growth factor. J Mol Biol 425 ment synthesis catalyzed by bacteriophage T7 (12):2247–2259. https://doi.org/10.1016/j. DNA polymerase. J Biol Chem 258 jmb.2013.03.020 (18):11174–11184 24. Tonikian R, Sidhu SS (2012) Selecting and 32. Sidhu SS, Lowman HB, Cunningham BC, purifying autonomous human variable heavy Wells JA (2000) Phage display for selection of (VH) domains. Methods Mol Biol novel binding peptides. Methods Enzymol 911:327–353. https://doi.org/10.1007/ 328:333–363 978-1-61779-968-6_20 Chapter 4

Modular Construction of Large Non-Immune Human Antibody Phage-Display Libraries from Variable Heavy and Light Chain Gene Cassettes

Nam-Kyung Lee, Scott Bidlingmaier, Yang Su, and Bin Liu

Abstract

Monoclonal antibodies and antibody-derived therapeutics have emerged as a rapidly growing class of biological drugs for the treatment of cancer, autoimmunity, infection, and neurological diseases. To support the development of human antibodies, various display techniques based on antibody gene repertoires have been constructed over the last two decades. In particular, scFv-antibody phage display has been extensively utilized to select lead antibodies against a variety of target antigens. To construct a scFv phage display that enables efﬁcient antibody discovery, and optimization, it is desirable to develop a system that allows modular assembly of highly diverse variable heavy chain and light chain (Vκ and Vλ) repertoires. Here, we describe modular construction of large non-immune human antibody phage-display libraries built on variable gene cassettes from heavy chain and light chain repertoires (Vκ- and Vλ-light can be made into independent cassettes). We describe utility of such libraries in antibody discovery and optimization through chain shufﬂing.

Key words Antibody gene diversity library, Kappa light chain, Lambda light chain, ScFv phage display, Chain shufﬂing, Antibody afﬁnity maturation, Antibody optimization, Human monoclonal antibody

1 Introduction

Antibody gene repertoires from non-immune (naive) human sources have been frequently used to construct antibody-display libraries. To date, several antibody formats, such as single-chain variable fragment (scFv), fragment antigen binding (Fab), or single-domain antibody (sdAb), have been utilized for human antibody-display library generation [1–11]. The scFv form, in which variable heavy chain (VH) and light chain (VL) genes are connected with a ﬂexible linker (typically (Gly4Ser)3, has been widely applied in phage display [12–20]. Peripheral blood mononuclear cells (PBMC) have been commonly used as a source of human B cells to generate the antibody variable gene libraries. In addition, various primary or secondary lymphoid tissues including

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_4, © Springer Science+Business Media LLC 2018 61 62 Nam-Kyung Lee et al.

bone marrow, lymph node, tonsil, or spleen have also been used as a source of VH and VL (κ/λ) gene repertoires [13, 21–23]. In a commonly used method to create the scFv gene, VH and VL (κ/λ) fragments are separately amplified by two independent polymerase chain reactions (PCR) and then assembled by overlap extension PCR [24, 25]. The heavy and light chain repertoires are thus PCR-amplified at least twice before being spliced into the phage-display vector. In addition, library construction using overlapping PCR fragments is often inefficient even with the aid of electroporation. Finally, once assembled and cloned into the display vector, the variable heavy or light chain genes cannot be readily removed and replaced due to the lack of appropriate restriction sites. Some research groups have constructed scFv display libraries by the sequential two-step cloning of VL and VH genes into a phagemid vector [26–28]. This has resulted in improved efficiency in library construction. These methods, however, have generally used two-step PCR amplification strategies with the initial set of primers matching the antibody variable genes and the second set of primers containing overhangs for cloning. Ideally, re-amplification steps should be avoided to reduce bias during scFv gene preparation. In addition, previous two-step cloning strategies either do not have restriction sites to allow ready insertion and removal of the variable gene cassettes [17] or less than optimal selection of restriction sites such as the Hind III site [26–28] that is also present in some antibody heavy and light chain genes (http://www2.mrc- lmb.cam.ac.uk/vbase/). The protocol detailed below describes a modular scheme of library construction where the variable heavy and light chain repertoires are made into cassettes and spliced into the display vector following a single PCR amplification step. VH and VL (κ/λ) genes are amplified by primer sets that encode properly selected flanking restriction enzymes. Separate human Vκ and Vλ gene cassettes are amplified and cloned into a newly designed phagemid vector by the one-step cloning strategy. Likewise, the heavy chain gene cassette is spliced into the phagemid vector by a single cut and paste action, resulting in an independently constructed heavy chain library. The VH cassette is then restriction-digested and spliced into the light chain display libraries, resulting in independent billion-member VH-Vκ and VH-Vλ scFv phage-display libraries. The antibody gene diversity and library size are assessed. These libraries are utilized to select for scFvs binding to target antigens and to optimize lead antibodies by chain shuffling, which is readily performed due to the modular nature of these display libraries. Modular Construction of scFv Phage Display Library 63

2 Materials

® 2.1 One-Step Human 1. Eppendorf Mastercycler pro (Eppendorf, Hauppauge, USA). ® Antibody Gene 2. OneTaq PCR master mix (New England BioLabs, Ipswich, Ampliﬁcation USA). 3. Oligo-nucleotide primers (see Table 1). 4. Agarose low-EEO (Thermo Fisher Scientiﬁc, Waltham, USA). 5. TAE buffer (40 mM Tris, 20 mM glacial acetic acid, 1 mM EDTA, pH 8.0). 6. DNA loading buffer [(0.25% Bromophenol blue, 30% Glycerol (v/v)]. 7. QIAquick gel extraction kit (Qiagen, Germantown, USA). 8. Peripheral Blood Mononuclear Cells (PBMCs) cDNA (Bio- chain, Newark, USA). 9. Human Lymph Node QUICK-Clone™ cDNA (Clontech Laboratories, Mountain View, USA).

2.2 Vκ and Vλ 1. XbaI (New England BioLabs). Library Construction 2. NotI-HF (New England BioLabs). ® 3. CutSmart buffer (New England BioLabs). 4. QIA quick PCR purification kit (Qiagen). 5. T4 DNA ligase (New England BioLabs). 6. Microcon Ultracel YM-10 centrifugal filter (Millipore, Biller- ica, USA). 7. TG1 electrocompetent cells (Lucigen, Middleton, USA). E.coli TG1 genotype: [F´ traD36 proAB lacIqZ ΔM15] supE thi-1 À À Δ(lac-proAB) Δ(mcrB-hsdSM)5 (rK mK ). ® 8. Gene Pulser electroporation cuvettes, 0.1 cm gap (Bio-Rad, Hercules, USA). 9. Electroporator 2510 (Eppendorf). 10. 2xYT broth (Thermo Fisher Scientific). 11. 50% glucose (Thermo Fisher Scientific), filter-sterilized. 12. 100 mg/ml Ampicillin sodium salt (Sigma, St. Louis, USA), filter-sterilized. 13. Polystyrene petri-dishes, 150 mm Â 15 mm (United Scientific Supplies, Waukegan, USA). 14. 2xYT-AG agar plates (2xYT, 100 μg/ml Ampicillin, 2% glucose, 1.5% agar (w/v)). 15. Glycerol (Thermo Fisher Scientific), 60% (v/v). ® 16. QIAGEN plasmid mini kit (Qiagen). 64 Nam-Kyung Lee et al.

Table 1 Oligo-nucleotide primers for one-step PCR amplification of antibody genes and colony PCR

Primer Oligo-nucleotide sequence (50 -30)

VH gene PCR ampliﬁcation NcoVH1aF TCGCAACTGCAATTGCCATGGCCCAGGTKCAGCTGGTGCAG NcoVH1bF TCGCAACTGCAATTGCCATGGCCCAGGTCCAGCTTGTGCAG NcoVH1cF TCGCAACTGCAATTGCCATGGCCSAGGTCCAGCTGGTACAG NcoVH1dF TCGCAACTGCAATTGCCATGGCCCARATGCAGCTGGTGCAG NcoVH2aF TCGCAACTGCAATTGCCATGGCCCAGATCACCTTGAAGGAG NcoVH2bF TCGCAACTGCAATTGCCATGGCCCAGGTCACCTTGARGGAG NcoVH3aF TCGCAACTGCAATTGCCATGGCCGARGTGCAGCTGGTGGAG NcoVH3bF TCGCAACTGCAATTGCCATGGCCCAGGTGCAGCTGGTGGAG NcoVH3cF TCGCAACTGCAATTGCCATGGCCGAGGTGCAGCTGTTGGAG NcoVH4aF TCGCAACTGCAATTGCCATGGCCCAGSTGCAGCTGCAGGAG NcoVH4bF TCGCAACTGCAATTGCCATGGCCCAGGTGCAGCTACAGCAG NcoVH5aF TCGCAACTGCAATTGCCATGGCCGARGTGCAGCTGGTGCAG NcoVH6aF TCGCAACTGCAATTGCCATGGCCCAGGTACAGCTGCAGCAG NcoVH7aF TCGCAACTGCAATTGCCATGGCCCAGGTSCAGCTGGTGCAA NheJH1-2R TCTAATTATGGCGCTAGCTGAGGAGACRGTGACCAGGGTGCC NheJH3R TCTAATTATGGCGCTAGCTGAAGAGACGGTGACCATTGTCCC NheJH4-5R TCTAATTATGGCGCTAGCTGAGGAGACGGTGACCAGGGTTCC NheJH6R TCTAATTATGGCGCTAGCTGAGGAGACGGTGACCGTGGTCCC VK gene PCR ampliﬁcation XbaVK1aF TCTGGCGGTGGCTCTAGARACATCCAGATGACCCAG XbaVK1bF TCTGGCGGTGGCTCTAGAGMCATCCAGTTGACCCAG XbaVK1cF TCTGGCGGTGGCTCTAGAGCCATCCRGATGACCCAG XbaVK1dF TCTGGCGGTGGCTCTAGAGTCATCTGGATGACCCAG XbaVK2aF TCTGGCGGTGGCTCTAGAGATATTGTGATGACCCAG XbaVK2bF TCTGGCGGTGGCTCTAGAGATRTTGTGATGACTCAG XbaVK3aF TCTGGCGGTGGCTCTAGAGAAATTGTGTTGACRCAG XbaVK3bF TCTGGCGGTGGCTCTAGAGAAATAGTGATGACGCAG XbaVK3cF TCTGGCGGTGGCTCTAGAGAAATTGTAATGACACAG XbaVK4aF TCTGGCGGTGGCTCTAGAGACATCGTGATGACCCAG XbaVK5aF TCTGGCGGTGGCTCTAGAGAAACGACACTCACGCAG (continued) Modular Construction of scFv Phage Display Library 65

Table 1 (continued)

Primer Oligo-nucleotide sequence (50 -30)

XbaVK6aF TCTGGCGGTGGCTCTAGAGAAATTGTGCTGACTCAG XbaVK6bF TCTGGCGGTGGCTCTAGAGATGTTGTGATGACACAG NotJK1R AGTCATTCACGACTTGCGGCCGCACGTTTGATTTCCACCTTGGTCCC NotJK2-4R AGTCATTCACGACTTGCGGCCGCACGTTTGATCTCCASCTTGGTCCC NotJK3R AGTCATTCACGACTTGCGGCCGCACGTTTGATATCCACTTTGGTCCC NotJK5R AGTCATTCACGACTTGCGGCCGCACGTTTAATCTCCAGTCGTGTCCC Vλ gene PCR ampliﬁcation XbaVλ1aF TCTGGCGGTGGCTCTAGACAGTCTGTGCTGACTCAG XbaVλ1bF TCTGGCGGTGGCTCTAGACAGTCTGTGYTGACGCAG XbaVλ1cF TCTGGCGGTGGCTCTAGACAGTCTGTCGTGACGCAG XbaVλ2F TCTGGCGGTGGCTCTAGACAGTCTGCCCTGACTCAG XbaVλ3aF TCTGGCGGTGGCTCTAGATCCTATGWGCTGACTCAG XbaVλ3bF TCTGGCGGTGGCTCTAGATCCTATGAGCTGACACAG XbaVλ3cF TCTGGCGGTGGCTCTAGATCTTCTGAGCTGACTCAG XbaVλ3dF TCTGGCGGTGGCTCTAGATCCTATGAGCTGATGCAG XbaVλ4F TCTGGCGGTGGCTCTAGACAGCYTGTGCTGACTCAA XbaVλ5dF TCTGGCGGTGGCTCTAGACAGSCTGTGCTGACTCAG XbaVλ6dF TCTGGCGGTGGCTCTAGAAATTTTATGCTGACTCAG XbaVλ7dF TCTGGCGGTGGCTCTAGACAGRCTGTGGTGACTCAG XbaVλ8dF TCTGGCGGTGGCTCTAGACAGACTGTGGTGACCCAG XbaVλ4-9dF TCTGGCGGTGGCTCTAGACWGCCTGTGCTGACTCAG XbaVλ10dF TCTGGCGGTGGCTCTAGACAGGCAGGGCTGACTCAG NotJλ123R AGTCATTCACGACTTGCGGCCGCACCTAGGACGGTSASCTTGGTCCC NotJλ4-5R AGTCATTCACGACTTGCGGCCGCACCTAAAACGGTGAGCTGGGTCCC NotJλ7R AGTCATTCACGACTTGCGGCCGCACCGAGGACGGTCAGCTGGGTGCC Primer for colony PCR of VH and VK/Vλ ColLMB3F CAGGAAACAGCTATGAC Colfdseq1R GAA TTT TCT GTA TGA GGG Restriction enzyme sites are underlined 66 Nam-Kyung Lee et al.

17. Oligo-nucleotide primer set for colony PCR (see Table 1). 18. 14 ml round bottom culture tube (Corning, New York, USA).

2.3 VH Cassette 1. NcoI-HF (New England BioLabs). Cloning and scFv 2. NheI-HF (New England BioLabs). Library Generation

2.3.1 ScFv Library with naı¨ve Human VH Genes

0 2.3.2 ScFv-Shufﬂe 1. Primers: NcoU2scFvF (5 ATTCCATGGCCCAGGTG- 0 0 Library with a Given Human CAGCTGCAGGAG 3 ) and NheU2scFvR (5 CAGGC- 0 VH Gene TAGCTGAGGAGACGGTGACCAG 3 ).

2.4 scFv-Phage 1. M13KO7 helper phage (Thermo Fisher Scientific). Packaging and Library 2. 70 mg/ml Kanamycin (Sigma), filter-sterilized. Preparation 3. Polyethylene Glycol (PEG) 8000 (Thermo Fisher Scientific). 4. PEG/NaCl solution, 5Â (20% PEG8000 (w/v), 2.5 M NaCl). 5. Phosphate buffer saline (PBS): 137 mM NaCl, 2.68 mM KCl, 10 mM Na2HPO4, 2 mM KH2PO4, pH 7.4.

6. Sodium Azide (NaN3) (Sigma). 7. 2xYT-AK medium (2xYT, 100 μg/ml Ampicillin, 70 μg/ml Kanamycin).

2.5 Selection 1. pFUSE-hIgG1-Fc2 plasmid (InvivoGen, San Diego, USA). of Phage Antibody 2. Protein A agarose (Thermo Fisher Scientific). Display Library 3. EZ-Link™ Sulfo-NHS-Biotin (Thermo Fisher Scientific). on Recombinant ® Antigen 4. Dynabeads M-280 Streptavidin (Thermo Fisher Scientific). ® 5. Dynal Magnetic rack (Thermo Fisher Scientific). 2.5.1 ScFv Phage- 6. End-over-end rotator (Barnstead International, Dubuque, Display Selection USA). 7. PBSM: PBS, 2% nonfat dry milk (LabScientific, Highlands, USA). 8. PBSMT: PBS, 2% nonfat dry milk, 0.1% Tween20 (Acros, Geel, Belgium). 9. Triethylamine (TEA) (Sigma).

2.5.2 Screen by Phage 1. 96-well MaxiSorp™ ﬂat-bottom plate (Corning). ELISA 2. PBST: PBS, 0.1% Tween20. 3. Biotin-labeled rabbit anti-fd bacteriophage antibody (Sigma). 4. Streptavidin-HRP (Horseradish Peroxidase) (Sigma). 5. TMB substrate solution (Thermo Fisher Scientiﬁc). Modular Construction of scFv Phage Display Library 67

6. Hydrochloric acid (Thermo Fisher Scientiﬁc). 7. Plate reader (Synergy HT from Biotek, Winooski, USA).

2.5.3 Flow Cytometry 1. Flow cytometry buffer (FCB): PBS, 2% fetal bovine serum Analysis of Monoclonal (FBS) (Thermo Fisher Scientific). Phage 2. Streptavidin conjugated with phycoerithrin (PE) (Thermo Fisher Scientific). 3. HEK293 cell line (ATCC, Manassas, USA). 4. BD Accuri™ C6 flow cytometer (BD Biosciences, San Jose, USA). 5. OptiMEM I serum-free medium (Thermo Fisher Scientific). 6. TransIT 2020 transfection reagent (Mirus Bio, Madison, USA).

3 Methods

The protocols describe modular construction of large naive human scFv phage-display libraries from independent heavy and light chain gene cassettes. These cassettes are derived from libraries constructed independently from naive human heavy and light chain (Vκ and Vλ) gene repertoires by one-step PCR amplification using primer sets with two restriction enzyme site overhangs. First, the pHEN1 phagemid [29] is modified by inserting stuffer sequences, a(G4S)3 linker flanked by restriction enzyme sites for cloning to generate a new display vector pHEN1-NX (Fig. 1a). Next, primers matched with the N- and C-terminal sequences of VH, Vκ, and Vλ are used for the PCR amplification of each sub-family of antibody variable genes (Fig. 1b). The variable gene fragments are directly digested by two distinct restriction enzymes pairs, NcoI/NheI for VH and XbaI/NotI for VL (κ/λ), and ligated into the modified pHEN1-NX phagemid. Ligation products are desalted and concentrated using a centrifugal filter unit and electro-transformed into electro-competent TG1. In addition to modular construction of the scFv phage-display library, the separately constructed Vκ and Vλ libraries can be used to generate chain shuffled libraries anchored on a previously identified heavy chain for optimization and affinity maturation studies.

3.1 One-Step Human 1. Human PBMC and lymph node-derived cDNA samples Antibody Gene obtained from bio-sample preparation vendors can be conve- Ampliﬁcation niently used as an alternative source for antibody gene fragments (see Note 1). 2. Forward primers used for VH, Vκ, and Vλ PCR ampliﬁcation are described in Table 1. Mix equal molar concentration of each subfamily-based forward primer designated below: 68 Nam-Kyung Lee et al.

Fig. 1 Modified phagemid pHEN1-NX and antibody library generation scheme by one-step PCR and cut-and- paste cloning (a) The original pHEN1 phagemid was modified by introducing two stuffers with multiple stop codons, a flexible (G4S)3 linker, and two restriction enzyme sites (NheI and XbaI). (b) Different primer sets grouped by germline antibody sub-families are used to amplify VH (1, 2–7, 3, 5, 6), Vλ (1, 2, 3, 4–10), and Vκ (1, 2, 3, 4–6). Primer sets include flanking NcoI and NheI sites for VH or XbaI and NotI sites for Vλ and Vκ. Variable fragments amplified by one-step PCR were directly utilized for cloning

Antibody gene Primer set (ﬁnal concentration, 10 μM)

VH Set#1: NcoVH1aF/NcoVH1bF/NcoVH1cF/NcoVH1dF Set#2–7: NcoVH2aF/NcoVH2bF/NcoVH4aF/ NcoVH4bF/ NcoVH7aF Set#3: NcoVH3aF/NcoVH3bF/NcoVH3cF Set#5: NcoVH5aF Set#6: NcoVH6aF Vκ Set#1: XbaVK1aF/XbaVK1bF/XbaVK1cF/XbaVK1dF Set#2: XbaVK2aF/XbaVK2bF Set#3: XbaVK3aF/XbaVK3bF/XbaVK3cF Set#4–6: XbaVK4aF/XbaVK5aF/XbaVK6aF/XbaVK6bF Vλ Set#1: XbaVλ1aF/XbaVλ1bF/XbaVλ1cF Set#2: XbaVλ2F Set#3: XbaVλ3aF/XbaVλ3bF/XbaVλ3cF/XbaVλ3dF Set#4–10: XbaVλ4F/XbaVλ5dF/XbaVλ6dF/XbaVλ7dF/ XbaVλ8dF/XbaVλ4-9dF/XbaVλ10dF

3. Make each reverse primer mixture (ﬁnal concentration, 10 μM) for VH, Vκ, and Vλ by mixing equal molar concentration of each reverse primer (Table 1). Set up mixture for one-step PCR as described below (see Note 2): Modular Construction of scFv Phage Display Library 69

Component (concentration) Volume (μl/reaction)

ddH2O 22–20 OneTaq master mix (2Â)25 Forward primer mixture (10 μM) 1 Reverse primer mixture (10 μM) 1 cDNA template 1–3

4. Add cDNA templates and carry out the PCR for 30 cycles (30 s at 95 C, 1 min at 55 C, 1 min at 72 C) after preincubation for 3 min at 95 C, then complete the PCR by incubating the samples for 7 min at 72 C. Four or six separate PCR ampliﬁca- tions for Vκ/Vλ or VH subfamilies can be performed with reverse primer mixture and each forward primer set. 5. Analyze PCR products on 1.4% TAE agarose gel by electrophoresis. Cut out separately the ampliﬁed Vκ,Vλ (~350 bp), and VH (~380 bp) and purify them from agarose gel with a gel extraction kit (see Note 3). For each Vκ,Vλ, and VH subfamily, pool the extraction product together (see Note 4). 6. Determine the DNA concentration of pooled antibody genes. Store the pooled DNA at 4 C and use them for the following restriction enzyme digestion step.

3.2 Vκ and Vλ 1. Prepare pHEN1-NX phagemid using a plasmid mini prep kit. Library Construction Perform a single (XbaI or NotI-HF) and double digestion (XbaI/NotI-HF) to test whether the phagemid can be fully digested. 2. Digest 5 μg of pHEN1-NX phagemid with XbaI and NotI-HF (5 U/μg of DNA) in 1Â buffer provided by the manufacturer for 4 h at 37 C. Analyze the digestion product on 0.7% TAE agarose gel and purify the digested vector from agarose gel (see Note 5). Determine the DNA concentration and store the samples at 4 C. 3. Digest 1 μgofVκ or Vλ PCR product with XbaI and NotI-HF (20 U/μg of DNA) for 6 h at 37 C. Purify the digested Vκ or Vλ fragment by a PCR puriﬁcation kit (see Note 6) and determine the DNA concentration. 4. Perform ligation reaction with the digested pHEN1-NX and Vκ or Vλ as shown below (see Note 7): 70 Nam-Kyung Lee et al.

Component (concentration) Volume (μl/reaction)

ddH2O Add up to 100 T4 DNA ligase buffer (10Â)10 pHEN1-NX (40–50 ng/μl) 20–25 Vκ or Vλ (10–20 ng/μl) 15–30 T4 DNA ligase (400,000 U/ml) 1

5. Incubate the ligation mixture at 16 C for 16 h and inactivate the ligase at 65 C for 10 min. 6. Concentrate and desalt the ligation product by using a centrifugal ﬁlter with a 50 kDa molecular weight cutoff according to the manufacturer’s instruction (see Note 8). 7. Thaw one vial of 50 μl electrocompetent TG1 bacteria on ice for 10 min (see Note 9). Distribute 25 μl of TG1 into two prechilled 1.5 ml Eppendorf tubes and transfer 5 μl of concentrated and desalted Vκ or Vλ ligation product into each tube. Incubate for 5 min and transfer the mixture into two prechilled 0.1 cm electroporation cuvettes. Electroporate at the setting of 1.8 kV/600 ohms/10 μF and add immediately 950 μl recovery medium pre-warmed at 37 C(see Note 10). 8. Transfer the culture into a 14 ml round bottom tube and incubate at 37 C for 1 h. in a shaker-incubator at 250 rpm. 9. Spread 1:105 and 1:106 diluents of the culture to determine the titer of transformed bacteria and spread the remaining culture onto three large 2xYT-AG plates (150 mm Â 15 mm) per one electroporation reaction. Incubate the plates at 37 C overnight. 10. Calculate colony titer of each sublibrary and scrape transformed bacteria by using 2xYT-AG supplemented with 20% glycerol (~3 ml/large plate). Freeze collected transformants at À80 C as a sublibrary stock. 11. To check Vκ or Vλ gene insertion, randomly pick single colonies from each transformation and perform colony PCR by using a primer set (Table 1). Analyze correct PCR products (~800 bp) on TAE agarose gel and estimate VL-insertion rate based on the electrophoresis result (see Note 11). 12. Repeat digestion, ligation, and electro-transformation steps to achieve expected size of each Vκ or Vλ library (> ~108 cfu/ library) (see Note 12). After ﬁnalizing transformation reactions, thaw all of collected transformants and mix them together proportionally based on each sublibrary size. Aliquot into cryovials (~1 ml) and store each reconstituted Vκ or Vλ library at À80 C. Modular Construction of scFv Phage Display Library 71

3.3 VH Cassette 1. VH cassette library is separately generated using the same Cloning and scFv methods as described in Subheading 3.2 (see Note 13). Library Generation 2. Prepare VH- and VL (Vκ or Vλ)-library phagemid using a plasmid mini prep kit. Perform double digestion by using 3.3.1 ScFv Library NcoI-HF and NheI-HF to test digestion of the stuffer 1 for with naive Human VH VH cloning (see Note 5). Genes 3. Digest separately 5 μg of VH- and VL-library phagemid with NcoI-HF and NheI-HF (5 U/μg of DNA) in 1Â buffer provided by the manufacturer for 4 h at 37 C. Analyze the digestion described on 0.7% TAE agarose gel and purify the digested VH gene cassettes (~350 bp) and VL phagemid vector from agarose gel (see Note 5). 4. Determine the DNA concentration and store samples at 4 C. 5. Ligate the digested VL-library phagemid and VH fragments as shown below (see Note 7):

Component (concentration) Volume (μl/reaction)

ddH2O Add up to 100 T4 ligase buffer (10x) 10 VL-library phagemid (40–50 ng/μl) 20–25 VH gene cassette (12–16 ng/μl) 15–20 T4 ligase (400,000 U/ml) 1

6. Incubate the ligation reaction at 16 C for 16 h and inactivate the ligase at 65 C for 10 min. 7. Concentrate and desalt the ligation product by using a centrifugal ﬁlter with a 50 kDa molecular weight cutoff according to the manufacturer’s instruction (see Note 8). 8. Thaw one vial of 50 μl electrocompetent TG1 bacteria on ice for 10 min (see Note 9). Distribute 25 μl of bacteria cells into two prechilled 1.5 ml Eppendorf tubes and transfer 5 μlof concentrated and desalted VH ligation product into each tube. Incubate for 5 min and transfer the mixture into two prechilled 0.1 cm electroporation cuvettes. Electroporate at the setting of 1.8 kV/600 ohms/10 μF and add immediately 950 μl recovery medium pre-warmed at 37 C(see Note 10). 9. Transfer the culture into a 14 ml round-bottom tube and incubate at 37 C for 1 h. in a shaker-incubator at 250 rpm. 10. Spread 1:105 and 1:106 diluents of the culture to determine the titer of transformed bacteria and spread the rest onto three large 2xYT-AG plates (150 mm Â 15 mm) per one electroporation reaction. Incubate the plates at 37 C overnight. 72 Nam-Kyung Lee et al.

Fig. 2 Sequence analysis of VH-Vκ and VH-Vλ scFv phage libraries. (a) Germline subfamily distribution of light and heavy chain sequences in VH-Vκ and VH-Vλ scFv libraries. 182 Vκ (panel a, left), 170 Vλ (panel a, right), or 252 VH (panel b) sequences were obtained from each scFv library and analyzed for frequency of representation of light chain variable region subfamilies. It should be noted that the distribution of subfamilies can be adjusted by adjusting the relative amount of PCR product from each subfamily during VL and VH cassette generation. (c) 252 VH sequences derived from constructed scFv libraries were analyzed for length distribution of CDR-H3 regions

11. Calculate colony titer of each VH-inserted sublibrary as a scFv sublibrary. Scrape transformed bacteria by using 2xYT-AG supplemented with 20% glycerol (~3 ml/large plate). Freeze collected transformants at À80 C as a scFv sublibrary stock. 12. To check full-length scFv insertion, randomly pick single colonies from each transformation and perform colony PCR by using ColLMB3F (forward) and Colfdseq1R (reverse) primers (Table 1). Analyze correct PCR products (~950 bp) on TAE agarose gel and estimate the percent of scFv-insertion based on the electrophoresis result (see Note 11). 13. Sequence scFv from step 12 above and analyze VH, Vκ, and Vλ sequences by IgAT tool [30] to evaluate the distribution of antibody subfamilies and the sequences of the third complementarity determining region of the antibody heavy chain (CDR-H3). The subfamily representation of Vκ, and Vλ cassettes (Fig. 2a) or the VH cassette (Fig. 2b) is estimated according to the frequency of each subfamily. CDR-H3 sequences (6 ~ 25 amino acids) from 252 VH-Vκ and VH-Vλ scFvs are analyzed for CDR-H3 length distribution (Fig. 2c). 14. Repeat digestion, ligation, and electro-transformation steps to achieve expected size (at least ~109 cfu) of the ﬁnal scFv library. After ﬁnalizing transformation reactions, thaw all of collected transformants and mix them together proportionally based on each sublibrary size. Aliquot into cryovials (~1 ml) and store at À80 C. Modular Construction of scFv Phage Display Library 73

3.3.2 ScFv-Shuffle 1. Perform PCR to amplify the VH fragment of interest with Library with a Given Human flanking NcoI and NheI sites by using the primer set 0 VH Gene NcoU2scFvF (5 ATTCCATGGCCCAGGTGCAGCTGCAG- GAG 30) and NheU2scFvR (50 CAGGCTAGCTGAGGA- GACGGTGACCAG 30) under the following condition: 30 cycles (30 s at 95 C, 1 min at 55 C, 1 min at 72 C) after preincubation for 3 min at 95 C, with final extension for 7 min at 72 C. 2. Analyze PCR products (~380 bp) on 1.4% TAE agarose gel by electrophoresis. Gel-purify the amplified VH fragment with a gel extraction kit (see Note 3), and determine the DNA concentration. 3. Carry out the VH cloning into VL libraries (Vκ or Vλ)to generate a VL-shuffled library as described previously (see Sub- heading 3.3.1). Repeat digestion, ligation, and electro- transformation steps to achieve desired library size (> ~2 Â 108 cfu).

3.4 ScFv-Phage 1. Thaw out scFv-phagemid library TG1 stocks and transfer into Packaging and Library 500 ml 2xYT-AG medium (OD600 at 0.05 ~ 0.1). Inoculate Preparation the culture for 2 h in a shaker-incubator at 250 rpm until reaching exponential growth phase (OD600 at ~0.5). 2. Add M13KO7 helper phage at a MOI (multiplicity of infection) of 20 and incubate the culture at 37 C for 30 min without shaking, followed by 30 min shaking at 120 rpm (see Note 14). 3. Harvest the bacteria by centrifugation (3500 Â g, 10 min) and remove as much residual glucose as possible. 4. Resuspend the pellet with 500 ml 2xYT-AK medium in a 2 L shaker ﬂask. Incubate the culture overnight with proper aeration in a shaker-incubator at 250 rpm at 30 C to rescue scFv- phage particles. 5. Centrifuge the bacteria (4500 Â g, 15 min, 4 C) and transfer 120 ml of the supernatant into four 250 ml centrifuge bottles. Add 30 ml 5Â PEG/NaCl solution to each bottle and mix completely by vortexing. Precipitate the phage particles on ice for at least 1 h. 6. Centrifuge the precipitated phage suspension (4500 Â g, 15 min, 4 C) and remove as much supernatant as possible. Resuspend the phage pellet from each centrifuge bottle using 2 ml PBS, collect and mix the suspension in a 15 ml Falcon tube. 7. Centrifuge (4500 Â g, 15 min, 4 C) to remove remaining bacterial debris. Transfer the suspension into a new 15 ml 74 Nam-Kyung Lee et al.

Falcon tube and add 1/5 volume of 5Â PEG/NaCl solution, and precipitate the phage on ice for at least 2 h. 8. Centrifuge for 15 min at 4500 Â g at 4 C, and remove the supernatant. Resuspend the phage pellet in 3 ml sterile PBS/0.02% NaN3 (w/v). Keep phage stock at 4 C. 9. To determine phage library titer, dilute rescued phage in 2xYT (1:10 serial dilutions from 1:106 to 1:1010) and infect 180 μlof log phase TG1 with 20 μl of phage diluents. Incubate for 30 min at room temperature without shaking and an additional 30 min in a shaker-incubator at 250 rpm at 37 C, and plate the infected TG1 on 2xYT-AG plate. Incubate the plates at 37 C overnight. 10. Calculate the phage titer based on the number of colonies on the plate (see Note 15). Add glycerol to 20% into previously resuspended phage stock (see step 8 above) and freeze rescued phage library at À80 C for long-term storage and use.

3.5 Selection 1. Clone the cDNA of the target antigen of interest into pFUSE- of Phage Antibody hIgG1-Fc2 to produce a recombinant Fc-fusion molecule. Display Library Purify the fusion protein on a protein A column and biotin- on Recombinant label it using a biotinylation reagent (see Note 16). Antigen 2. Make 15 ml 4% PBSM for each round of the selection procedure. Mix an equal volume of phage library and 4% PBSM 3.5.1 ScFv Phage- (total 1 ml, ﬁnal 2% PBSM) in a 1.5 ml Eppendorf microcen- Display Library Selection trifuge tube per antigen selection. Rotate the tube end-over- end to mix for 5 min at RT. 3. Equilibrate 50 μl of streptavidin-coated Dynabeads in PBS and wash once with PBSM. Draw the beads into a pellet with a magnetic rack and remove PBSM from the tube. Add phage antibody library solution (see above Subheading 3.3.1) and resuspend the beads thoroughly. Incubate for 1 h at RT using an end-over-end rotator. 4. Draw the beads into a pellet with a magnetic rack and transfer the supernatant containing phage library depleted against the streptavidin beads into a new 1.5 ml Eppendorf tube. Add 0.5 ~ 10 nM biotinylated Fc-fusion antigen directly to the counter-selected phage library (see Note 17) and incubate on an end-to-end rotator for 1 h at RT. While the incubation is in progress, place 50 μl of streptavidin-coated Dynabeads in a new 1.5 ml tube. Equilibrate, wash, and block the beads in 2% PBSM for 1 h. 5. Draw blocked beads into a pellet with a magnetic rack and remove the supernatant. Resuspend the beads with 1 ml of the phage library incubated with the antigen. Incubate the phage/biotinylated-antigen/streptavidin-beads mixture for 30 min at RT on an end-to-end rotator. Modular Construction of scFv Phage Display Library 75

6. Draw the phage/antigen/beads complex into a pellet with a magnetic rack for 3 min. Discard carefully the supernatants, wash the beads ﬁve times with 1 ml PBSMT (see Note 18). After the ﬁnal washing, resuspend the beads with 1 ml PBS and transfer into a new 1.5 ml Eppendorf tube. Wash twice with PBS. 7. Elute bound phage from the beads with 0.5 ml of 100 mM TEA by incubation for 10 min on a rotator (see Note 19). Draw the beads and transfer the eluent to a new 1.5 ml tube containing 250 μl of 1 M Tris–HCl (pH 6.8). Immediately neutralize the solution by vortexing. 8. Add neutralized phage eluent to 8.5 ml of exponential phase TG1 (OD600 at 0.7) and incubate for 1 h in a shaker-incubator at 120 rpm at 37 C. Plate 1 and 10 μl of the infected TG1 on 2xYT-AG plates (100 mm Â 15 mm) to estimate the number of eluted phage. 9. Centrifuge the remaining TG1 culture at 3000 Â g for 10 min, remove the supernatant, resuspend the pellet using 300 μl 2xYT medium, and plate on a large 2xYT-AG plate (150 mm Â 15 mm). 10. Grow overnight at 37 C and collect the TG1 output by scraping using 2.5 ml 2xYT-AG/20% glycerol medium.

11. Inoculate the TG1 output (OD600 at 0.05 ~ 0.1) in 2xYT-AG medium and prepare phage particles for the next round of selection as described in Subheading 3.4 (see Note 20). Cul- ture the bacteria and rescue phage particles as described in Subheading 3.4.

3.5.2 Screen by Phage 1. Refer to Subheading 3.5.1 for polyclonal phage preparation ELISA from each round output of the selections. 2. Coat a 96-well microtiter plate with the Fc-fusion antigen (2 ~ 10 μg/ml in PBS) by incubating overnight at 4 C. Alternatively, the antigen can be immobilized by incubation for 2 h at room temperature. 3. Discard the antigen solution and block the plate with PBSM (200 μl/well) for 1.5 h at room temperature. 4. Wash the plate three times with PBS and add polyclonal phages (109 ~1010 pfu/well) resuspended in PBSM for 1 h 5. Discard the solution. Wash the plate three times with PBST followed by two times with PBS. 6. Add biotinylated anti-fd bacteriophage antibody (100 μl/well) diluted at 1:1000 in PBSM and incubate for 1 h at room temperature. 76 Nam-Kyung Lee et al.

7. Discard the primary antibody solution and wash wells three times with PBST followed by two times with PBS. 8. Add streptavidin-HRP (100 μl/well) diluted at 1:2000 in PBSM and incubate for 30 min at room temperature. 9. Discard the solution and wash wells three times with PBST followed by two times with PBS. 10. Add 100 μl of TMB substrate solution into each well and incubate the plate for 1 ~ 10 min at room temperature. 11. Conﬁrm the blue-colored reaction and quench by adding 100 μl/well of 1 N HCl solution.

12. Measure the absorbance at OD450 nm using a plate reader and compare each binding activity of polyclonal phages rescued from unselected library-, first-, second-, and third- round outputs. Increasing signals are expected with progressing rounds of selection. Output from the 3rd round of selection is often used for screening of monoclonal phage antibody (described below). 13. To screen for monoclonal binding phage from polyclonal phage selection output showing positive ELISA signals, individual colonies are separately inoculated into a 96-well plate containing 150 μl/well of 2xYT-AG medium. 14. After overnight incubation, store the original plate in À80 C after mixing with 50 μl of 60% glycerol. Each bacteria culture (~20 μl) from each well of the original plate is inoculated in a new plate containing 150 μl/well of 2xYT-AG medium and incubated for ~2 h at 37 C. 15. Infect the bacteria with helper phages (~2 Â 109 pfu/well) for 30 min at 37 C without shaking and an additional 30 min with shaking at 150 rpm. 16. Pellet the culture by centrifugation at 2000 Â g for 10 min at 4 C. Resuspend the culture with 150 μl/well of 2xYT-AK medium and incubate the plate overnight at 37 Cina shaker-incubator at 150 rpm. 17. Pellet the bacteria by centrifugation at 2000 Â g for 10 min at 4 C and transfer the supernatant into a new 96-well plate to store at 4 C until analysis. 18. Use 50 μl of each phage supernatant for ELISA screening using procedures described above in steps 2–12 above in this section. Positive clones are identified and subjected to flow cytometry analysis for binding to the target antigen expressed on the surface of a living cell. Modular Construction of scFv Phage Display Library 77

3.5.3 Flow Cytometry 1. Refer to steps 13–17 in Subheading 3.5.2 for monoclonal Analysis of Monoclonal phage production. Use 50 μl of each phage supernatant to Phage analyze binding by flow cytometry. 2. Prepare GFP- and antigen-expression plasmids for transient transfection into HEK293 cells. Seed HEK293 cells (~1 Â 106 cells/well) in a 6-well plate on the day before transfection. 3. Mix 0.5 μg of GFP-expression plasmid, 1.5 μg of antigen- expression plasmid, and 7.5 μl of transfection reagent thoroughly and let the mixture for 15 min at room temperature. 4. Remove the growth medium from HEK293 culture and transfect the cells with the prepared DNA mixture. 5. Incubate the cells for 16 ~ 24 h, trypsinize and collect the cells by centrifugation at 1000 Â g for 5 min. Resuspend cell pellet with FCB at ~1 Â 106 cells/ml. 6. Transfer 100 μl of cell suspension into a 96-well V-bottom plate. Add 50 μl of monoclonal phage supernatant and incubate for 1 h at room temperature with shaking. 7. Centrifuge the 96-well plate and wash once with PBS (200 μl/ well). Discard PBS and resuspend the cell pellet with 100 μl biotinylated anti-fd bacteriophage antibody (3.5 μg/ml). Incu- bate the plate for 1 h at room temperature. 8. Perform centrifugation and washing steps as above. Resuspend the cell pellet with 100 μl PE-conjugated streptavidin (2 μg/ ml). Incubate the plate for 30 min at room temperature. 9. After washing the cells, resuspend the pellet thoroughly with 200 μl PBS. Determine cell-binding by flow cytometer by gating the GFP positive population and analyze mean fluorescence intensity in appropriate channel for PE. The non-transfected parental HEK293 is used as the negative control.

4 Notes

1. Pooled cDNA from over 400 healthy donors was used in our library construction although there is no linear correlation between the pool size and the diversity of the eventual antibody library. 2. The amount of cDNA template required for optimal antibody gene PCR ampliﬁcation should be determined by direct testing. 3. Because the purity of the PCR products is critical for downstream procedures including restriction enzyme digestion and ligation, we routinely incorporate additional column washing 78 Nam-Kyung Lee et al.

steps into the gel extraction protocol. After spinning the solubilized gel pieces through the nucleic acid purification column, we perform a column wash step with a gel solubilization solution to remove any remaining gel fragments. We then wash the column twice with wash buffer, taking care to ensure that all residual wash buffer is removed from the column prior to elution. 4. Consider the desired final library size and prepare enough PCR product for subsequent steps. Approximately 0.5 μgof digested PCR product is needed for each ligation reaction and three ligation reactions are sufficient for 2–4 transformations. Although commercially available electro-competent TG1 is rated at 4x1010–1011 cfu/μg (presumably tested using purified plasmids), in our hands, the transformation efficiency of ligation products generally falls in the range of 0.5–2.0 Â 108 cfu/transformation. 5. The pHEN1-NX phagemid has two stuffer sequences (stuffer 1 and 2) for VH and VL cloning. The stuffer 1 and 2 are sized at 198 bp and 150 bp, respectively. Digestion with NcoI-HF/ NheI-HF or XbaI/NotI-HF releases the stuffer 1 or 2 from the vector. Cut out carefully the larger vector band from TAE agarose gel to avoid contamination of the stuffer fragments. Both stuffer fragments contain multiple stop codons. If either stuffer remains in the phagemid, no gene III fusion product will be made. 6. Use ddH2O pre-warmed at 70 C to efficiently elute the digested antibody gene fragments from the DNA purification column. Incubate the DNA column for 5 min at room temperature and spin for DNA elution. 7. High transformation efficiency by electroporation is facilitated by high DNA concentration in the concentrated and desalted ligation product (100–300 ng/μl). To generate enough ligation products for subsequent concentration and desalting steps, it is preferable to set up multiple small volume ligations (100 μl) rather than one large volume ligation. The ligation reactions can then be pooled and concentrated together. 8. Desalting and concentrating the ligation products are crucial for efficient electro-transformation. The purification of ligation products using this column-based method can improve the electro-transformation efficacy by desalting and concentrating at the same time. About 10–20 μl of concentrated ligation reaction should be isolated from this step. Check the DNA concentration to confirm that it is in the desired range (100–300 ng/μl). 9. Pre-made TG1 electro-competent cells have high electro- transformation efficiency (~4 Â 1010 cfu/μg as claimed by the Modular Construction of scFv Phage Display Library 79

manufacturer, approximately 0.5 ~ 2.0 Â 108 cfu/transformation in our hands, see Note 4). 50 μl of TG1 electrocompetent cells are suitable for two separate electroporations according to the manufacturer’s instruction. 10. The recovery medium provided with the electro-competent TG1 cells should be used for this step according to the manufacturer’s instruction. 11. To exclude sublibraries with poor VL- or VH-insertion rate, only include potential sublibraries qualified by colony PCR (more than 80% of full-length VL or VH inserts). To validate the scFv-phage library, analyze DNA sequences of single colonies randomly picked from each sublibrary by using the ColLMB3F (forward) and Colfdseq1R (reverse) primers. Esti- mate the final complexity of the scFv-phage library by aligning all sequenced clones with germline sub-family sequences in antibody gene databases (e.g., VBASE2). 12. Theoretical diversity in variable heavy or light chains is determined by combinatorial rearrangements, junctional flexibility, addition of P-(palindromic) and N-(non-templated) nucleotides, and somatic hypermutation. Experimental data showed the estimated diversity of VL segments may be about 106–107 in a scFv library (31). Thus Vκ or Vλ libraries that possess more than 108 diverse clones would be sufficient for scFv-library generation and affinity engineering by light chain-shuffling. 13. All materials used for VH gene cassette library generation are the same as listed in Subheading 2.2, except NcoI-HF and NheI-HF restriction enzymes. Because the quality of antibody library is mainly determined by the diversity of VH genes rather than that of VL genes, sufficient VH library size should be achieved for the high diversity of final libraries. Thus, allowing for further utilization of VH library (e.g., heavy chain shuffling, common light chain library generation), the size of VH cassette library should reach at least 1 ~ 2 Â 109 diversity. 8 14. An OD600 at 0.5 is equivalent to about 4.0 Â 10 cells/ml. So the number of M13KO7 helper phage that should be added into the bacteria culture can be estimated as follows: 4.0 Â 108 (cells/ml) Â 500 (ml) Â 20 (MOI) ¼ 4.0 Â 1012 pfu. The M13KO7 from the manufacturer (Thermo Fisher Scientific) has a titer of 1 Â 1011 pfu/ml. We routinely prepare a higher titer M13KO7 stock from infected TG1 to make the concentration at 1012 pfu/ml. 15. The phage titer rescued from 500 ml culture is generally expected at about 1012 cfu/ml or higher. 16. pFUSE-hIgG1-Fc2 plasmid encoding cDNA of target antigen can be transiently transfected into cell lines suitable for protein production, such as HEK-293A or CHO cell lines. Transient 80 Nam-Kyung Lee et al.

transfection is commonly performed using polyethylenimine (PEI), a stable cationic polymer. Transfected cells are maintained in serum-free media, and the media can be collected twice over a period of 1 ~ 2 weeks. Produced Fc-fusions are purified using a gravity column packed with Protein A beads and buffer-exchanged to PBS. EZ-Link™ Sulfo-NHS-Biotin was used to biotin-label purified Fc-fusion followed by neutralization with 1 M Tris-HCl and re-purification by gel filtration. 17. To control stringency during the selection rounds, the concentration of biotinylated Fc-fusion target antigen can be decreased as the selection proceeds (e.g., 5 nM for first round, 1 nM for second round, etc.). 18. To remove weak binders, washing steps with PBSMT can be increased in later selection rounds (e.g., 5 washes in the first round, 10 washes in the second round, etc.). 19. Exposure to the high pH TEA solution during elution can negatively affect the infectivity of eluted phage. Do not exceed 10 min during the elution step. 20. Total output of TG1 infected with phage eluent is generally between 105 ~107 cfu. It is generally recommended to perform a next round of selection with several hundred copies of each phage clone. Therefore, using a smaller volume (~ 50 ml) to inoculate the output bacteria from previous round is sufficient to amplify and prepare phages for the next round.

Acknowledgment

Work in our laboratory is supported by grants from the National Institutes of Health/National Cancer Institute (R01 CA171315, R01 CA118919, and R01 CA129491). NKL received fellowship support from Basic Science Research Program of the National Research Foundation of Korea (NRF) that is funded by the Ministry of Education, Science and Technology (2013R1A6A3A03060495).

References

1. Nelson AL (2010) Antibody fragments: hope 4. Clackson T, Hoogenboom HR, Grifﬁths AD and hype. MAbs 2:77–83 et al (1991) Making antibody fragments using 2. Chan CEZ, Lim APC, MacAry PA et al (2014) phage display libraries. Nature 352:624–628. The role of phage display in therapeutic anti- https://doi.org/10.1038/352624a0 body discovery. Int Immunol 26:649–657. 5. de Haard HJ, van Neer N, Reurs A, Hufton SE, https://doi.org/10.1093/intimm/dxu082 Roovers RC, Henderikx P, de Bruı¨ne AP, 3. Marks JD, Hoogenboom HR, Bonnert TP et al Arends JW, Hoogenboom HR (1999) A large (1991) By-passing immunization. Human non-immunized human fab fragment phage antibodies from V-gene libraries displayed on library that permits rapid isolation and kinetic phage. J Mol Biol 222:581–597 analysis of high afﬁnity antibodies. J Biol Chem 274:18218–18230 Modular Construction of scFv Phage Display Library 81

6. Barbas CF 3rd, Kang AS, Lerner RA et al selection of single-chain Fv antibody libraries (1991) Assembly of combinatorial antibody on prostate cancer cells. Cancer Res libraries on phage surfaces: the gene III site. 64:704–710 Proc Natl Acad Sci U S A 88:7978–7982 17. Sheets MD, Amersdorfer P, Finnern R et al 7. Hawkins RE, Russell SJ, Winter G (1992) (1998) Efficient construction of a large nonim- Selection of phage antibodies by binding affin- mune phage antibody library: the production ity. Mimicking affinity maturation. J Mol Biol of high-affinity human single-chain antibodies 226:889–896 to protein antigens. Proc Natl Acad Sci U S A 8. Reiter Y, Schuck P, Boyd LF et al (1999) An 95:6157–6162 antibody single-domain phage display library 18. Ruan W, Sassoon A, An F et al (2006) Identifi- of a native heavy chain variable region: isolation cation of clinically significant tumor antigens of functional single-domain VH molecules by selecting phage antibody library on tumor with a unique interface. J Mol Biol cells in situ using laser capture microdissection. 290:685–698. https://doi.org/10.1006/ Mol Cell Proteomics 5:2364–2373. https:// jmbi.1999.2923 doi.org/10.1074/mcp.M600246-MCP200 9. Lorimer IA, Keppler-Hafkemeyer A, Beers RA 19. An F, Drummond DC, Wilson S et al (2008) et al (1996) Recombinant immunotoxins spe- Targeted drug delivery to mesothelioma cells cific for a mutant epidermal growth factor using functionally selected internalizing human receptor: targeting with a single chain antibody single-chain antibodies. Mol Cancer Ther variable domain isolated by phage display. Proc 7:569–578. https://doi.org/10.1158/1535- Natl Acad Sci U S A 93:14815–14820 7163.MCT-07-2132 10. Dooley H, Flajnik MF, Porter AJ (2003) Selec- 20. Zhu X, Bidlingmaier S, Hashizume R et al tion and characterization of naturally occurring (2010) Identification of internalizing human single-domain (IgNAR) antibody fragments single-chain antibodies targeting brain tumor from immunized sharks by phage display. Mol sphere cells. Mol Cancer Ther 9:2131–2141 Immunol 40:25–33 21. Venet S, Kosco-Vilbois M, Fischer N (2013) 11. Hairul Bahara NH, Chin ST, Choong YS et al Comparing CDRH3 diversity captured from (2016) Construction of a semisynthetic human secondary lymphoid organs for the generation VH single-domain antibody library and selec- of recombinant human antibodies. MAbs tion of domain antibodies against alpha- 5:690–698. https://doi.org/10.4161/mabs. crystalline of mycobacterium tuberculosis. J 25592 Biomol Screen 21:35–43. https://doi.org/ 22. Yip YL, Hawkins NJ, Clark MA et al (1997) 10.1177/1087057115609144 Evaluation of different lymphoid tissue sources 12. Sanchez-Martin D, Sorensen MD, Lykkemark for the construction of human immunoglobu- S et al (2015) Selection strategies for anticancer lin gene libraries. Immunotechnology antibody discovery: searching off the beaten 3:195–203 path. Trends Biotechnol 33:292–301. 23. Vaughan TJ, Williams AJ, Pritchard K et al https://doi.org/10.1016/j.tibtech.2015.02. (1996) Human antibodies with 008 sub-nanomolar affinities isolated from a large 13. Schwimmer LJ, Huang B, Giang H et al (2013) non-immunized phage display library. Nat Bio- Discovery of diverse and functional antibodies technol 14:309–314. https://doi.org/10. from large human repertoire antibody libraries. 1038/nbt0396-309 J Immunol Methods 391:60–71. https://doi. 24. Lennard S (2002) Standard protocols for the org/10.1016/j.jim.2013.02.010 construction of scFv libraries. Methods Mol 14. Liu B, Marks JD (2000) Applying phage anti- Biol 178:59–71 bodies to proteomics: selecting single chain Fv 25. Hust M, Dubel S (2004) Mating antibody antibodies to antigens blotted on nitrocellu- phage display with proteomics. Trends Bio- lose. Anal Biochem 286:119–128. https:// technol 22:8–14. https://doi.org/10.1016/j. doi.org/10.1006/abio.2000.4788 tibtech.2003.10.011 15. Liu B, Huang L, Sihlbom C et al (2002) 26. Welschof M, Terness P, Kipriyanov SM et al Towards proteome-wide production of mono- (1997) The antigen-binding domain of a clonal antibody by phage display. J Mol Biol human IgG-anti-F(ab’)2 autoantibody. Proc 315:1063–1073. https://doi.org/10.1006/ Natl Acad Sci U S A 94:1902–1907 jmbi.2001.5276 27. Hust M, Frenzel A, Meyer T et al (2012) Con- 16. Liu B, Conrad F, Cooperberg MR et al (2004) struction of human naive antibody gene Mapping tumor epitope space by direct libraries. Methods Mol Biol 907:85–107. 82 Nam-Kyung Lee et al.

https://doi.org/10.1007/978-1-61779-974- 30. Rogosch T, Kerzel S, Hoi KH et al (2012) 7_5 Immunoglobulin analysis tool: a novel tool 28. Kugler J, Wilke S, Meier D et al (2015) Gener- for the analysis of human and mouse heavy ation and analysis of the improved human and light chain transcripts. Front Immunol HAL9/10 antibody phage display libraries. 3:176. https://doi.org/10.3389/fimmu. BMC Biotechnol 15:10. https://doi.org/10. 2012.00176 1186/s12896-015-0125-0 31. Glanville J, Zhai W, Berka J et al (2009) Precise 29. Hoogenboom HR, Griffiths AD, Johnson KS determination of the diversity of a combinato- et al (1991) Multi-subunit proteins on the sur- rial antibody library gives insight into the face of filamentous phage: methodologies for human immunoglobulin repertoire. Proc Natl displaying antibody (fab) heavy and light Acad Sci U S A. 106: 20216–20221. doi:10. chains. Nucleic Acids Res 19:4133–4137 1073/pnas.0909775106 Chapter 5

Construction of Macaque Immune-Libraries

Arnaud Avril, Sebastian Miethe, Michael Hust, and Thibaut Pelat

Abstract

Rapidly after the clinical success of the first murine therapeutic antibody licensed in 1985 (muromomab- CD3), the first limits of the therapeutic use of antibodies deriving from hybridoma technology appeared. Indeed, the nonhuman nature of these therapeutic antibodies makes them immunogenic when administrated to patients, which develop anti-drug antibodies (ADA). If repeated drug-administrations are needed, the immune response will accelerate the elimination of the drug, leading to a therapeutic failure, or in the worst case to an anaphylactic reaction against the murine protein. Several antibody generations were then developed to obtain better-tolerated molecules: chimeric, humanized, and fully human antibodies. The first antibody generation is fully based on cellular technology (mice hybridoma technology), but the next generations are improved by molecular engineering. Immune antibody phage-display libraries are one successful approach to isolating such engineered antibodies. One strategy to isolate high-affinity and well-tolerated antibodies when no immunized patients are available is based on the phage-display- screening of immune libraries deriving from immunized nonhuman primates, which are phylogenetically close to humans. This chapter presents the strategy for the construction of macaque antibody immune- libraries.

Key words Phage-display, Antibody, Non-human-primate, Macaque, Screening, Panning, Antibody library, Phagemid, Phages, scFv, Antibody fragment

1 Introduction

1.1 Choice The first monoclonal antibodies developed for therapy were of the Nonhuman isolated by the mice hybridoma-technology. [1] Unfortunately, Primate Approach the nonhuman nature of these antibodies quickly limits their therapeutic efficacy, excepted in some specific cases, such as with Muromomab-CD3, used to prevent or treat acute rejection of organ transplants, and which induced its own tolerance by inducing a complete depletion of T cell population. [2] To overcome this limitation, different antibody generations were developed to decrease the proportion of nonhuman amino acids in the therapeutic antibodies. First, the chimeric antibodies were obtained by grafting the nonhuman antibody variable-domains on human constant domains (~33% of nonhuman amino acids in the final

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_5, © Springer Science+Business Media LLC 2018 83 84 Arnaud Avril et al.

antibody), and then humanized antibodies were developed by grafting only the amino acids of the complementary determining regions (CDR) onto the frameworks of the variable light (VL) and variable heavy (VH) domains of human antibodies (~10% of nonhuman amino acids in the final antibody). [3] Despite these innovations, several antibodies were still immunogenic, leading to the development of the last generation, the fully human antibodies. [4] Fully human antibodies are mainly isolated by two technologies: transgenic-mice or rats (“humanized-mice”) [5, 6] and antibody phage display [7]. Despite the recent progress done in the development of transgenic mice, which now express a huge part of the human antibody repertoire and which allow natural (but murine) V-(D)-J gene-rearrangement, the generation and phage-display screening of antibody gene-libraries remain the technology of reference. Phage-display is the most widespread method for the display and selection of large collections of antibodies. [8, 9] Naive- libraries, which are composed of rearranged V-genes from B cells of non-immunized donors (i.e., the IgM repertoire) [10], are frequently opposed to immune-libraries as high-affinities are generally not reached without affinity-maturation engineering process. [10, 11] Inversely, human immune-libraries are generated after the active and controlled immunization of humans or from convales- cent patients. [12–14] Well-tolerated high-affinity antibodies are generally isolated from such immune-libraries as they have under- gone the natural affinity maturation. Immune-libraries generated from blood-derived human B-lymphocytes are the most suitable option to obtain efficient and well-tolerated antibodies. Neverthe- less, it is sometimes impossible to actively immunized humans due to ethical or legal aspects. When human-immunization is not possible, it was suggested to immunize nonhuman primates (NHP), because the human-like nature of their antibodies was revealed early by several studies [15, 16] and confirmed recently [17–19]: nonhuman primate antibodies are close but nonidentical to human ones. In Europe the utilization of chimpanzee for the isolation of antibodies is not authorized, and in the United States of America it is rigorously restricted. Contrary to chimpanzees, macaques are not listed on appendix I of the Convention for International Trade of Endangered Species (CITES) (excepted Macaca silenus), and are easily accessible. It was demonstrated that chimpanzee’s antibodies are not closer to human’s antibodies than macaque’s antibodies. [18] Consequently, macaques represent a model of choice for the generation of immune-libraries and this strategy already proved its ® success. Four “primatized antibodies” (i.e., antibodies where the nonhuman primate variable domains were fused on human constant domains) are currently, or were, in clinical development: Galiximab (anti-human CD80) is in phase II against non-hodgkin’s lymphoma and psoriasis, Lumiliximab (anti-human FCER2) is in Construction of Macaque Immune-Libraries 85

phase II/III against allergic asthma and chronic lymphocytic leukemia, Clenoliximab (anti-human CD4) was discontinued after being in phase II against asthma and Keliximab (anti-human CD4) was discontinued after being in phase III against rheumatoid arthritis and asthma, after proving to be well tolerated by the patients, [20–23] and many others were isolated. [18, 24–28] Another advantage of nonhuman primates is that nanomolar, or even picomolar antibodies [ 26], against any region of the antigen could be isolated, when an immune-library with an adequate size and diversity (108 different clones) was generated [18]. The following protocols describe the generation of single-chain Fragment variable (scFv) immune-libraries by a “two-step cloning strategy” which already proved its success for the development of naive [29] and immune libraries [24, 25, 30]. In this strategy, the ampliﬁed repertoire of light chain genes is cloned ﬁrst into the phage-display vector (phagemid vector), followed by the second cloning step the heavy chain gene repertoire into the phagemids containing the light chain gene repertoire [29]. An overview of antibody gene libraries and vectors is given in various reviews [29, 31–33].

1.2 Legal Aspects IDEC Biogen Inc. initially patented in Europe (patents EP0605442A1, EP1266965B1, EP1715045A2) and in the United States of America (patent 5,658,570) the strategy consisting in the development of chimeric antibodies obtained by grafting the variables domains of a nonhuman primate antibody on the ® variables domains of a human antibody (“Primatized antibodies”). These patents were based on the paradoxical and question- able notion that macaques are sufﬁciently distant from humans to raised antibodies against human proteins (even against very conserved proteins such as CD4 or CD54) and in the same time sufﬁciently similar to humans to have antibodies which do not trigger the host anti-antibody immune response when administrated in humans. According to this notion, antibodies against human proteins can be isolated from immunized macaques; moreover in Europe the macaque’s antibodies against human proteins were protected by the patents. This notion was based on several old studies comparing human and nonhuman primate’s antibodies sequences and analyzing the (absence of) macaque’s immune response when human IgG1 were injected to them. [34–36] These results are consistent with the clinical studies of three “Pri- ® matized antibodies,” where anti-drug antibodies were not detected in patients’ sera, [37, 38] but are contradicting recent studies demonstrating that human and nonhuman primate’s antibody sequences are statistically different, justifying the humanization of nonhuman primate’s antibodies. [18, 19, 39] Based on these studies, the validity of EP1266965B1 was challenged and invalidated on 20/08/2010. In addition, the parental patent 86 Arnaud Avril et al.

EP0605442A1 expired and EP1715045 was withdrawn on 20/08/2010, leading to a formal expiry of these patents in 25/07/2012. [40] Following the expiration, the utilization of ® “Primatized antibodies” in Europe is no longer regulated by ® these patents and it is now possible to develop “Primatized antibodies” in Europe. In the United States of America, the patent 5,658,570 is still in force, but it will expire in 2017.

2 Materials

2.1 Immunization 1. A specimen of the Macaca fascicularis species around 4–6 kg of Macaques can be bought from an approved supplier, under the approval of a relevant ethical committee. A special attention should be paid to the quality of breeding, because healthy animals are required; animals should be strictly selected according to blood count and clinical observation. Authorization of importation is restricted by the Convention for International Trade of Endangered Species (CITES, https://cites.org/). A quarantine is required before importation and the animal must be controlled for the absence of tuberculosis and herpes) and in the animal house facilities; 2. At least 0.5 mg of pure antigen in the form of soluble protein, 1011 inactivated virions or a relevant quantity of life virus. The quality of the immunogen must be controlled to be sure that the epitopes are in their native conformation. 3. Freund’s complete (Sigma-Aldrich, Saint-Quentin Fallavier, France) and incomplete (Sigma-Aldrich) liquid adjuvant. Alum adjuvant (such as Imject™ Alum Adjuvant, Thermo- Fisher Scientiﬁc, Courtaboeuf, France) may be used as an alternative. 4. Ketamine chlorhydrate, Tileamine, or similar for the anesthesia (bought from a provider approved by the relevant authorities). 5. Buprenorphine or similar (bought from a provider approved by the relevant authorities) if administration of an analgesic is required. 6. Sterile anticoagulant citrate dextrose (ACD) (Terumo BCT, Lakewood, USA). 7. 25 mm 17G Mallarme Trocar (Thiebaud Biomedical devices, Margencel, France).

2.2 Determination 1. DeepWell™ plate ﬂat-bottom. ® of the Serum-Titer 2. 96-well NUNC MaxiSorp ﬂat-bottom. ® 3. 0.1% Tween 20. 4. Plate washer, e.g., Well Wash 4 MK 2. Construction of Macaque Immune-Libraries 87

5. Anti-monkey IgG whole-molecule (Sigma-Aldrich). 6. TMB and TMB stop solution (e.g., HCl 1 N). 7. Multiskan FC microplate-spectrophotometer (ThermoFisher Scientiﬁc).

2.3 Total RNA 1. Sonicator Sonica Q700 (QSonica, Newton, USA) with 1/200 Isolation diameter probe for 50 mL plastic vials, or similar system. 2. Spectrophotometer (NanoDrop ND-1000, NanoDrop-Ther- moFisher, Wilmington, USA). 3. RNAse AWAY™ (Molecular Bioproducts, San Diego, Califor- nia, USA). ® 4. TRI Reagent (Molecular Research Center, Cincinnati, USA). 5. BCP (1–bromo–3–chloropropane) (see Note 1). 6. Isopropanol molecular-grade. 7. PCR-grade water. 8. Ethanol 75% v/v diluted in PCR-grade water. 9. PCR grade water. 10. Glycogen. 11. Phosphate-buffered saline (PBS) pH 7.4: 8.0 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4*2H2O, 0.24 g KH2PO4 in 1 L.

2.4 cDNA Synthesis 1. SuperScript IV First-Strand kit (Thermo Fisher Scientiﬁc) þ 5Â RT buffer þ0.1 m DTT. 2. RNAseOUT (Thermo Fisher Scientiﬁc). 3. dNTP mix (10 mM each).

® 2.5 First Antibody 1. Platinium Taq polymerase High Fidelity (Thermo Fisher Gene PCR Scientiﬁc). 2. Nuclease-free water.

3. 50 mM MgCl2. 4. dNTP mix. 5. Oligonucleotide primers (see Table 1). 6. Agarose molecular-biology grade. 7. TAE-buffer 10Â (40 mM Tris–Acetate, 1 mM EDTA, 1 mM EDTA pH 8.3). 8. Ethidium Bromide. 9. DNA size marker (e.g., Smart Ladder MW1700–10, Eurogen- tec, Liege, Belgium). 10. NucleoSpin Gel and PCR Clean-up (Macherey-Nagel, Duren,€ Germany). 88 Arnaud Avril et al.

Table 1 List of the forward and reverse primers useful for the amplification of VH, VLκ, and VLλ DNA-coding regions from CDNA

Primers for VH ampliﬁcation (First PCR)

Name 50- > 30 sequence (with XhoI or SpeI restriction sites)

MacVH1 Cag GTG cag CTC gag cag TCT GGG MacVH2 Cag GTG cag CTC CTC gag TCT GGG MacVH3 Cag GTG cag CTA CTC gag TCG GG MacVH4 Gag GTG cag CTC gag gag TCT GGG MacVH5 Gag GTG cag CTG CTC gag TCT GGG MacVH6 Cag GTA cag CTC gag cag TCA GG MacVH7 AGG TGC AGC TGC TCG AGT CTG G MacVH8 Cag GTG cag CTG CTC gag TCG GG MacVH9 Cag GTG cag CTA CTC gag TGG GG MacCH AGG TTT act AGT ACC ACC ACA TGT TTT gat CTC

Primers for VLκ ampliﬁcation (First PCR)

Name 50- > 30 sequence (with SacI or XbaI restriction sites)

MacVLκ1 GAC ATC gag CTC ACC cag TCT CCA MacVLκ2 GAC ATC gag CTC ACC cag TCT cc MacVLκ3 Gat ATT gag CTC act cag TCT CCA MacVLκ4 GAA ATT gag CTC ACG cag TCT CCA MacVLκ5 GAA ATT gag CTC ACA cag TCT CCA MacVLκ6 GAA ATT gag CTC ACA cag TCT CCA gag CCG CAC gag ccc gag CTC cag ATG ACC cag TCT cc MacVLκ7 Gag CCG CAC gag ccc gag CTC GTG ATG ACA cag TCT cc MacVLκ GCG CCG TCT Aga ATT AAC act CTC ccc TGT TGA AGC TCT TTG TGA CGG GCG AAC TCAG (continued) Construction of Macaque Immune-Libraries 89

Table 1 (continued)

Primers for VLλ ampliﬁcation (First PCR)

Name 50- > 30 sequence

MacVLλ1 Cag TCT GTG CTG act cag CCA cc MacVLλ2 Cag TCT GTG YTG ACG cag CCG cc MacVLλ3 Cag TCT GCC CTG act cag CCT MacVLλ4 TCC tat GWG CTG ACW cag CCA cc C MacVLλ5 TCT TCT cag CTG act cag GAC cc MacVLλ6 CTG CCT GTG CTG act CAA TCR YC MacVLλ7 Cag CYT GTG CTG act CAA TCR YC MacVLλ8 Cag SCT GTG CTG act cag cc MacVLλ9 AAT TTT ATG CTG act cag ccc CA MacVLλ10 Cag RCT GTG GTG ACY cag gag cc MacVLλ11 Cag SCW GKG CTG act cag CCA cc MacCLλ TGA ACA TTC TGT AGG GGC CAC TG Restrictions sites are shown in bold. DNA IUB codes: Y ¼ CorT,W¼ AorT,R¼ AorG,S¼ G or C and K ¼ Gor T. For degenerated primers, order equimolar concentration of each primer

2.6 Second Antibody 1. GoTaq þ5Â buffer (Promega, Mannheim, Germany). Gene PCR 2. Nuclease-free water.

3. 50 mM MgCl2. 4. dNTP mix. 5. Oligonucleotide primers (see Table 2). 6. Low-melting point agarose molecular-biology grade. 7. TAE-buffer 10Â (40 mM Tris-Acetate, 1 mM EDTA, 1 mM EDTA pH 8.3). 8. Ethidium Bromide. 9. DNA size marker (e.g., Smart Ladder MW1700–10, Eurogen- tec, Liege, Belgium). 10. NucleoSpin Gel and PCR Clean-up (Macherey-Nagel, Duren,€ Germany).

2.7 Cloning of VL 1. MluI-HF(NEB, Frankfurt, Germany). then VH 2. NotI-HF(NEB). 3. SﬁI (NEB). 90 Arnaud Avril et al.

Table 2 Primer sequences for the second PCR that add restriction sites

Primers for VH ampliﬁcation (Second PCR)

Name 50- > 30 sequence (with NcoI or HindIII restriction sites)

MHMacVH-NcoI_f1 GTCCTCGCACCATGGCCSAGGTGCAGCTCGAGSAGTCTGGG MHMacVH-NcoI_f2 GTCCTCGCACCATGGCCCAGGTGCAGCTRCTCGAGTCKGG MHMacVH-NcoI_f3 GTCCTCGCACCATGGCCSAGGTGCAGCTGCTCGAGTCKGG MHMacVH-NcoI_f4 GTCCTCGCACCATGGCCCAGGTACAGCTCGAGCAGTCAGG MHMacVH-NcoI_f5 GTCCTCGCACCATGGCCAGGTGCAGCTGCTCGAGTCTGG MHMacVH-NcoI_f6 GTCCTCGCACCATGGCCCAGGTGCAGCTACTRGAGTSGGG MHMacIgGCH1scFv- GTCCTCGCAAAGCTTTGGGCCCTTGGTGGA HindIII_r

Primers for VLκ ampliﬁcation (Second PCR)

Name 50- > 30 sequence (with MluI or NcoI restriction sites)

MHMacVK-MluI_f1 ACCGCCTCCACGCGTAGAHATCGAGCTCCANCAGTCTCC MHMacVK-MluI_f6 ACCGCCTCCACGCGTAGAGCTWCAGATGACMCAGTCTCC MHMacKappaCL-NotI_r ACCGCCTCCGCGGCCGCGACAGATGGTGSAGCCAC

Primers for VLλ ampliﬁcation

Name 50- > 30 sequence (with MluI or NotI restriction sites)

MHMacVL-MluI_f1 ACCGCCTCCACGCGTACAGTCTGTGCTGACTCAGCCRCC MHMacVL-MluI_f2 ACCGCCTCCACGCGTACAGTCTGCCCTGACTCAGCCT MHMacVL-MluI_f3 ACCGCCTCCACGCGTATCCTATGAGCTGACWCAGCCACC MHMacVL-MluI_f4 ACCGCCTCCACGCGTATCTTCTGAGCTGACTCAGGACCC MHMacVL-MluI_f5 ACCGCCTCCACGCGTACWGCCTGTGCTGACTCAGCC MHMacVL-MluI_f6 ACCGCCTCCACGCGTACAGCCGGCCTCCCTCTCAGCATCT MHMacVL-MluI_f7 ACCGCCTCCACGCGTACAGRCTGTGGTGACYCAGGAGCC MHMacVL-MluI_f8 ACCGCCTCCACGCGTACAGCCTGTGCTGACTCAGCCA MHMacLambdaCL-NotI_r ACCGCCTCCGCGGCCGCAGAGGAGGGCGGRAAWAGAGTGAC The restriction sites are shown in bold. DNA IUB codes: Y ¼ CorT,W¼ AorT,R¼ AorG,S¼ GorC,K¼ GorT, B ¼ not A (G or C or T), D ¼ not C (A or G or T), H ¼ not G (A or C or T) and V ¼ not T/U (A or C or G). For degenerated primers, order equimolar concentration of each primer

4. HindIII-HF (NEB). 5. Glycerol molecular-biology grade. 6. BSA 100Â molecular-biology grade. Construction of Macaque Immune-Libraries 91

7. Calf intestine phosphatase (CIP) (NEB). 8. T4 DNA ligase (Promega) 3 M sodium acetate pH 5.2 (Ther- moFisher Scientiﬁc). 9. E. coli XL1-Blue MRF’ genotype: Δ(mcrA)183 Δ(mcrCB- hsdSMR-mrr)173 endA1 supE44 thi-1 recA1 gyrA96 relA1 lac [F´ proAB lacIqZΔM15 Tn10 (Tetr)] (Agilent). 10. Electroporator MicroPulser (BIO-RAD, Munchen,€ Germany). 11. 2 M glucose (sterile ﬁltered).

12. 2 M Magnesium solution: 1 M MgCl2, 1 M MgSO4 (autoclaved). 13. SOC medium pH 7.0: 2% (w/v) tryptone, 0.5% (w/v) yeast extract, 10 mM NaCl 2.5 mM KCL, 10 mM MgSO4,10mM MgCl2 and 20 mM glucose. 14. 2ÂTY-media pH 7,0: 1.6% (w/v) tryptone, 1% (w/v) yeast extract, 0.5% (w/v) NaCl. À 15. Ampicilline 100 mg.mL 1 stock (Sigma Aldrich). À 16. Tetracycline 10 mg.mL 1 stock (Sigma Aldrich). 17. 2ÂYT medium: 1.6% (w/v) tryptone, 1% (w/v) yeast extract, 0.5% (w/v) NaCl. À 18. 2ÂYT-GAT: 2ÂTY þ 100 mM glucose þ100 μg.mL 1 ampi- À cilline þ20 μg.mL 1 tetracycline. 19. 9.4 cm Petri dishes. 20. 144 cm square Petri dishes. 21. 2ÂYT-GAT agar plates: 2ÂTY-GAT þ1.5% (w/v) agarose. 22. Nucleobond Extra Midi Kit (Macherey-Nagel).

2.8 Colony PCR 1. Oligonucleotide primers (see Table 3). 2. GoTaq þ5Â buffer (Promega).

2.9 Library 1. 2ÂYT-media pH 7.0. Packaging and scFv 2. 2ÂYT-GAT. Phage Production 3. M13 K07 helperphage for monovalent display (Agilent, Santa Clara, CA, USA).

Table 3 Sequences of the primers for single-colony PCR

Speciﬁcity Name 50- > 30 sequence

VH MHLacZPro_f 10 μM GGC TCG TAT GTT GTG TGG VLκ MHgIII_r10 μM CTA AAG TTT TGT CGT CTT TCC 92 Arnaud Avril et al.

4. Hyperphage M13KO7ΔpIII for oligovalent display (Progen, Heidelberg, Germany). 5. Kanamycin B sulfate (Sigma-Aldrich). À À 6. 2ÂYT-AK: 2ÂYT þ100 μg.mL 1 ampicillin þ50 μg.mL 1 kanamycin. 7. Sorval Centrifuge RC5B Plus, rotor GS3 and SS34 (Thermo Fisher Scientiﬁc). 8. Polyethylenglycol (PEG) solution: 20% (w/v) PEG 6000, 2.5 M NaCl. 9. Phage dilution buffer: 10 mM TrisHCl pH 7.5, 20 mM NaCl, 2 mM EDTA. 10. Mouse anti-pIII monoclonal antibody (clone 10C3, Mobitec, Go¨ttingen, Germany). 11. Polyclonal goat anti-mouse IgG whole-molecule alkaline phosphatase conjugate (Sigma-Aldrich).

2.10 Temperature In the following protocols, room temperature is deﬁned as a tem- and Humidity perature comprised between 20 and 25 C. Relative humidity should be maintained between 40 and 60%.

3 Methods

3.1 Macaque 1. After macaque quarantine, about 1 month-acclimation is immunization required in the animal-house facility. The macaque should be and Bone Marrow continuously protected from environmental pathogens to pre- Sampling vent the stimulation of the animal-immune system with undesirable antigens. The macaque could share its cage with other healthy-macaques (according to the European laws). If an infectious antigen is used, the animals should be isolated from the other animals. During the immunization process (Fig. 1a), the animals must be under regular examination of animal kee- pers, for well-being controls and injection sites checkup. 2. At the end of the acclimation 4–5 mL of bone marrow should be sampled under general anesthesia (intra-muscular injection À of 10 mg.kg 1 of ketamine or any relevant and approved anesthetic) with a trocar and a syringe (or alternatively with a syringe and a 23 G needle) and transferred in a 15 mL centrifuge tube containing 10–15% of citrate. All bone marrow sampling realized during this protocol should be realized alternatively in both humerus and iliac crests, to decrease the pain and allowing the local regeneration of the bone marrow. No more than 2 samplings should be realized in a 6 days-window. Analgesics have to be administrated to the animal in case of pain signs. Construction of Macaque Immune-Libraries 93

Fig. 1 Follow-up of the immune response during the hyper-immunization protocol. (a)Schematic representation of the evolution of the immune-response during the hyper-immunization. The immune response increases faster and faster after each immunization. When the immune-response is equal to that obtained after the previous immunization, the hyper-immunization protocol was successful (dotted line). (b) Typical representation of the evolution of amplification of the DNA region coding for the variables domains (here VH and VLκ)of the antibodies. Before the final boost, no significant amplification is observed. After the final immunization, the amplification increases quickly until a maximum of intensity and then decreases slowly 94 Arnaud Avril et al.

3. Starting from the first pre-immunization bone marrow sample, total RNA is retro-amplify and controlled for the absence amplification of the DNA region coding the antibody variable domains, according to Subheadings 3.4 and 3.5. Immunization protocol should not be started if significant amplification is observed. The results of this control (Fig. 1b)andtheserum obtained after bone marrow centrifugation (stored at À20 C) will serve as a reference for the follow-up of the immune response during the immunization protocol. 4. Once no significant PCR amplification is observed, immunize the macaque with a single-site or two-site subcutaneous injection (s) of a total of 100 μg (or 1010 inactivated virions) of the immunogen (see Note 2) diluted in 200 μLofPBSandmixwith200μL of Freund’s complete adjuvant (400 μL injected in total). 5. About 10 days after the immunization, sample ~5 mL of blood and control the immunization titer according to Subheading 3.2. 6. Immunize (“boost”) 30 and 60 days after the first immunization, as described previously, using incomplete Freund’s adjuvant. About 10 days after each immunization, 5 mL of bone whole blood and control the immunization titer according to Subheading 3.2. 7. Ten days after the third immunization a titer >1:100,000 should be reached. If the serum-titer significantly increases between the second and the third immunization realize additional immunization(s) until a stable titer superior to 1:100,000 is reached. If the titer did not increase between the second and the third immunization, that means that a hyper-immunization status is reached and that the final boost can be realized. 8. After a resting period of at least 120 days after the last immunization, sample ~ 4 mL of bone marrow and check for PCR-products according to Subheadings 3.4 and 3.5.No significant amplification is expected, meaning that the macaque’s immune response has returned to background level, and then realize a final boost with the immunogen. In the opposite case, if significant amplification is observed, leave the animal without any antigenic stimulation for some additional weeks, until no significant amplification could be observed. 9. After the final injection, realize iterative 4–6 mL bone marrow sampling; both first ones should be realized 2/3 and 6/7 days after the final immunization. The date and the frequency of the other samplings should be adapted according to the response of the animal. Stop the samplings when two consecutive decreases of PCR-products bands intensity on agarose-gel are observed after the maximum of amplification. Delay the sampling or stop it if there is any issue with the animal welfare. Sampling twice in a 6 days-window is normally not deleterious for the animal. Construction of Macaque Immune-Libraries 95

3.2 Determination Starting material: antigen(s) of interest and serum obtained after of the Serum-Titer blood-sampling (and stored at À20 C). in ELISA 1. Coat an ELISA plate (100 μL per well) with the antigen of interest and a negative control (such as KLH) diluted in PBS À (5 μg.mL 1). Coat a quantity of wells sufficient for two-fold dilutions of each serum from ~1:1000 to 1:2,048,000 (2 Â 12 wells). Each serum dilution has to be tested in parallel against the antigen of interest and the negative control. Incubate for 16 h (~overnight) at 4 C or alternatively, incubate for 2 h at 37 C. Wash the plate 3 times with 300 μL of PBS þ with a plate washer. 2. Add 200 μL of saturation solution (200 μL of PBS þ 2% skim milk), 2 h at 37 C. 3. In parallel of the saturation plate step, saturate also a 2.0 mL DeepWell™ plate using the same solution for 1.5 h at 37 C then prepare your serum dilution in the saturated-DeepWell™ plate (250–300 μL of each dilution). Serums should be diluted from 1:1000 to 1:2,048,000 in PBS þ 2% skim milk. 4. After the 2 h of saturation of the main-plate, wash three times. 5. Incubate 100 μL of each serum dilution during 2 h at 37 C. 6. A couple of minutes before the end of the incubation prepare the detection antibody according to the manufacturer’s recom- mendation. Use an antibody directly conjugated to the peroxidase and specific of the whole monkey-IgG (see Note 3). 7. Wash the plate three times. 8. Incubate 100 μL of the detection antibody in each well and incubate for 1 h and 37 C. 9. Wash the plate three times. 10. Add 90 μL of TMB and incubate under gentle orbital agitation and shielded from the light, until coloration is observed for the wells of interest. Read directly at 650 nM with a microplate spectrophotometer. If the plate cannot be read directly, stop the reaction by the addition of 90 μL of TMB stop solution and read in the 2 h at 450 nM. Determine the titer by diving the specific signal (serum dilution incubated with the antigen) by three times the background signal (same serum dilution incubated with the nonspecific antigen). The titer is defined as the lowest serum dilution giving a normalized-ratio superior to 1. 96 Arnaud Avril et al.

3.3 Total RNA Starting biological material: 4–6 mL of fresh bone marrow sampled Extraction in 15 mL centrifuge tube containing 10–15% of citrate. for the Follow-Up Before the beginning of the protocol, treat all materials (soni- to the Immune cator’s probe, micropipets, benchtop, gloves, etc.) with RNAse Response AWAY to remove RNAse and DNA contamination (sterile plastic tubes do not need to be treated) and prechilled the centrifuge at 4 C. 1. Centrifuge the bone marrow during 10 min at 500 Â g and 4 C. 2. After centrifugation, aliquot and store the serum (brown–yellow upper fraction) at À20 C; the serum will be used for the determination of the immunization’s titer according to Sub- heading 3.2. Transfer the bone-marrow (dark-red lower fraction) in a 50 mL centrifuge tube containing 0.75 volume of ® TRI Reagent LS for 0.25 volume of bone-marrow. Lyse the cells by passing the suspension several times though a 10 mL serological pipet. Try to minimize the loose of material on the surface of the serological pipet and the centrifuge tube (the bone marrow sticks on the plastic walls). 3. Place the tube on ice and sonicate the cells for 1 min with 6 s–pulses of 10 watt separated by pauses of 3 s. Incubate for 5 min at room temperature to ensure complete dissociation of nucleoprotein-complexes. After cell-lysis, samples can be stored at À70 C for up to 1 month. 4. Centrifuge for 10 min at 12.000 Â g and 4 C. After centrifugation, take the tube very carefully because the pellet is not solid and not clearly visible. 5. Transfer the supernatant (containing total-RNA, proteins, and low molecular mass DNA) very slowly and carefully in a new 50 mL centrifuge tube (see Note 4). ® 6. Add 0.1 mL of BCP per 0.75 mL of TRI Reagent used. Vigorously vortex the mixture during 15 s. Incubate at room temperature during 15 min. 7. Divide the mixture in 2 mL colorless centrifuge tubes. Centri- fuge for 15 min at 12000 g and 4 C. Centrifugation separates the mixture into three phases: a lower red organic phase (proteins), a white interphase (DNA), and a colorless upper aqueous phase (total RNA). 8. Angling the tubes at 45 and transfer the upper aqueous phase in a 50 mL centrifuge collection tube. To avoid DNA contamination, carefully transfer the aqueous phase without touching the interphase. Generally, the yield of RNA extraction is widely sufﬁcient for RT-PCR and it is preferable to leave a part of the aqueous phase rather than taking part of the interphase. Construction of Macaque Immune-Libraries 97

9. Add 0.5 mL of isopropanol (at room temperature) per 0.75 mL ® of TRI Reagent used and add 20 μg of ultra-pure and RNase- free glycogen (glycogen acts as a carrier for RNA precipitation), and does not inhibit RT-PCR (at concentrations 4 mg. À mL 1). Vortex during 15 s and incubate for 10 min at room temperature (to prevent that the salts in excess precipitate with the RNA). Distribute in a 2 mL colorless centrifuge tube with conical bottom. 10. Centrifuge at 12000 Â g for 15 min at 4 C. Place all the tubes in the same sense to facilitate the observation of the pellet after the centrifugation. 11. Pool all the RNA pellet together and discard carefully the supernatant with a micropipette (see Note 5). 12. Wash the pellets by adding ~2 mL of 75% ethanol (ethanol remove residual salt from the pelleted RNA). Mix gently by pipetting (do not vortex) and incubate for 10 min at room temperature (see Note 6). 13. Centrifuge 10 min at 7500 Â g. If the pellet ﬂoat of accumu- lates on a side of the tube, centrifuge at 12000 Â g (do not centrifuge at higher speed). 14. Discard the maximum of the supernatant. Dry the pellet at room temperature until almost all the ethanol is evaporated (generally about ~10 min). Do not dry the pellet by vacuum centrifuge. 15. Once almost all the ethanol is evaporated (if the pellet dries completely it can lose solubility) dissolve the pellet by pipetting with 50 μL of ultra-pure RT-PCR-grade water. The RNA pellet can be stored in ethanol up to 1 year at À20 C or at least 1 week at 2–8 C. 16. Sample 3 μL of the RNA and determine its concentration with ® a spectrophotometer (NanoDrop , see Note 7). The optimal concentration for the pursuit of the protocol is about700 μg. À mL 1. Approximately, 200 μg of total RNA should have been puriﬁed from 5 mL of bone marrow and 20 μg are required for reverse transcription. The A260 /A280 ratio should be 1.7 and an optimum ~2.0 is preferable to consider the RNA pure from proteins contamination. The A260/A230 ratio should be 1.8 and a ratio between 2.0 and 2.4 is preferable, to consider the RNA pure from organic compounds. The A260/A240 ratio should be 1.4 to consider the RNA pure from salts. If ratios are below these thresholds, new RNA precipitation is required. 17. RNA preparation quality controls can be realized on standard agarose gel. 98 Arnaud Avril et al.

3.4 cDNA Synthesis Starting material: 20 μg of RNA extracted from the bone marrow À sampling with a minimal concentration of 650 μg.mL 1. 1. Reverse transcription is performed using the SuperScript IV First-Strand kit. The mixes should be prepared on ice on a 200 μL PCR-grade tube. Set up mixture for the ﬁrst-strand cDNA synthesis as below:

Volume or Final concentration or Solution or component quantity quantity

À Total RNA 20 μg20μg (500 μg.mL 1) À Oligo dT 20 4 μL 5 ng.μL 1 dNTP-mix (10 mM 4 μL 1 mM each each) DEPC-treated water Up to 40 μL

2. Denature the RNA for 5 min at 70 C. Afterward directly chill down on ice for 5 min. 3. During the denaturation prepare the following cDNA synthesis mix. Each component has to be added in the indicated order.

Volume Final Solution or component (μL) concentration

RT buffer (5Â)81Â

25 mM MgCl2 16 5 mM 0.1 M DTT 8 10 mM RNAseOUT 4 4 μL Superscript IV reverse transcriptase 4 200 U (200 U/μL)

4. Directly on the thermocycler, add the 40 μL of the cDNA synthesis mix to the 40 μL of RNA/Primers mixture. Incubate the 80 μL mixture for 10 min at 25 C for primer annealing. Afterward incubate for 50 min at 50 C for ﬁrst-strand synthesis. 5. Denature the RNA/DNA hybrids and the enzyme for 5 min at 85 C and thereafter keep the cDNA mixture on ice until PCR ampliﬁcation. Sample 3 μL of the cDNA to measure the RNA ® concentration in the NanoDrop to determine if the reverse- transcription was performed correctly (use reaction buffer as blank). cDNA can be stored at least 1 year at À20 C. Construction of Macaque Immune-Libraries 99

3.5 PCR Starting material: cDNA from RT-PCR. Amplification ® 1. PCR amplification is performed with Invitrogen™ Platinium of the Heavy and Light Taq polymerase High Fidelity (see Note 8). The cDNA will be Chain Variable Domain used as a template to amplify the variable domain of the heavy and light chains. The PCR amplification is realized with two “reverse” primers specific of the constant domain of the IgGγ1 heavy (VH) or of the κ and λ light (VL) chain and with several “forward” primers specific of the N-terminal extremity of the VH, VLκ,andVLλ domain. Set up the PCR reactions as follows (prepare 30Â the master mix for the 27 PCR reactions):

Per reaction For Final concentration Component (48 μL) 30 reactions or quantity

Nuclease-free water To 48 μL To 1.44 mL 10Â PCR buffer 5 μL 150 μL1Â without magnesium

50 mM MgCl2 1.5 μL45μL 1.5 mM dNTP mix (10 mM 1 μL30μL 0.2 μM each each) À cDNA 100 ng 3 μg2μg.mL 1 PlatiniumTM Taq 0.2 μL6μL 2 U per reaction DNA polymerase

2. Divide the master mix in 480 μL for VH, 384 μL for Kappa and 576 μL for Lambda and add the relevant constant (“revers”) primer in each mix:

For 10 VH For 8 VLκ For 12 VLλ Final Component reactions reactions reactions concentration

Master mix 480 μL 384 μL 576 μL 5 μMof 20 μLof 16 μLof 24 μLof 0.2 μM relevant MHMacIg MHMac MHMacL C-terminal GCH1scFv- KappaCL- ambdaCL- reverse HindIII_r NotI_r NotI_r primers speciﬁc of VH, VLκ or VLλ constant domains.

On ice, add directly 2 μL of each forward (see Table 1) primer on 200 μL PCR tubes (1 forward-primer per tube): prepare 9 tubes for the VH amplification, 7 tubes for VLκ amplification, and 11 tubes for VLλ amplifications. 100 Arnaud Avril et al.

3. Add 48 μL of the corresponding master mix to the 2 μL of each forward primer and gently mix. 4. Carry out the PCR using the following program:

Step Temperature Time

Initial denaturation 94 C 2 min Denature 94 C 1 min 30Â Primer annealing 55 C 1 min

DNA extension 72 C 2 min Hold 8 C Stand-by

5. Control the PCR products by electrophoresis-migration on agarose-gel (see Notes 9–12). Use 1Â TBE gels containing 0.8% agarose and ethidium bromide (concentration according to the manufacturer’s recommendations); run with 0.5Â TBE at 120 V. Use a relevant size marker (such as Smart Ladder MW1700–10, Eurogentec). Determine the DNA concentration with the NanoDrop (use the reaction buffer as blank). Analyze the gel with a GelDoc system, under UV-light. The PCR products of amplification are about 750 bp. Pool VH, VLκ and VLλ subfamilies separately (i.e., three different pools). For each subfamily, pool only the amplicons presenting the best amplification compared to the amplification obtained before the last immunization. Store the products of amplification at À20 C. 6. During the days following the final immunization, an optimum of amplification should be obtained. This optimal amplification corresponds to the PCR products of higher intensity on agarose gel. The VH, VLκ, and VLλ for the selected-PCR-products of amplification should be used for the pursuit of the protocol.

3.6 Second Antibody Starting material: pools of VH, VLκ, and VLλ first PCR products. Gene PCR 1. During the second PCR the DNA will be re-amplified to introduce restriction sites for library cloning. Then after, the DNA coding for VLκ or VLλ will be introduced first in the phagemid vector and finally the DNA coding for VH will be introduced. 2. Set up the PCR reactions as follows (prepare 30Â mastermix for 27 PCR reactions): Construction of Macaque Immune-Libraries 101

Solution or component Volume (μL) Final concentration or quantity dH2O 2200 Buffer with MgCL2 (5Â) 600 1Â dNTPs (10 mM each) 60 200 μM each À GoTaq 5 U.μL 1 15 2.5 U

3. Divide the master mix in 900 μL for VH, 700 μL for VLκ, and 1100 μL for VLλ. 4. Add to each of the three reactions the corresponding reverse primers as follows:

Antibody Volume Final concentration gene Primer (μL) (μM)

VH MHMacIgGCH1scFv- 18 0.2 HindIII_r Kappa MHMacKappaCL-NotI_r 14 0.2 Lambda MHMacLambdaCL- 22 0.2 NotI_r

5. Add the corresponding PCR products of the ﬁrst PCR as follows:

PCR product Volume (ng)

VH 900 VL kappa 700 VL lambda 1100

6. Divide the solutions into 9 (VH), 7 (VLκ),and11(VLλ)PCR reactions, each with 98 μLofmastermix.Add2μL(10μM, 0.2 μM ﬁnal concentration) of the subfamily speciﬁc forward primer (see Table 2): Carry out the PCR using the following program:

Step Temperature (C) Time

Initial denaturation 94 1 min Denature 94 1 min 20Â Primer annealing 57 1 min

DNA extension 72 1.5 min Final extension 72 10 min Hold 8 Stand-by 102 Arnaud Avril et al.

7. Separate the PCR products with 1Â TAE gels containing 1.5% agarose and BET (concentration according to the manufacturer’s recommendations); run with 0.5Â TAE at 120 V. Use a relevant marker of size (such as Smart Ladder MW1700–10, Eurogentec). Analyze the gel with a GelDoc system (or analog system), under UV-light. The PCR amplified products are about 400 bp size. Cut out the amplified antibody genes on preparative low-melting point agarose gel and purify the PCR products using a gel extraction kit according to the manufacturer’s instructions. Pool all VH, kappa, and lambda subfamilies separately. Determine the DNA concentration with the NanoDrop (use the reaction buffer as blank). Store the three purified second PCR pools at À20 C.

3.7 First Cloning Starting material: second-PCR pools. Step: VL 1. Prepare the mastermix for the digestion of 5 μg of pHAL35 phagemid vector (5 μg for the cloning and 5 μg for the controls), 2 μgofVLκ pool and 2 μgofVLλ during 2 h at 37 C, according to the tables below (see Note 12). 2. In parallel, control the success of the restriction-reaction by digesting pHAL35 with only one enzyme (MluI or NotI). Analyze the products of singles digestions by gel-electrophoresis with 1Â TAE gels containing 1.5% agarose and BET (concentration according to the manufacturer’s recommendations); run with 0.5Â TAE at 120 V. After migration compare with the undigested phagemid: undigested phagemid migration is slower than linear (digested) phagemid. For the cloning, use only material where single digestions are successful and where no degradation is visible in the double digest.

pHAL35 digestion

Volume or Final concentration or Solution or compound quantity quantity

dH2O To 83 μL À pHAL35 5 μg60μg.mL 1 NEB buffer 3 (10Â)10μL1Â BSA (100Â)1μL1Â NEB MluI (10 U.μ 3 μL30U À L 1) NEB NotI (10 U.μ 3 μL30U À L 1) Construction of Macaque Immune-Libraries 103

VLκ or VLλ digestion

Volume or Final concentration or Solution or compound quantity quantity dH2OTo83μL À VLκ or VLλ pooled 2 μg24μg.mL 1 NEB buffer 3 (10Â)10μL1Â BSA (100Â)1μL1Â NEB MluI (10 U.μ 3 μL30U À L 1) NEB NotI (10 U.μ 3 μL30U À L 1) 3. Incubate the digestion-mixture at 37 C for 2 h. 4. After incubation, control the digestion of the vector by using a 5 μL aliquot on 1Â TAE 1% agarose gel electrophoresis. If the vector is not fully digested, extend the incubation time. 5. When double digestion is complete, inactivate the enzymes at 65 C for 10 min. À 6. Add 0.5 μL CIP (1 U.μL 1) and incubate at 37 C for 30 min. Repeat this step once. 7. Purify the digested-pHAL35 phagemid and the digested-VLκ/ VLλ with a PCR purification kit (such as Purelink PCR purification kit, Thermofisher Scientific), according to the manufacturer’s instructions and elute with 50 μL elution buffer or water. During the double digestion, the short stuffer fragment containing multiple stop codons between MluI and NotI in pHAL35 was removed. Determine the DNA concentration with a NanoDrop. 8. Ligate the pHAL35 phagemid (4243 bp) and VL (~380 bp) as follows:

Ligation of pHAL35 with VLκ or VLλ

Volume or Final concentration or Solution or compound quantity quantity dH2O To 89 μL pHAL35 1000 ng 1000 ng VLκ or VLλ 270 ng 270 ng Promega T4 ligase buffer 10 μL1Â (10Â) À T4 ligase (3 U.μL 1)1μL3U 104 Arnaud Avril et al.

9. Incubate for 16 h (overnight) at 16 C. 10. Inactivate the ligation at 65 C for 10 min. 11. Precipitate the ligation with 10 μL of 3 M sodium acetate pH 5.2 and 250 μL of ethanol, incubate for 2 min at room temperature, and centrifuge for 5 min at 16,000 Â g and 4 C. 12. Wash the pellet with 500 μL 70% (v/v) ethanol and pellet the DNA for 2 min at 16,000 Â g and 4 C. Repeat this step once and resolve the DNA pellet in 35 μL sterile dH2O. 13. Slowly thaw 25 μL electrocompetent E. coli XL1-Blue MRF’ on ice and mix with the ligation reaction. 14. Mix the 35 μL of product of ligation with the 25 μLofE. coli cells and transfer to a prechilled 1 mm electroporation cuvette. Before electroporation, dry the electrode of the cuvette with a tissue paper. 15. Perform a 1.7 kV pulse using an electroporator (4–5 ms pulse for optimal electroporation efﬁciency). Immediately, add 1 mL 37 C pre-warmed SOC medium, transfer the suspension to a 2 mL cap, and shake for 1 h at 600 rpm and 37 C. À 16. To determine the amount of transformants, use 10 μL(¼10 2 dilution) of the transformation and perform a dilution series À À down to 10 6 dilution. Plate out the 10 6 dilution on a 12 cm Â 12 cm 2ÂYT-GAT agar Petri dish and incubate overnight at 37 C. 17. Plate out the remaining 990 μL on a 24 cm Â 24 cm 2ÂYT- GAT agar Petri dish and incubate for ~16 h (overnight) at 37 C. 18. The day after Petri dishes incubation, calculate the amount of À transformants on the 12cmx12cm Petri dishes (10 6 dilution). The number of colony-forming units (cfu) should be between 1 Â 106 and 5 Â 108 cfu. Control the presence of full-size insert in several colonies by colony PCR (according to Sub- heading 3.9.) and keep only the sublibraries with more than 75% of full-size inserts (see Note 13). 19. Float off the colonies on the 24 cm Â 24 cm Petri dish with the 40 mL 2ÂYT medium using a Drigalsky spatula. Use 5 out of 40 mL bacteria solution for midi plasmid preparation according to the manufacturer’s instructions. Determine the DNA concentration with a NanoDrop.

3.8 Second Cloning Starting material: pHAL35 þ VL library. Step: VH 1. Digest the pHAL35-VL repertoire and the VH PCR products. Always perform additional single-enzyme digestions of the vector in parallel (as done for VL cloning): Construction of Macaque Immune-Libraries 105 pHAL35 þ VL digestion

Volume or Final concentration or Solution or compound quantity quantity dH2OTo82μL pHAL35 þ VL 5 μg5μg NEB CutSmart (10Â)10μL1Â NEB SﬁI-HF (20 U/μL) 1.5 μL30U NEB HindIII-HF (20 U/ 1.5 μL30U μL)

VH digestion

Volume or Final concentration or Solution or compound quantity quantity dH2OTo82μL VH 2 μg2μg NEB Cutsmart (10Â)10μL1Â NEB SfiI- (20 U/μL) 1.5 μL30U NEB HindIII-HF (20 U/ 1.5 μL30U μL) 2. Incubate at 37 C for 2 h. Control the digest of the vector by using a 5 μL aliquot on 1% agarose gel electrophoresis (see Note 14). 3. Inactivate the digestion at 80 C for 20 min. À 4. Add 0.5 μL CIP (1 U.μL 1) and incubate at 37 C for 30 min. Repeat this step once. 5. Purify the vector and the PCR product using a PCR purification kit (such as Purelink PCR purification kit, ThermoFisher Scientific), according to the manufacturer’s instructions and elute with 50 μL elution buffer or water. During the double digestion, the short stuffer fragment containing multiple stop codons between MluI and NotI in pHAL35 was removed. Determine the DNA concentration with a NanoDrop. 6. Ligate the vector pHAL35 þ VL (~4610 bp) and VH (~380 bp) as follows: 106 Arnaud Avril et al.

Ligation of pHAL3535 þ VL with VLκ or VLλ

Solution or Volume or Final concentration or compound quantity quantity

dH2OTo89μL pHAL35 þ VL 1000 ng 1000 ng VH 250 ng 250 ng T4 ligase buffer 10 μL1Â (10Â) À T4 ligase (3 U.μL 1)1μL3U

7. Incubate at 16 C ~ 16 h (overnight). 8. Inactivate the ligation at 65 C for 10 min. 9. Precipitate the ligation with 10 μL of 3 M pH 5.2 sodium acetate and 250 μL ethanol, incubate for 2 min at RT, and centrifuge for 5 min at 16,000 Â g and 4 C. 10. Wash the pellet with 500 μL 70% (v/v) ethanol and pellet the DNA for 2 min at 16,000 Â g and 4 C. Repeat this step once and resolve the pellet in 35 μL sterile dH2O. 11. Slowly thaw 25 μL electrocompetent E. coli XL1-Blue MRF’ on ice and mix with the ligation reaction. 12. Mix the 35 μL of product of ligation with the 25 μLofE. coli cells and transfer to a prechilled 1 mm electroporation cuvette. Before electroporation, dry the electrode of the cuvette with a tissue paper. 13. Perform a 1.7 kV pulse using an electroporator (4–5 ms pulse for optimal electroporation efﬁciency). Immediately, add 1 mL 37 C pre-warmed SOC medium, transfer the suspension to a 2 mL cap, and shake for 1 h at 600 rpm and 37 C. À 14. To determine the amount of transformants, use 10 μL(¼10 2 dilution) of the transformation and perform a dilution series À À down to 10 6 dilution. Plate out the 10 6 dilution on a 12 cm x12cm2ÂYT-GAT agar Petri dish and incubate overnight at 37 C. 15. Plate out the remaining 990 μL on a 24 cm x 24 cm 2ÂYT- GAT agar Petri dish and incubate for ~16 h (overnight) at 37 C. 16. The day after Petri dishes incubation, calculate the amount of À transformants on the 12cm Â 12cm Petri dishes (10 6 dilution). The number of cfu should be comprised between 1 Â 106 and 5 Â 108 cfu. Control the presence of full-size insert in several colonies by colony PCR (according to Construction of Macaque Immune-Libraries 107

Subheading 3.9) and keep only the sublibraries with more than 75% of full-size inserts (see Note 13). 17. Float off the colonies on the 24cmx24cm Petri dishes with 40 mL 2ÂYT medium using a Drigalsky spatula. Use 5 out of 40 mL bacteria solution for midi plasmid preparation according to the manufacturer’s instructions. Determine the DNA concentration with a NanoDrop. When all transformations are done, thaw one aliquot of each sublibrary on ice, mix all sublibraries, and make new aliquots for storage at À80 C.

3.9 Single- Starting material: Clones isolated on petri dishes (from the previous Colony PCR transformations). 1. Choose ~16 single colonies per transformation. Set up the 10 μL PCR reaction per colony as follows (for primer see Table 3):

Colony PCR mix

Solution or Volume or Final concentration or compound quantity quantity

dH2O 7.5 μL GoTaq buffer (5Â)2μL1Â dNTPs (10 mM each) 0.2 μL 200 μM each MHLacZPro_f 0.1 μL 0.1 μM 10 μM MHgIII_r10 μM 0.1 μL 0.1 μM À GoTaq (5 U.μL 1) 0.1 μL 0.5 U Template Freshly handpicked single colony from a petri dish

2. After PCR, analyze the products of single digestions by gel-electrophoresis with 1Â TAE gels containing 1.5% agarose and run with 0.5Â TAE at 120 V. The product of ampliﬁcation is about 1100 bp. Throw away all sublibraries with less than 75% of full-size inserts.

3.10 Library Starting material: pHAL35 þ VH þ VLκ/VLλ DNA (see Notes 15 Packaging and scFv and 16). Phage Production 1. To package the library, inoculate 400 mL 2ÂYT-GA in a 1 L Erlenmeyer ﬂask with 1 mL antibody gene library stock. Grow at 250 rpm at 37 C up to an O.D.600 nm ~ 0.5. 2. Infect 25 mL bacteria culture (~1.25 Â 1010 cells) with 2.5*1011 colony forming units (cfu) of the helper-phage M13 K07 or Hyperphage according to a multiplicity of infection (moi) ¼ 1:20 (see Note 17). Incubate for 30 min at 37 C 108 Arnaud Avril et al.

without shaking and the following 30 min with 250 rpm at 37 C. 3. To remove the glucose which represses the lac promoter of pHAL35 and therefore the scFv::pIII fusion protein expression, harvest the cells by centrifugation for 10 min at 3.200 Â g in 50 mL polypropylene tubes. [41] 4. Resuspend the pellet in 400 mL 2ÂYT-AK in a 1 L Erlenmeyer ﬂask. Produce scFv-phage overnight at 250 rpm and 30 C. 5. Pellet the bacteria by centrifugation for 10 min at 10,000 Â g in two 500 mL centrifuge tubes. If the supernatant is not clear, centrifuge again to remove remaining bacteria. 6. Precipitate the phage from the supernatant by adding 1/5 volume of PEG 6.000 solution in two 500 mL tubes. Incubate for 1 h at 4 C with gentle shaking, followed by 1 h centrifugation at 10.000 Â g. 7. Discard the supernatant, resolve each pellet in 10 mL phage dilution buffer in 50 mL centrifuge tubes and add 1/5 volume PEG 6.000 solution. 8. Incubate on ice for 20 min and pellet the phage by centrifugation for 30 min at 10,000 Â g. 9. Discard the supernatant and put the open tubes upside down on tissue paper. Let the viscous PEG 6.000 solution move out completely. Resuspend the phage pellet in 1 mL phage dilution buffer. Titer the phage preparation (as described in Subheading 3.11). Store the packaged antibody phage library at 4 C. 10. The library packaging should be controlled by 10% SDS-PAGE and western-Blot with anti-pIII immunostain (mouse anti-pIII 1:2000, goat anti-mouse IgG AP conjugate 1:10,000). Wild- type pIII has a calculated molecular mass of 42.5 kDa, but it runs at an apparent molecular mass of 65 kDa in SDS-PAGE. Accordingly, the scFv::pIII fusion protein runs at about 95 kDa [42].

3.11 Phage Titration 1. Inoculate 5 mL 2ÂYT-T in a 100 mL Erlenmeyer ﬂask with E. coli XL1-Blue MRF0 and grow overnight at 37 C and 250 rpm. 2. Inoculate 50 mL 2 Â YT-T with 500 μL overnight culture and grow at 250 rpm at 37 CuptoOD600 ~ 0.5. 3. Make serial dilutions of the phage-suspension in 1Â PBS. The phage preparation after ampliﬁcation should have a titer À between 1012 and 1014 phages.mL 1. 4. Infect 50 μL bacteria with 10 μL of phage-dilution and incubate for 30 min at 37 C. 5. You can perform titrations in two different ways: Construction of Macaque Immune-Libraries 109

(a) Plate the 100 μL infected bacteria on 2 Â TY-GA agar plates (9 cm Petri dishes). À (b) Serially dilute phages-preparation in 2XYT from 10 6 to À 10 16 Pipet 10 μL of each dilution and spot-it (in triplicate, without touching the Petri dish with the tip) on 2 Â TY-GA agar plates. About 12 titering spots can be placed on one 9 cm Petri dish. 6. Incubate the plates overnight at 37 C. À 7. Count the colonies and calculate the cfu.mL 1 titer according to the dilution.

4 Notes

1. Chloroform could alternatively be used instead of BCP with only few adaptations of the protocol, but chloroform is most dangerous for health and not compatible with plastic consumable. 2. Immunization with several antigens is possible if there is no dominant epitope. Immunize in the shaven back of the animal. A particular attention should be done to the injection sites as superficial granuloma can appear in the days/weeks following immunization with complete Freund’s adjuvant (in a less extent with incomplete Freund’s adjuvant). Such granulomas have to be removed by surgery to prevent their ulceration. Blood sample should be harvested 10 days after each antigen injection to follow the rise of the immune response by standard ELISA. 3. Anti-human whole IgG may cross-react with macaque IgG. 4. If a part of the pellet is also transferred, the quality (purity) of the RNA extraction will be decreased. 5. Generally the pellet is very small, white or colorless and on the side of the bottom of the tube. ® 6. Use at least 1 mL of 75% ethanol per 0.75 mL of TRI Reagent LS initially used. ® 7. If a NanoDrop device is not available, it is possible to determine the RNA concentration with a conventional spectrophotometer. In this case, apply the convention that 1 OD260 equals À a RNA concentration of 40 μg.mL 1. 8. It is possible to realize several PCR amplifications in parallel because cDNA is more stable than RNA for long-term storage. 9. If there are significant unspecific bands, the migration can be realized with 1Â TAE gels containing 1.5% agarose and BET. Then, the band of interest can be cut out and purified using a 110 Arnaud Avril et al.

gel extraction kit according to the manufacturer’s instructions. Pool all VH, VLκ, and VLλ subfamilies separately. 10. The VH amplifications of VH subfamilies sometimes result also in longer PCR products. Cut out only the ~380 bp fragment. The amplifications of kappa subfamilies should always give a clear ~650 bp fragment (complete light chain). When amplifying lambda subfamilies often other PCR products are generated, especially the amplification of the lambda2 subfamily results often in slushy bands. If some subfamilies are bad amplified and no clear ~650 bp fragment is detectable, use only the ~650 bp fragments from the well-amplified subfamilies. 11. As the first PCR amplifies the full LC, it is also possible to construct Fab or scFab libraries from this material. 12. Any relevant phagemid vector can be used. If you use another vector, adapt the protocol with the correct restriction enzymes. 13. Transformation rates between 107 and 108 clones per transformation are expected. 14. Generally, the digestion with HindIII is incomplete after 2 h. In this case, inactivate the enzymes by heating up to 65 C for 10 min, add additional 5 μL of HindIII and incubate overnight. You can use also higher concentrated HindIII. Alterna- tively, perform the SfiI digest first for 2 h, inactive the digest, and afterward perform the HindIII digest. This problem only occurs when HindIII is used instead of HindIII-HF. 15. To minimize loss of diversity, avoid too many freeze and thaw steps, e.g., when constructing an immune library make eight transformations in parallel and directly package the immune library. 16. When making a large immune library, combine only a glycerol stock of each sublibrary which corresponds to a maximum of 2 Â 109 independent clones to ensure that the library diversity can be kept when packaging 1 mL of mixed library glycerin stock. When the library size is bigger than 2 Â 109 independent clones, do not package the library as complete library, package “blocks” of sublibraries. Combine the phage particles of each “block” before panning to get the final complete library. 17. The use of Hyperphage as helperphage instead of M13 K07 offers oligovalent phage-display, facilitates the selection of specific binders in the first and most critical panning round by avidity effect. The Hyperphage should be only used for library packaging. For the following panning rounds use M13 K07 to enhance the stringency of the panning process. Construction of Macaque Immune-Libraries 111

Acknowledgments

This review contains updated and revised parts of Pelat et al. (2010) [43].

References

1. Ko¨hler G, Milstein C (1975) Continuous cul- activity analyzed in vitro and in vivo. J Antivir tures of fused cells secreting antibody of pre- Antiretrovir 1(1):36–42 defined specificity. Nature 256 13. Thie H, Toleikis L, Li J et al (2011) Rise and (5517):495–497 fall of an anti-MUC1 specific antibody. PLoS 2. Emmons C, Hunsicker LG (1987) One 6(1):e15921 Muromonab-CD3 (Orthoclone OKT3): the 14. Trott M, Weiβ S, Antoni S et al (2014) Func- first monoclonal antibody approved for thera- tional characterization of two scFv-fc antibo- peutic use. Iowa Med 77(2):78–82 dies from an HIV controller selected on soluble 3. Presta LG (2006) Engineering of therapeutic HIV-1 Env complexes: a neutralizing V3- and a antibodies to minimize immunogenicity and trimer-specific gp41 antibody. PLoS One 9(5): optimize function. Adv Drug Deliv Rev 58 e97478 (5–6):640–656 15. Kawamura S, Omoto K, Ueda S (1990) Evolu- 4. Getts DR, Getts MT, McCarthy DP, Chastain tionary hypervariability in the hinge region of EML, Miller SD (2010) Have we overesti- the immunoglobulin alpha gene. J Mol Biol mated the benefit of human(ized) antibodies? 215(2):201–206 MAbs 2(6):682–694 16. Andris JS, Miller AB, Abraham SR et al (1997) 5. Ma B, Osborn MJ, Avis S et al (2013) Human Variable region gene segment utilization in antibody expression in transgenic rats: compar- rhesus monkey hybridomas producing human ison of chimeric IgH loci with human VH, D red blood cell-specific antibodies: predomi- and JH but bearing different rat C-gene nance of the VH4 family but not VH4-21 regions. J Immunol Methods 400–401:78–86 (V4-34). Mol Immunol 34(3):237–253 6. Bruggemann€ M, Osborn MJ, Ma B et al 17. Thullier P, Chahboun S, Pelat T (2010) A (2015) Human antibody production in trans- comparison of human and macaque (Macaca genic animals. Arch Immunol Ther Exp Mulatta) immunoglobulin germline V regions (Warsz) 63(2):101–108 and its implications for antibody engineering. 7. Frenzel A, Schirrmann T, Hust M (2016) MAbs 2(5):528–538 Phage display-derived human antibodies in 18. Avril A, Froude JW, Mathieu J, Pelat T, Thul- clinical development and therapy. MAbs 8 lier P (2014) Isolation of antibodies from (7):1177–1194 non-human primates for clinical use. Curr 8. Hoogenboom HR (2005) Selecting and Drug Discov Technol 11(1):20–27 screening recombinant antibody libraries. Nat 19. Thullier P, Huish O, Pelat T, ACR M (2010) Biotechnol 23(9):1105–1116 The humanness of macaque antibody 9. Chan CEZ, Lim APC, MacAry PA, Hanson BJ sequences. J Mol Biol 396(5):1439–1450 (2014) The role of phage display in therapeutic 20. Poole JA, Meng J, Reff M, Spellman MC, antibody discovery. Int Immunol 26 Rosenwasser LJ (2005) Anti-CD23 monoclo- (12):649–657 nal antibody, lumiliximab, inhibited allergen- 10. Kugler€ J, Wilke S, Meier D et al (2015) Gener- induced responses in antigen-presenting cells ation and analysis of the improved human and T cells from atopic subjects. J Allergy Clin HAL9/10 antibody phage display libraries. Immunol 116(4):780–788 BMC Biotechnol 15:10 21. Gottlieb AB, Kang S, Linden KG, et al. Evalu- 11. Amersdorfer P, Wong C, Smith T et al (2002) ation of safety and clinical activity of multiple Genetic and immunological comparison of doses of the anti-CD80 monoclonal antibody, anti-botulinum type a antibodies from immune galiximab, in patients with moderate to severe and non-immune human phage libraries. Vac- plaque psoriasis. Clin Immunol 2004;111 cine 20(11–12):1640–1648 (1):28–37 12. Duan T, Ferguson M, Yuan L, Xu F, Li G 22. Mould DR, Davis CB, Minthorn EA et al (2009) Human monoclonal fab antibodies (1999) A population pharmacokinetic- against West Nile virus and its neutralizing pharmacodynamic analysis of single doses of 112 Arnaud Avril et al.

clenoliximab in patients with rheumatoid properties, characteristics and construction. J arthritis. Clin Pharmacol Ther 66(3):246–257 Mol Biol 417(3):129–143 23. Bugelski PJ, Herzyk DJ, Rehm S et al (2000) 33. Zhao A, Tohidkia MR, Siegel DL, Coukos G, Preclinical development of keliximab, a Prima- Omidi Y (2016) Phage antibody display tized anti-CD4 monoclonal antibody, in libraries: a powerful antibody discovery plat- human CD4 transgenic mice: characterization form for immunotherapy. Crit Rev Biotechnol of the model and safety studies. Hum Exp Tox- 36(2):276–289 icol 19(4):230–243 34. Ehrlich PH, Moustafa ZA, Justice JC et al 24. Miethe S, Rasetti-Escargueil C, Avril A et al (1988) Human and primate monoclonal anti- (2015) Development of human-like scFv-fc bodies for in vivo therapy. Clin Chem 34 neutralizing Botulinum neurotoxin E. PLoS (9):1681–1688 One 10(10):e0139905 35. Ehrlich PH, Moustafa ZA, Justice JC, Harfeldt 25. Avril A, Miethe S, Popoff MR et al Isolation of KE, Ostberg L (1988) Further characterization nanomolar scFvs of non-human primate origin, of the fate of human monoclonal antibodies in cross-neutralizing botulinum neurotoxins A1 rhesus monkeys. Hybridoma 7(4):385–395 and A2 by targeting their heavy chain. BMC 36. Ehrlich PH, Harfeldt KE, Justice JC, Moustafa Biotechnol 15(1):86 ZA, Ostberg L (1987) Rhesus monkey 26. Noy-Porat T, Rosenfeld R, Ariel N et al (2016) responses to multiple injections of human Isolation of anti-ricin protective antibodies monoclonal antibodies. Hybridoma 6 exhibiting high affinity from immunized (2):151–160 non-human primates. Toxins (Basel) 8(3) 37. Reichert JM (2010) Antibodies to watch in 27. Pelat T, Hust M, Hale M et al (2009) Isolation 2010. MAbs 2(1):84–100 of a human-like antibody fragment (scFv) that 38. Ranuio J, Estrellado A, Harris S et al (2015) neutralizes ricin biological activity. BMC Bio- Anti-drug antibody responses to Lumiliximab technol 9:60 are not detected in relapsed refractory CLL 28. Pelat T, Hust M, Laffly E et al (2007) High- patients treated with Lumiliximab in combina- affinity, human antibody-like antibody frag- tion with FCR in a phase 1/II study. Blood ment (single-chain variable fragment) neutra- 112(11):4210 lizing the lethal factor (LF) of bacillus anthracis 39. Abhinandan KR, Martin ACR (2007) Analyz- by inhibiting protective antigen-LF complex ing the “degree of humanness” of antibody formation. Antimicrob Agents Chemother 51 sequences. J Mol Biol 369(3):852–862 (8):2758–2764 40. Storz U (2016) Rituximab: how approval his- 29. Hust M, Meyer T, Voedisch B et al (2011) A tory is reflected by a corresponding patent fil- human scFv antibody generation pipeline for ing strategy. MAbs 6(4):820–837 proteome research. J Biotechnol 152 41. Bazan J, Całkosin´ski I, Gamian A (2012) Phage (4):159–170 display--a powerful technique for immunother- 30. Rasetti-Escargueil C, Avril A, Chahboun S et al apy: 1. Introduction and potential of therapeu- (2015) Development of human-like scFv-fc tic applications. Hum Vaccin Immunother 8 antibodies neutralizing Botulinum toxin sero- (12):1817–1828 type B. MAbs 7(6):1161–1177 42. Hust M, Jostock T, Menzel C et al (2007) 31. Hust M, Dubel€ S (2004) Mating antibody Single chain fab (scFab) fragment. BMC Bio- phage display with proteomics. Trends Bio- technol 7:14 technol 22(1):8–14 43. Pelat T, Hust M, Thullier P (2010) Antibody 32. Qi H, Lu H, Qiu H-J, Petrenko V, Liu A engineering. Springer, Berlin (2012) Phagemid vectors for phage display: Chapter 6

Construction of Bovine Immunoglobulin Libraries in the Single-Chain Fragment Variable (scFv) Format

Ulrike S. Diesterbeck

Abstract

Recombinant immunoglobulins are an excellent tool for diagnosis, treatment, and passive immunization. Phage display offers a robust technique for the selection of recombinant antibodies from immunoglobulin libraries. The construction of immunoglobulin libraries for veterinary purposes was restricted by the lack of knowledge about species-specific diversities. The now available data enable the construction of highly diverse libraries in livestock like cattle. Using diverse primer sets, the immunoglobulin repertoire is amplified and ligated into a phagemid. Infection of E. coli with filamentous phages allows the display of the immunoglobulin fragments on the surface as a fusion protein to the phage’s minor coat protein 3.

Key words Bos taurus, Bovine immunoglobulins, Immunoglobulin libraries, Bovine scFv, Heavy chain variable region, Light chain variable region

1 Introduction

The following protocol describes the construction of bovine single- chain Fragment variable (scFv) libraries for phage display based on the current knowledge about immunoglobulin diversity in cattle. Recombinant human antibodies possess a successful history in the therapy of inflammatory and autoimmune diseases, and cancer [1]. Species-specific monoclonal antibodies are not only useful to protect, for example, new born animals from infections and to treat infectious diseases and cancer but also in diagnostic test systems. Hybridoma-derived monoclonal antibodies from rodent animal models are not always available. The first construction of a bovine phage-display Fab (Fragment antigen-binding) library used a limited primer set to select a panel of Fab fragments against recombinant GST/BPV-4 L2 fusion protein [2]. In order to construct recombinant species-specific immunoglobulin libraries and to engineer selected fragments, it is critical

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_6, © Springer Science+Business Media LLC 2018 113 114 Ulrike S. Diesterbeck

to investigate the characteristic diversity of a species’ immunoglobulin gene loci. In early studies, the genomic locations of immunoglobulin loci were mapped using classical fluorescence in situ hybridization (FISH) experiments. The equine heavy chain locus was mapped to the Equus caballus autosome 24 [3]. First analyses about its potential diversity were performed using cDNA clones [4, 5] and sequencing of equine-human chimeric scFv [6]. Equine immunoglobulin diversity and locus organization were enabled by examining the whole genome [7]. In contrast, on the bovine genome the immunoglobulin heavy chain locus is still incompletely covered [8, 9]. The heavy chain locus was assigned to Bos taurus autosome (BTA) 21 by FISH spanning approximately 250 kb [9, 10]. Furthermore, unplaced contigs and genomic sequences gave insights into the putative diversity of bovine immunoglobulin heavy chains [8, 9, 11]. The bovine genome consists of a limited number of heavy and κ and λ light variable segments (IGHV, IGKV, IGLV) [8, 9, 11–14]. In total, 50 different IGHV segments are described [8, 9, 11]. All 32 functional bovine IGHV belong to IGHV family 1 (boVH1) while boVH2 consists exclusively of pseudogene variable segments [8, 9, 11, 12, 15–19]. The ten heavy chain diversity segments (IGHD), classified into four families, revealed huge size differences from 14 to 148 nucleotides [20, 21]. Using hybridization, six heavy chain joining segments (IGHJ) were identified [22]. Only two are functional and IGHJ1 is used predominantly over IGHJ2 [18, 22, 23]. The bovine κ-light chain locus was mapped to BTA11 spanning approximately 280 kb [14], although the entire locus mapped with BAC clones extends to 412 kb [24]. In silico analysis revealed 22 IGKV with 8 functional segments, 3 IGKJ, and 1 gene for the constant region (IGKC) [14]. The bovine λ light chain locus was mapped to BTA17 [14, 25]. In total 63 IGLV segments were identified and located either on the genome or on unplaced contigs. Of these, 25 are potentially functional [14]. Four constant region genes (IGLC) are preceded by IGLJ segments. Only IGLC2 and IGLC3 are functional and IGLC3 is preferentially used [14, 24, 26–28]. A fifth IGLC might be present but its chromosomal localization has not yet been described [14]. Bovine immunoglobulin heavy chains pair mainly with light chains of the λ type [29]. In addition to classical somatic hypermu- tations, immunoglobulin λ, light chains were found to mature using a process called gene conversion where pseudogenes are used as donor sequences [27]. Most recently, Walther et al. [18] found support for the same mechanism in heavy chains. Gene conversion is known to increase diversity also in chicken and rabbit immunoglobulins [30–34]. Bovine Immunoglobulin Libraries 115

The most striking feature of bovine antibodies is a special group heavy chains possessing exceptionally long complementarity determining regions 3 (CDR3H) with up to 67 amino acid residues (aa) [8]. During junctional combination of IGHV10/34_ [8] and IGHD2_ [21] (the IGHD of 148 nucleotides) conserved short nucleotide sequences of 13 to 18 nucleotides are introduced, which are responsible for those exceptionally long CDR3H sequences [20, 21]. A high number of cysteine residues allow for disulﬁde bonding within the CDR3H. As a result, these specialized CDR3H form a stalk-knob structure uniquely found in cattle [35]. The only paired light chain variable region is IGLV1x [36, 37] which is thought to stabilize the exceptionally long CDR3Hs [35]. In cattle all known heavy chain isotype classes are present. The IGHC (heavy chain constant region) locus spans approximately 150 kb assembling the constant regions in the order μ, δ, γ3, γ1, γ2, ε, and α [23]. Phylogenetic analyses revealed a duplication of the bovine μ gene resulting in the δ gene. The ﬁrst exon of μ and δ share a homology of 95% and unlike other species the bovine δ locus has a switch region, which may permit class switch recombination [38]. The accessible genetic information about bovine immunoglobulin diversity enables the construction of immunoglobulin libraries analogue to human and mouse.

2 Materials

2.1 Construction 1. Freshly prepared lymphocytes (see Note 1). of Single-Chain 2. RNeasy Mini Kit (Qiagen) for RNA isolation. Fragment Variables 3. Invitrogen™ SuperScript™ III First-Strand Synthesis System (scFvs) (Fisher Scientiﬁc) for cDNA synthesis. 4. Chemicals for PCRs: 10 mM dNTPs (Bioline), 10Â PCR buffer (Biotools), Dimethyl sulfoxide (DMSO), DNA polymerase (Biotools), Phusion High-Fidelity DNA Polymerase, Phusion GC Buffer (both NEB). The primer sets for the individual steps are listed in Tables 1, 2, 3, 4, 5, 6, 7 (see Note 2). Primers for Glyceraldehyde 3-phosphate dehydrogenase (GAPDH): GAPDH_for TGG TCA CCA GGG CTG CT, GAPDH_rev GGA GGG GCC ATC CAC AGT CT, [8]to test the integrity of the cDNA. 5. DNA Clean & Concentrator™ (Zymo Research) for the puri- ﬁcation of PCR products (see Note 3). 6. MinElute Gel Extraction Kit (Qiagen). Ultra-pure low melting agarose for gel extraction of PCR products and cleaved phagemid (vector/plasmid). 116 Ulrike S. Diesterbeck

Table 1 Forward (P1) and reverse primer set (P2) for the amplification of bovine immunoglobulin heavy chains

Forward primer (P1) Sequence 50 –30

boVH1BACK CAG GTG CAG CTG CGS GAG TCR GG boVH2BACK MAG GTG CAG CTG CAG GAG TCR GG boVH3BACK CAG GTG CAG CTA CAG GAG TCR GG boVH4BACK CAG GTG CAG CTA GGG GAG TCR GG

Reverse primer (P2)

BoIgMCH2_FOR TGC CGT CAC CAG AGA GGC TGT BoIgDCH2_FOR TGC GTG CTG ACC GCC TTG TT BoIgG1-3CH1_FOR GGC ACC CGA GTT CCA GGT CA BoIgECH1_FOR GCC CAG CCT TAC ACG GGC TT BoIgACH1_FOR GCC AGC ACG GCA GGG AAG TT The forward primers anneal at the 50 end of the heavy chain variable region. Wobble bases are deﬁned according to the IUPAC nomenclature and marked in bold. Reverse primers anneal in the constant regions for the ampliﬁcation of different heavy chain isotypes

Table 2 Bovine primer set used for the amplification of the immunoglobulin kappa light chains in the first PCR (P3)

Forward primer (P3) Sequence 50 –30

boVκ2BACK GAT GTT GTG CTG ACC CAG ACT CC boVκ4BACK GAC ATC CAG GTG ACC CAG TCT CC

Reverse primer (P4)

boCκFOR ACA CTC RTT CTT ACT GAA GCT CTT The forward primers bind at the 50 end of the variable regions. The reverse primer (P4) anneals at the 30 end of the κ light chain constant region and is based on known alleles. Wobble bases are shown in bold

50Â TAE (tris-acetate-ethylenediaminetetraacetic acid) running buffer: 2 M Tris base, 1 M glacial acetic acid, 0.05 M EDTA (ethylenediaminetetraacetic acid), dilute with H2Oto 1Â TAE running buffer. Ethidium bromide (5 mg/ml), add 20 μl Ethidium bromide/ 100 ml heated and cooled to <60 C agarose solution. Bovine Immunoglobulin Libraries 117

Table 3 Forward primer set used for the amplification of the immunoglobulin λ light chains (P5). The forward primers bind at the 50 end of IGLV

Forward primer (P5) Sequence 5’ – 3’

boVλ1aBACK CAG GMT GTG CTG ACT CAG CCR TC boVλ1bBACK CAG GCT GTG CTG ACY CAG CCG CC boVλ2BACK CAG TCT GGC CTG ACT CAG CCT TC boVλ6aBACK TCC TAT GAA CTG ACC CAG CCG AC boVλ6bBACK TCC TAT GAA CTG ACA CAG TTG AC boVλ6cBACK TCT TCT CAG CTG ACT CAG CCG CC boVλ7BACK CAG CCT GTG CTG ACT CAG YCA GW boVλ8BACK CAG AYT GTG ATC CAG GAA CC

Reverse primer (P6)

boCλ2FOR AGA ACA CGC TGA GGT CTT CAC TGT boCλ3FOR AGR ACA CTC TGA GGG CTT CAC TGT The reverse primers (P6) anneal to the 30 end of IGLC2 and 3 in the 1st PCR, which are the functional constant regions. Wobble bases are in bold

Table 4 Bovine primer set used for the amplification of heavy chain variable regions in the second PCR

Reverse primer (P7)

Linker-boJH1FOR AGA ACC ACC TCC GCC TGA ACC GCC TCC ACC TGA GGA GAC GGT GAC CAG GAG TC Linker-boJH2FOR AGA ACC ACC TCC GCC TGA ACC GCC TCC ACC TGA GGA GAC GGT GAC CTC GAT CC The forward primer set is set P1 (see list 1). The reverse primers (P7) anneal at the 30 end of the heavy chain variable regions. The reverse primers have an overhang coding for the

N-terminal part of the (G4S)3-linker (italic). Wobble bases are shown in bold and deﬁned according to the IUPAC code

2.2 Ligation of scFvs 1. Purified scFv constructs, a phagemid (pCANTAB5E, see Note 4) Into Phagemid containing SfiIandNotI restriction sites, SfiI, NotI-HF restriction and Transformation enzymes with CutSmart Buffer (NEB), Alkaline Phosphatase, into E. coli TG1 Calf Intestinal (NEB) T4 DNA Ligase with 10Â T4 DNA Ligase buffer (NEB). DNA Clean & Concentrator™ (Zymo Research) Thermoblock with heating and cooling options. 118 Ulrike S. Diesterbeck

Table 5 Bovine primer set used for the amplification of kappa light chain variable regions in the second PCR

Forward primer (P8) Sequence 50 –30

Linker-boVκ2BACK GT TCA GGC GGA GGT GGT TCT GGC GGT GGC GGA TCG GAT GTT GTG CTG ACC CAG ACT CC Linker-boVκ4BACK GT TCA GGC GGA GGT GGT TCT GGC GGT GGC GGA TCG GAC ATC CAG GTG ACC CAG TCT CC

Reverse primer (P9)

boJκ1FOR TTT GAT CTC TAC CTT GGT TCC TTG boJκ3FOR ATT GAT TTC CAC CTT GGT CCC GCC 0 The forward primers (P8) bind at the 5 end and have an overhang coding for the C-terminal part of the (G4S)3-linker (italic). The reverse primers (P9) anneal at the 30 end of the variable regions

Table 6 Bovine primer set used for the amplification of lambda light chain variable regions in the second PCR

Forward primer (P10) Sequence 50 –30

Linker-boVλ1aBACK GT TCA GGC GGA GGT GGT TCT GGC GGT GGC GGA TCG CAG GMT GTG CTG ACT CAG CCR TC Linker-boVλ1bBACK GT TCA GGC GGA GGT GGT TCT GGC GGT GGC GGA TCG CAG GCT GTG CTG ACY CAG CCG CC Linker-boVλ2BACK GT TCA GGC GGA GGT GGT TCT GGC GGT GGC GGA TCG CAG TCT GGC CTG ACT CAG CCT TC Linker-boVλ6aBACK GT TCA GGC GGA GGT GGT TCT GGC GGT GGC GGA TCG TCC TAT GAA CTG ACC CAG CCG AC Linker-boVλ6bBACK GT TCA GGC GGA GGT GGT TCT GGC GGT GGC GGA TCG TCC TAT GAA CTG ACA CAG TTG AC Linker-boVλ6cBACK GT TCA GGC GGA GGT GGT TCT GGC GGT GGC GGA TCG TCT TCT CAG CTG ACT CAG CCG CC Linker-boVλ7BACK GT TCA GGC GGA GGT GGT TCT GGC GGT GGC GGA TCG CAG CCT GTG CTG ACT CAG YCA GW Linker-boVλ8BACK GT TCA GGC GGA GGT GGT TCT GGC GGT GGC GGA TCG CAG AYT GTG ATC CAG GAA CC

Reverse primer (P11)

boJλ2FOR CAG GAC GGT CAC TCT GGT CCC GCC boJλ3FOR CAG GAC GGT CAG TGT GGT CCC GCT 0 The forward primers (P10) bind at the 5 end and have an overhang coding for the C-terminal part of the (G4S)3-linker. The reverse primers (P11) anneal at the 30 end of the variable regions Bovine Immunoglobulin Libraries 119

Table 7 Connection of the heavy chain variable regions with the light chain variable regions and reamplification of the resulting scFv’s with primers possessing overhangs for restriction enzyme cleavage (italic)

Forward primer (P12) Sequence 50 –3’

boVH1BACKSfiI GTC CTC GCA ACT GCG GCC CAG CCG GCC ATG GCC CAG GTG CAG CTG CGS GAG TCR GG boVH2BACKSfiI GTC CTC GCA ACT GCG GCC CAG CCG GCC ATG GCC MAG GTG CAG CTG CAG GAG TCR GG boVH3BACKSfiI GTC CTC GCA ACT GCG GCC CAG CCG GCC ATG GCC CAG GTG CAG CTA CAG GAG TCR GG boVH4BACKSfiI GTC CTC GCA ACT GCG GCC CAG CCG GCC ATG GCC CAG GTG CAG CTA GGG GAG TCR GG

Reverse primer (P13)

boJκ1FORNotI GAG TCA TTC TCG ACT TGC GGC CGC TTT GAT CTC TAC CTT GGT TCC TTG boJκ3FORNotI GAG TCA TTC TCG ACT TGC GGC CGC ATT GAT TTC CAC CTT GGT CCC GCC

Reverse primer (P14)

boJλ2FORNotI GAG TCA TTC TCG ACT TGC GGC CGC CAG GAC GGT CAC TCT GGT boJλ3FORNotI GAG TCA TTC TCG ACT TGC GGC CGC CAG GAC GGT CAG TGT GGT CCC GCT Forward primers (P12) anneal within the 50 end of the heavy chain variable region. The reverse primers anneal either at the 30 end of the joining segment of the κ light chains (P13) or they anneal at the 30 end of the joining segment of the λ light chains (P14)

2. M9 minimal agar:

M9 salts solution: 12.8 g Na2HPO4, 3.0 g KH2PO4, 0.5 g NaCl, 1.0 g NH4Cl to 478 ml demineralized H2O. Stir until dissolved and autoclave to sterilize.

Agar solution: heat 15 g of agar in 500 ml demineralized H2O until dissolved and autoclave. Combine both cooled (50 C) solutions aseptically. Add ﬁlter- sterilized solutions of 20% glucose (20 ml), 1 M MgSO4 (2 ml), 1 M CaCl2 (0.1 ml), and 0.5% w/v Thiamine (0.1 ml). Dis- pense into 9 cm petri dishes and store at 4 C(see Note 5). 3. One aliquot of E. coli TG1 (genotype: K-12 supE thi-1 Δ(lac-- proAB) Δ(mcrB-hsdSM)5, (rK-mK-), plasmid: F0 [traD36 proABþ lacIq lacZΔM15]) (see Note 6). 120 Ulrike S. Diesterbeck

4. Preparation of electrocompetent cells: LB broth: 10 g tryptone, 5 g yeast, and 10 g NaCl in 1 l water. Autoclave to sterilize the solution.

Demineralized H2O, 15% glycerol solution, and liquid nitrogen. All solutions have to be ice cold (see Note 7). 5. Electroporator, 0.1 cm electroporation cuvettes (both Bio-Rad).

2.3 Packing 1. 50% glucose solution, ﬁlter-sterilized of the Unselected 100 mg/ml ampicillin in 70% ethanol, 50 mg/ml kanamycin in Library Into Phages demineralized H2O, ﬁlter-sterilized and stored at À20 C. All (Rescue) solutions, broth or solid media, which require ampicillin (see Note 8) or kanamycin, have to be cooled to <60 C. 2. 2 Â TY: 16 g tryptone, 10 g yeast, 5 g NaCl in 900 ml demineralized H2O, autoclave. After addition of additional components (see below), use autoclaved water to adjust the volume to 1 l. Plates: add 15 g agar before autoclaving. Broth or solid media, which require 2% glucose (2%G), add 4 ml of 50% glucose. If the recipe requires ampicillin (A) or kanamycin (K), add 1 ml of the respective stock solution to 1 l broth or agar: 2 Â TY-broth, 2 Â TY2%G-broth, 2 Â TY2%GA-broth/plates (9 cm petri dishes for titration, 24 Â 24 NUNC Bioassay dishes for growing the library), and 2 Â TYAK broth Incubator for E. coli on plates. Incubator with shaking function for E. coli in broth. 3. Packing the library into phages: Hyperphage (Progen) or Helperphage M13K07 (see Note 9). 4. Precipitation of phages: 20% PEG 6000, 2.5 M NaCl. PBS pH 7.4: 8 g NaCl (137 mM), 0.2 g KCl (2.7 mM), 0.24 g KH2PO4 (1.8 mM), 1.42 g Na2HPO4, anhydrous (10 mM). PBS pH 7.4 with 15% (v/v) Glycerol. High-speed centrifuge, 250 ml, 500 ml centrifugation bottles.

3 Methods

3.1 Ampliﬁcation 1. Use the appropriate amount of cells for total RNA extraction as of scFvs recommended by the manufacturer of your RNA isolation kit and measure the concentration (e.g., with a Nanodrop spectrophotometer) (see Note 10). Bovine Immunoglobulin Libraries 121

2. Convert the RNA into cDNA using random hexamer primers (see Note 11). 3. For the ampliﬁcation of the variable regions combine the following reagents for each master mix:

Final Solution or component Volume concentration

cDNA/no template (H2O) 1–2 μl n/a Forward primer (10 μM) 2.5 μl 0.4 μM Reverse primer (10 μM) 2.5 μl 0.4 μM dNTPs 1 μl 200 μM DMSO 10 μl5% 10Â PCR buffer 5 μl1Â DNA polymerase (1 U/μl) 2 μl2U

H2O Add to a total volume of 50 μl

For the ampliﬁcation of the immunoglobulins combine one primer P1 with one suitable isotype primer P2. Also, combine always one primer P3 or P5 with an equimolar mix of corresponding primers P4 and P6. The no-template control serves as an indicator of reagent purity. GAPDH ampliﬁcation tests for integrity of the cDNA template. PCR is performed in a thermocycler under following cycling conditions:

Temperature (C) Time (min)

95 5 95 1 58 1 35Â 72 2 72 10

4. Evaluate the quality of PCR products on a 1% agarose gel in 1Â TAE running buffer (see Note 12). 5. Purify the PCR products with a DNA Clean & Concentrator™ Kit and measure the concentration. Keep all products separated. 122 Ulrike S. Diesterbeck

6. During the 2nd PCR only the variable regions of heavy and light chains with overhangs for the (G4S)3 linker will be ampliﬁed in a semi-nested PCR using the puriﬁed product from the 1st PCR as template:

Final Solution or component Volume concentration

1st PCR product n/a 50 ng Forward primer (10 μM) 2.5 μl 0.4 μM Reverse primer mix (10 μM) 2.5 μl 0.4 μM dNTPs (10 mM) 1 μl 200 μM 5Â GC buffer 10 μl1Â Phusion polymerase (2 U/μl) 1 μl2U

H2O Add to a total volume of 50 μl

Use a no-template control to verify the purity of your reagents. Use corresponding single forward primers (P1, P8, or P10) and combine with reverse primers in equal molarities (P7, P9, or P11). PCR is performed under following cycling conditions:

Temperature (C) Time

98 30 s 98 30 s 60 30 s

72 30 s 35Â 72 10 min

7. Resolve PCR products 2 on a 1.5% agarose gel in 1Â TAE running buffer and isolate the product bands using the gel extraction kit (see Note 13). Measure the concentration of purified products. 8. The next step is the connection and reamplification of heavy and light chain variable regions (splicing by overlap extension (SOE)). Thus, mix the gel-purified products of the variable regions of the heavy, κ, and λ light chains in equal amounts, respectively. The first step of the splicing-by-overlap extension PCR is the connection of heavy and light chain variable regions without the addition of primers (see Note 14): Bovine Immunoglobulin Libraries 123

Final Solution or component Volume concentration

Heavy chain variable region linker n/a 300 ng mix Linker light chain variable region n/a 300 ng mix (either κ or λ) dNTPs (10 mM) 1 μl 200 μM 5Â GC buffer 10 μl1Â Phusion polymerase (2 U/μl) 1 μl2U

H2O Add to a total volume of 50 μl PCR conditions are:

Temperature (C) Time

98 30 s 98 30 s 60 30 s 35Â 72 30 s

72 10 min 9. During the reampliﬁcation of the scFvs, primers with overhangs for enzyme cleavage are used. Divide the PCR mix of the spliced variable regions into 10 aliquots:

Solution or component Volume (μl) Final concentration

First SOE PCR 5 n/a Forward primer (10 μM) 2.5 0.4 μM Reverse primer mix (10 μM) 2.5 0.4 μM dNTPs (10 mM) 1 200 μM 5Â GC buffer 10 1Â Phusion polymerase (2 U/μl) 1 2 U

H2O28 PCR conditions are the same as for the SOE 1 step except increase the ampliﬁcation cycle to 25. 10. Check the resulting scFvs on a preparative 1% agarose gel with 1Â TAE running buffer. You might see non-spliced products of variable regions. Cut the band of the size of the spliced product and purify. Measure the concentration afterward. 124 Ulrike S. Diesterbeck

3.2 Ligation of scFv 1. The advantage of using restriction enzymes like SﬁI and NotI- Genes into HF is the rare occurrence of their cleavage sites because the the Phagemid scFv genes should not be cleaved. For cleaving the insert (scFv genes: scFv κ or λ) with SﬁI combine for a 100 μl reaction volume:

Solution or Final component Volume concentration

Insert (scFv κ or λ) n/a 1 μg 10Â CutSmart buffer 10 μl1Â SﬁI (20 U/μl) 1 μl20U

H2O Add to a total volume of 100 μl

The phagemid is cleaved with:

Final Solution or component Volume concentration

Phagemid n/a 5 μg (pCANTAB5E) 10Â CutSmart buffer 10 μl1Â SﬁI (20 U/μl) 2.5 μl50U

H2O Add to a total volume of 100 μl Gently mix and brieﬂy centrifuge both samples. Incubation of the reaction occurs overnight at 50 C in a thermoblock. 2. Next day add to both sample tubes:

Solution or component Volume Final concentration

SﬁI cleavage reaction 100 μl n/a Insert (scFv κ or λ)or Phagemid (pCANTAB5E) 10Â CutSmart buffer 5 μl1Â NotI-HF (20 U/μl) 1 μl 20 U Insert (scFv κ or λ) 2.5 μl 50 U Phagemid (pCANTAB5E)

H2O 44 μl Insert (scFv κ or λ) 42.5 μl Phagemid (pCANTAB5E) Mix both samples brieﬂy and centrifuge down small droplets under the tube lid. Incubate overnight at 37 C in a thermoblock. Bovine Immunoglobulin Libraries 125

3. Inactivate NotI-HF at 65 C for 20 min. 4. Add 2.5 μl CIP to the cleaved phagemid sample and incubate for 1 h at 37 C. The CIP is inactivated at 65 C for 20 min. The alkaline phosphatase removes phosphates from the 50 end of DNA and thus, prevents self-ligation of the vector. 5. In order to remove the short DNA pieces cleaved from the insert use the DNA Clean & Concentrator™ kit. Elute with 8 μl Tris-HCl, pH 8.5. 6. The linearized vector has to be purified by preparative agarose gel extraction. The best resolution is achieved with a gel length of 30 cm. Run a thick 0.5% gel overnight in 1Â TAE running buffer at 70 V, 400 mA. Isolate the band corresponding to the cleaved vector and use the Qiagen Gel Extraction Kit for purification (see Note 15). 7. Determine the DNA concentration of insert and vector. 8. Both purified products will be ligated in an approximate 5 (insert): 1 (vector) ratio:

Final Solution or component Volume concentration

Puriﬁed linearized vector n/a 170 ng Puriﬁed scFv κ or λ n/a 330 ng 10Â T4 DNA ligase buffer 2 μl1Â T4 DNA ligase (2000 U/μl) 1 μl 2000 U

H2O Add to a total volume of 20 μl

Incubate the ligation reaction overnight at 16 C. Heat inactivate at 65 C for 10 min. Pool two reactions of the same kind and use the DNA Clean & Concentrator™ kit for removing salts and enzyme prior to electroporation. Elute in 6 μl Tris-HCl, pH 8.5 and chill on ice until transformation.

3.3 Production There are various protocols that are used for the production of of Electrocompetent electrocompetent E. coli. The following one yielded in higher E. coli transformation efﬁciencies than commercially available cells. It is crucial to work sterile because the cells do not have any antibiotic resistance and contamination can occur. 1. Plate fresh untransformed E. coli TG1 on a minimal agar plate. Incubate for 2 days at 37 C. 2. Choose one colony and transfer into 5 ml LB broth. Incubate overnight at 37 C, 250 rpm. 126 Ulrike S. Diesterbeck

3. Use the preculture to inoculate 500 ml LB broth with an OD600 nm of 0.1–0.2. 4. Incubate at 37 C, 250 rpm until an OD600nm of 0.4–0.5 is reached. 5. Cool the bacteria suspension down on ice for 1 h. 6. Centrifuge at 4000 Â g for 15 min at 4 C. Wash the pellet with 250 ml ice-cold water. 7. Repeat step 6. 8. Centrifuge again and wash the pellet with 200 ml 15% ice-cold glycerol. 9. After the ﬁnal centrifugation, remove as much of the supernatant as possible. 10. Dissolve the pellet in 15% ice-cold glycerol. 11. Prepare a 1:1000 dilution in 15% glycerol (e.g., 10 μl in 10 ml) and measure the OD600nm (see Note 16). 12. Aliquot the cells in 70 μl aliquots and shock freeze in liquid nitrogen. Store cells at À80 C. 13. Check the transformation efﬁciency with a pUC19 plasmid and examine untransformed cells for the absence of resistance on agar plates with and without antibiotics (see Note 17).

3.4 Transformation 1. Thaw an aliquot of electrocompetent E. coli, TG1 on ice and into E. coli TG1 pipet 1 μl of puriﬁed ligation product to the cells. Do not and Library Assembly mix them. 2. Transfer cells and vector into a 1 mm electroporation cuvette. Transform with electropulse of 1.8 kV, 200 Ω, and a capacitance of 25 μF. Immediately, add 1 ml pre-warmed 2Â TY2%G broth. Combine four transformations in a 15 ml reaction tube and incubate at 37 C and 250 rpm for 1 h. 3. Titrate the pooled transformation reactions in log10 steps. Add À 100 μl transformation reaction to 900 μl2Â TY2%G (10 1). À Mix well and take 100 μl into fresh 900 μl2Â TY2%G (10 2). À Dilute until 10 7. Take 100 μl (1/10 of each dilution!) and plate on 9 cm 2Â TY2%GA dishes. Take the remaining transformation reaction culture and plate on a BioAssay Dish. Incu- bate overnight at 30 C to ensure equal growing of the colonies. 4. Count the colonies and calculate the colony-forming units (cfu) per ml. 5. Scrape the colonies grown in one BioAssay Dish into 5 ml 2Â TY2%GA15%glycerol and divide into 1 ml aliquots. Store sublibraries at À80 C. Bovine Immunoglobulin Libraries 127

6. To check the quality of the libraries, transfer single colonies from the titration plates into fresh 2Â TY2%GA and prepare the plasmids. 7. Sequence the insert and check for insert size (complete scFv), diversity (combination of different genes), and open reading frame. 8. Repeat ligation and transformation until the κ and λ sublibraries reach about 108 cfu, respectively. 9. To assemble the library, thaw aliquots of the κ and λ sublibraries on ice. Calculate the overall diversity and combine respective aliquots. The library can be stored in 1 ml aliquots at À80 C. 10. To determine the colony-forming units per ml (cfu/ml), dilute 20 μl of the library in log10 serial dilutions. Plate 100 μl of each dilution onto 2Â TY2%GA. Incubate overnight at 37 C. 11. The cell density can be determined as a 1:1000 dilution at 600 nm.

3.5 Recombinant 1. In Subheading 3.4 you have measured the density and cfu of Phage Production your library. Inoculate a proportion of bacteria: library diversity “Rescue” of 10:1 into 1 l 2Â TY2%GA. Incubate at 37 C, 225 rpm until the OD600nm reaches 0.4 to 0.6. 2. Take 150 ml and transduce with Hyperphage for polyvalent display or Helperphage M13KO7 for monovalent display. Use an MOI of 30:1. 3. Incubate at 37 C ﬁrst 30 min stationary and then for 30 min at 250 rpm to allow attachment and transduction of the phages. 4. Pellet the cells with 3000 rpm, at 15 min, and 4 C. 5. Solubilize in 1 l 2 Â TYAK and incubate overnight at 30 C and 250 rpm (see Note 18). 6. Pellet the bacteria suspension at 3000 Â g for 15 min at 4 C. 7. Take the supernatant harboring the phages and add 1/5 volume 20%PEG/2.5 M NaCl for phage precipitation. Incubate for 45 min on ice with gentle shaking followed by 45 min stationary. 8. Harvest the phages by centrifugation (1 h, maximum speed, 4 C) and solubilize in 200 ml PBS. 9. Repeat steps 7 and 8. 10. Finally solubilize the phage pellet in 1 ml PBS þ 15% Glycerol. Phages can be kept at 4 C. This is your unselected phage library. 128 Ulrike S. Diesterbeck

4 Notes

1. Lymphocyte sources might be, e.g., whole mononuclear cells isolated by Ficoll gradient from peripheral blood, isolated from appropriate tissues (e.g., spleen, bone marrow), or sorted by an fluorescence-activated cell sorter (FACS). Check cell viability by trypan blue staining under a microscope. Count cells in a Fuchs-Rosenthal chamber. If required, count cells in different dilutions of your sample. 2. Bovine primer sets, used for the amplification of the different immunoglobulin heavy chain isotypes as well as the variable regions, are based on the genomic analysis [8, 11]. The primer sets for the amplification of κ and λ light chains were designed based on published sequence analyses [14, 39, 40]. Although bovine immunoglobulins are predominantly of λ isotypes, amplification of κ light chains should be performed, too. 3. This kit allows the elution of DNA in minimal volume of 6 μl. The elution buffer provided by the kit contains ethylenediaminetetraacetic acid (EDTA), which might interfere with downstream PCRs. Instead, use an elution buffer with 10 mM Tris–HCl, pH 8.5 that lacks EDTA. 4. Other phagemids with different tags work equally well if they contain the SfiI cleavage site 50 of the scFv and the NotI cleavage site 30 of the scFv. For soluble scFv expression an amber stop is required between tag and the fusion gene 3 of a filamentous phage. 5. The transformation requires electrocompetent cells. Electro- competent E. coli exhibit higher transformation competence than chemical competent E. coli. Minimal agar facilitates selection for the F-plasmid, which is required for pili formation. 6. You are working with untransformed E. coli, which do not have an antibiotic resistance yet. Therefore, absolute sterility during each step is important. 7. If no liquid nitrogen is available dry-ice with ethanol will work, too. 8. Carbenicillin is a good alternative to ampicillin because it has longer storage stability. 9. There are two options for packing the library into phages. The Hyperphage (Progen) allows multivalent display due to a defective gene for major coat protein 3 or for monovalent display use Helperphage M13K07. 10. Extract lymphocytes from your target tissue according to your protocol. Count and check for viable cells using trypan blue staining. Bovine Immunoglobulin Libraries 129

11. PolydT primers may also work but best results can be obtained with random hexamers. 12. Three bands correspond to three groups of CDR3H lengths [8]. 13. This step ensures not only the removal of primers but also of template DNA. 14. This is crucial, because primers would tend to amplify the small products and the splicing would be less efﬁcient. 15. It is critical to remove uncut vector as best as possible. It might accumulate and overgrow insert-bearing plasmids.

16. If the OD600nm is too high, add more 15% glycerol solution, and if it is too low, centrifuge again and remove more supernatant. 17. Transformation efﬁciency should be >109/μg vector for one aliquot of cells. 18. Kanamycin is crucial at this point. Kanamycin resistance is provided by the phage. Only transduced E. coli are allowed to replicate. Filamentous phages are lytic and bacteria are able to replicate, albeit slower. Phages are released into the culture supernatant.

Acknowledgments

I gratefully thank Dr. Apostolos Gittis and Dr. David Garboczi (both Structural Biology Section, Research Technologies Branch, NIAID, NIH, Maryland, USA) for critical review of the manuscript. This material should not be interpreted as representing the viewpoint of the U.S. Department of Health and Human Services, the National Institutes of Health, or the Laboratory of Viral Diseases-Genetic Engineering Section of the National Institute of Allergy and Infectious Diseases.

References

1. Frenzel A, Schirrmann T, Hust M (2016) chain constant gene region reveals evidence Phage display-derived human antibodies in for seven IgG isotypes and for IgD in the clinical development and therapy. MAbs 8 horse. J Immunol 173(5):3230–3242. (7):1177–1194. https://doi.org/10.1080/ Doi:173/5/3230 [pii] 19420862.2016.1212149 4. Navarro P, Barbis DP, Antczak D, Butler JE 2. O’Brien PM, Aitken R, O’Neil BW, Campo (1995) The complete cDNA and deduced MS (1999) Generation of native bovine mAbs amino acid sequence of equine IgE. Mol by phage display. Proc Natl Acad Sci U S A 96 Immunol 32(1):1–8 (2):640–645 5. Schrenzel MD, King DP, McKnight ML, Fer- 3. Wagner B, Miller DC, Lear TL, Antczak DF rick DA (1997) Characterization of horse (2004) The complete map of the Ig heavy (Equus Caballus) immunoglobulin mu chain- 130 Ulrike S. Diesterbeck

encoding genes. Immunogenetics 45 15. Berens SJ, Wylie DE, Lopez OJ (1997) Use of (6):386–393 a single VH family and long CDR3s in the 6. Almagro JC, Martinez L, Smith SL, Alagon A, variable region of cattle Ig heavy chains. Int Estevez J, Paniagua J (2006) Analysis of the Immunol 9(1):189–199 horse V(H) repertoire and comparison with 16. Sinclair MC, Gilchrist J, Aitken R (1997) the human IGHV germline genes, and sheep, Bovine IgG repertoire is dominated by a single cattle and pig V(H) sequences. Mol Immunol diversified VH gene family. J Immunol 159 43(11):1836–1845. Doi:S0161-5890(05) (8):3883–3889 00374-3 [pii]. https://doi.org/10.1016/j. 17. Lopez O, Perez C, Wylie D (1998) A single Molimm.2005.10.017 VH family and long CDR3s are the targets for 7. Sun Y, Wang C, Wang Y, Zhang T, Ren L, hypermutation in bovine immunoglobulin Hu X, Zhang R, Meng Q, Guo Y, Fei J, Li N, heavy chains. Immunol rev 162:55–66 Zhao Y (2010) A comprehensive analysis of 18. Walther S, Tietze M, Czerny CP, Konig S, germline and expressed immunoglobulin rep- Diesterbeck US (2016) Development of a bio- ertoire in the horse. Dev comp Immunol 34 informatics framework for the detection of (9):1009–1020. Doi:S0145-305X(10)00103- gene conversion and the analysis of combinato- 5 [pii]. https://doi.org/10.1016/j.dci.2010. rial diversity in immunoglobulin heavy chains 05.003 in four cattle breeds. PLoS one 11(11): 8. Walther S, Czerny CP, Diesterbeck US (2013) e0164567. https://doi.org/10.1371/journal. Exceptionally long CDR3H are not isotype Pone.0164567 restricted in bovine immunoglobulins. PLoS 19. Tutter A, Riblet R (1989) Conservation of an one 8(5):e64234. https://doi.org/10.1371/ immunoglobulin variable-region gene family journal.Pone.0064234 indicates a specific, noncoding function. Proc 9. Niku M, Liljavirta J, Durkin K, Schroderus E, Natl Acad Sci U S a 86(19):7460–7464 Iivanainen A (2012) The bovine genomic 20. Koti M, Kataeva G, Kaushik AK (2008) Orga- DNA sequence data reveal three IGHV sub- nization of D(H)-gene locus is distinct in cat- groups, only one of which is functionally tle. Dev biol (Basel) 132:307–313 expressed. Dev comp Immunol 37 21. Koti M, Kataeva G, Kaushik AK (2010) Novel (3–4):457–461. https://doi.org/10.1016/j. atypical nucleotide insertions specifically at dci.2012.02.006 VH-DH junction generate exceptionally long 10. Miller JR, Thomsen PD, Dixon SC, Tucker CDR3H in cattle antibodies. Mol Immunol 47 EM, Konfortov BA, Harbitz I (1992) Synteny (11–12):2119–2128. Doi:S0161-5890(10) mapping of the bovine IGHG2, CRC and 00064-7 [pii]. https://doi.org/10.1016/j. IGF1 genes. Anim genet 23(1):51–58 Molimm.2010.02.014 11. Liljavirta J, Niku M, Pessa-Morikawa T, 22. Hosseini A, Campbell G, Prorocic M, Aitken R Ekman A, Iivanainen A (2014) Expansion of (2004) Duplicated copies of the bovine JH the preimmune antibody repertoire by junc- locus contribute to the Ig repertoire. Int tional diversity in Bos Taurus. PLoS one 9(6): Immunol 16(6):843–852. https://doi.org/ e99808. https://doi.org/10.1371/journal. 10.1093/intimm/dxh085. dxh085 [pii] Pone.0099808 23. Zhao Y, Kacskovics I, Rabbani H, Hammar- 12. Saini SS, Hein WR, Kaushik A (1997) A single strom L (2003) Physical mapping of the bovine predominantly expressed polymorphic immu- immunoglobulin heavy chain constant region noglobulin VH gene family, related to mam- gene locus. J biol Chem 278 malian group, I, clan, II, is identified in cattle. (37):35024–35032. https://doi.org/10. Mol Immunol 34(8–9):641–651 1074/jbc.M301337200. M301337200 [pii] 13. Zhao Y, Jackson SM, Aitken R (2006) The 24. Pasman Y, Saini SS, Smith E, Kaushik AK bovine antibody repertoire. Dev comp Immu- (2010) Organization and genomic complexity nol 30(1–2):175–186. Doi:S0145-305X(05) of bovine lambda-light chain gene locus. Vet 00148-5 [pii]. https://doi.org/10.1016/j. Immunol Immunopathol 135(3–4):306–313. dci.2005.06.012 Doi:S0165-2427(10)00006-1 [pii]. https:// 14. Ekman A, Niku M, Liljavirta J, Iivanainen A doi.org/10.1016/j.Vetimm.2009.12.012 (2009) Bos Taurus genome sequence reveals 25. Tobin-Janzen TC, Womack JE (1992) Com- the assortment of immunoglobulin and surro- parative mapping of IGHG1, IGHM, FES, and gate light chain genes in domestic cattle. BMC FOS in domestic cattle. Immunogenetics 36 Immunol 10:22. Doi:1471-2172-10-22 [pii]. (3):157–165 https://doi.org/10.1186/1471-2172-10-22 Bovine Immunoglobulin Libraries 131

26. Parng CL, Hansal S, Goldsby RA, Osborne BA light chain gene is limited to the rearranged (1995) Diversification of bovine lambda-light variable gene segment. Cell 48(3):369–378. chain genes. Ann N Y Acad Sci 764:155–157 Doi:0092-8674(87)90188-7 [pii] 27. Parng CL, Hansal S, Goldsby RA, Osborne BA 35. Wang F, Ekiert DC, Ahmad I, Yu W, Zhang Y, (1996) Gene conversion contributes to Ig light Bazirgan O, Torkamani A, Raudsepp T, chain diversity in cattle. J Immunol 157 Mwangi W, Criscitiello MF, Wilson IA, Schultz (12):5478–5486 PG, Smider VV (2013) Reshaping antibody 28. Chen L, Li M, Li Q, Yang X, An X, Chen Y diversity. Cell 153(6):1379–1393. https:// (2008) Characterization of the bovine immu- doi.org/10.1016/j.Cell.2013.04.049 noglobulin lambda light chain constant IGLC 36. Saini SS, Allore B, Jacobs RM, Kaushik A genes. Vet Immunol Immunopathol 124 (1999) Exceptionally long CDR3H region (3–4):284–294 with multiple cysteine residues in functional 29. Arun SS, Breuer W, Hermanns W (1996) bovine IgM antibodies. Eur J Immunol 29 Immunohistochemical examination of light- (8):2420–2426. Doi:10.1002/(SICI)1521- chain expression (lambda/kappa ratio) in 4141(199908)29:08<2420::AID-IMMU242 canine, feline, equine, bovine and porcine 0>3.0.CO;2-A [pii]. 10.1002/(SICI)1521- plasma cells. Zentralbl Veterinarmed a 43 4141(199908)29:08<2420::AID- (9):573–576 IMMU2420>3.0.CO;2-A 30. Reynaud CA, Dahan A, Anquez V, Weill JC 37. Saini SS, Farrugia W, Ramsland PA, Kaushik (1989) Somatic hyperconversion diversifies AK (2003) Bovine IgM antibodies with excep- the single Vh gene of the chicken with a high tionally long complementarity-determining incidence in the D region. Cell 59 region 3 of the heavy chain share unique struc- (1):171–183. Doi:0092-8674(89)90879-9 tural properties conferring restricted VH þ [pii] Vlambda pairings. Int Immunol 15 31. Reynaud CA, Anquez V, Grimal H, Weill JC (7):845–853 (1987) A hyperconversion mechanism gener- 38. Zhao Y, Kacskovics I, Pan Q, Liberles DA, ates the chicken light chain preimmune reper- Geli J, Davis SK, Rabbani H, Hammarstrom toire. Cell 48(3):379–388. Doi:0092-8674 L (2002) Artiodactyl IgD: the missing link. J (87)90189-9 [pii] Immunol 169(8):4408–4416 32. Becker RS, Knight KL (1990) Somatic diversi- 39. Stein SK, Diesterbeck US, Aboelhassan DM, fication of immunoglobulin heavy chain VDJ Czerny CP (2012) Comparison of joining and genes: evidence for somatic gene conversion in constant kappa-light chain regions in different rabbits. Cell 63(5):987–997. Doi:0092-8674 cattle breeds. Anim Genet 43(6):776–780. (90)90502-6 [pii] https://doi.org/10.1111/j.1365-2052.2012. 33. Reynaud CA, Bertocci B, Dahan A, Weill JC 02324.x (1994) Formation of the chicken B-cell reper- 40. Diesterbeck US, Aboelhassan DM, Stein SK, toire: ontogenesis, regulation of Ig gene rear- Czerny CP (2012) Detection of new allotypic rangement, and diversification by gene variants of bovine lambda-light chain constant conversion. Adv Immunol 57:353–378 regions in different cattle breeds. Dev Comp 34. Thompson CB, Neiman PE (1987) Somatic Immunol 36(1):130–139. https://doi.org/ diversification of the chicken immunoglobulin 10.1016/j.dci.2011.06.011 Chapter 7

Construction of Rabbit Immune Antibody Libraries

Thi Thu Ha Nguyen, Jong Seo Lee, and Hyunbo Shim

Abstract

Rabbits have distinct advantages over mice as a source of target-specific antibodies. They produce higher affinity antibodies than mice, and may elicit strong immune response against antigens or epitopes that are poorly immunogenic or tolerated in mice. However, a great majority of currently available monoclonal antibodies are of murine origin because of the wider availability of murine fusion partner cell lines and well- established tools and protocols for fusion and cloning of mouse hybridoma. Phage-display selection of antibody libraries is an alternative method to hybridoma technology for the generation of target-specific monoclonal antibodies. High-affinity monoclonal antibodies from nonmurine species can readily be obtained by constructing immune antibody libraries from B cells of the immunized animal and screening the library by phage display. In this article, we describe the construction of a rabbit immune Fab library for the facile isolation of rabbit monoclonal antibodies. After immunization, B-cell cDNA is obtained from the spleen of the animal, from which antibody variable domain repertoires are amplified and assembled into a Fab repertoire by PCR. The Fab genes are then cloned into a phagemid vector and transformed to E. coli, from which a phage-displayed immune Fab library is rescued. Such a library can be biopanned against the immunization antigen for rapid identification of high-affinity, target-specific rabbit monoclonal antibodies.

Key words Rabbit antibody, Antibody library, Phage display, Immune antibody library, Fab library, Rabbit monoclonal antibody

1 Introduction

Traditionally, most monoclonal antibodies have been generated by immunizing mice and preparing hybridomas [1]. Murine myeloma cell lines as fusion partners, as well as the protocols for fusion, cloning, and screening, have been well established. On the other hand, the generation of non-murine monoclonal antibodies by hybridoma technology has been hampered by the lack of suitable fusion partner cell lines and/or difﬁculty in the growth and cloning of the fused cells [2, 3]. Although some fusion partner cell lines and monoclonal antibodies are available from animals such as rats and rabbits [4–7], the technology is less well established than mouse hybridoma and/or not widely accessible. Because these non-murine animals are considerably larger than mice and thus

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_7, © Springer Science+Business Media LLC 2018 133 134 Thi Thu Ha Nguyen et al.

have a greater number and diversity of B-cells, they may elicit stronger humoral immune response and produce higher-affinity antibodies against wider range of antigens or epitopes [4, 8]. There- fore, alternative technologies to hybridoma that allow the production of monoclonal antibodies from non-murine sources are highly desirable. One of such technologies is antibody phage display, which enables rapid in vitro selection of target-specific antibody clones from a large pool of antibody fragments displayed on the surface of bacteriophage [9]. Through physical linkage of the antibody fragment and its gene on the same bacteriophage particle, target- binding antibody clones can be rapidly selected, amplified, and enriched from an antibody library by iterative rounds of biopanning. For successful antibody generation by phage display, it is important to prepare and use a large, high-quality antibody library. Antibodies can be isolated from either naive (unbiased) antibody libraries [9] or immune antibody libraries [10, 11]. The former can be constructed synthetically or from the B-cell repertoire of unim- munized animals, and need to be highly diverse (>109 independent clones) in order to yield specific binders against a target antigen. The latter are made from the B-cells of immunized animals, and because they are already enriched with binders produced by humoral immune response, their size may be much smaller; electroporation of E. coli routinely yields >107 transformants, which is sufficient for most immune library construction. When an immune library is constructed from immunized laboratory animals such as mice or rabbits, the animals are sacrificed after the completion of immunization schedule, and their spleens are removed. Spleen contains a large number of B cells and plasma cells, and hence is a good source of diverse antibody genes including those encoding target-specific antibodies. In this article, a detailed protocol for the construction of rabbit immune Fab library is provided. The PCR amplification and assembly protocols are based on Andris-Widhopf et al. [12], with modifications in primer sequences. Because the Fab format generally conserves the binding activity of the immunoglobulin antibodies better than scFv [13, 14], it may be the preferable format for immune library construction, although target-specific monoclonal antibodies have also been successfully isolated from immune scFv libraries. After the construction, the phage-displayed antibody library can be panned against the immunization antigen for the rapid isolation of multiple, high-affinity monoclonal antibody fragments. Construction of Rabbit Immune Antibody Libraries 135

2 Materials

Reagents and equipment suggested here can be substituted with equivalent products from different vendors. If not listed below, standard molecular biology laboratory equipment and molecular biology grade chemicals/reagents can be used. 1. Freund’s adjuvant: complete and incomplete (Sigma). 2. Maxi H minus First Strand cDNA synthesis kit # K1651 (Thermo Scientific). 3. Nuclease-free water. 4. Oligonucleotide primers for polymerase chain reaction: see Table 1. 5. DNA polymerase: Taq polymerase (New England Biolabs, Ips- wich, MA. USA) with vendor-provided reaction buffer. 6. dNTP mixture (10 mM each of four dNTPs, New England Biolabs). 7. T4 DNA ligase (New England Biolabs). 8. pCom3X-TT phagemid vector (Addgene, Cambridge, MA, USA. Plasmid #63891). 9. SfiI restriction endonuclease (New England Biolabs). 10. ER2537 electrocompetent cells (or other E.coli strains harboring F factor and supE44 mutation such as TG1 or XL1-Blue): prepared according to [15]. 11. SB medium: Dissolve 20 g of Yeast Extract, 30 g of Trypton, and 10 g of MOPS (3-(N-morpholino)propanesulfonic acid) in 1 L deionized water. Adjust pH to 7.2 and autoclave. Store at room temperature. 12. SOC medium: Dissolve 20 g of Trypton, 5 g of Yeast extract, and 0.5 g of NaCl in 950 mL of deionized water. Add 10 mL of 250 mM KCl. Adjust pH to 7.0 and add water to 1 L. Sterilize by autoclave, allow to cool down to 60 C or less, then add 10 mL of sterile 1 M MgCl2 and 20 mL of filter-sterilized 1 M glucose. 13. Agarose electrophoresis gel: for 1% agarose gel, use 1 g of agarose and 100 mL of TAE buffer (40 mM Tris, 20 mM acetic acid, and 1 mM EDTA, pH 8.0). 14. QIAquick Gel Extraction Kit (QIAGEN, Hilden, Germany). 15. LB-ampicillin glucose (LBAG) plate: dissolve 10 g of Trypton, 5 g of Yeast extract, and 10 g of NaCl in 1 L deionized water. Add 18 g of bacteriological agar and autoclave. After the autoclaved solution cools down to 45–60 C, add 50 mL of 40% (w/v) filter-sterilized glucose and 1 mL of filter-sterilized Ampicillin (100 mg/mL). Mix gently and pour on 100 mm 136 Thi Thu Ha Nguyen et al.

Table 1 List of primers used in this protocol

Name Sequences

HIgCH1-f GCCTCCACCAAGGGCCCA dpseq AGAAGCGTAGTCCGGAACG HKC-f ACTGTG GCT GCA CCA TCT G Lead-B GGCCATGGCTGGTTGGGC RVH1 GCCCAACCAGCCATG GCCCAG GAGCAGCTGAAGGAG RVH2 GCCCAACCAGCCATG GCCCAGGAGCAG CTG RTG GAG RVH3 GCCCAACCAGCCATGGCCCAGGAGCAGCTGGAGGAGTCC RVH4 GCCCAACCAGCCATGGCCCAGTCGSTGGAGGAGTCC RVH5 GCCCAACCAGCCATGGCCCAGTCGGTGAAGGAGTCC RVH6 GCCCAACCAGCCATGGCCCAGCAGCTGGAGCAGTCC RJH-b TGGGCCCTTGGTGGAGGCTGARGAGAYGGTGACCAGGGT RVK1 TAATTGGCCCAGGCGGCCGACCCTATGCTGACCCAG RVK2 TAATTGGCCCAGGCGGCCGATGTCGTGATGACCCAG RVK3 TAATTGGCCCAGGCGGCCGCAGCCGTGCTGACCCAG RVK4 TAATTGGCCCAGGCGGCCGCCATCGATATGACCCAG RVK5 TAATTGGCCCAGGCGGCCGCCCAAGTGCTGACCCAG RVK6 TAATTGGCCCAGGCGGCCGCCCTTGTGATGACCCAG RVK7 TAATTGGCCCAGGCGGCCGCTCAAGTGCTGACCCAG RVK8 TAATTGGCCCAGGCGGCCTATGTCATGATGACCCAG RJK1-b AGATGGTGCAGCCACAGTTCGTTTGATTTCCACATTGGT RJK2-b AGATGGTGCAGCCACAGTTCGTTYGACSACCACCTYGGT RJK3-b AGATGGTGCAGCCACAGTTCGTAGGATCTCCAGCTCGGT RJK4-b AGATGGTGCAGCCACAGTTCGTTTGATYTCCASCTTGGT RVL1 TAATTGGCCCAGGCGGCCCAGCCTGCCCTCACTCAG RVL2 TAATTGGCCCAGGCGGCCTCCTATGAGCTGACACAG RVL3 TAATTGGCCCAGGCGGCCTCCTTCGTGCTGACTCAG RVL4 TAATTGGCCCAGGCGGCCCAGCCTGTGCTGACTCAG RVL5 TAATTGGCCCAGGCGGCCAGCGTTGTGTTCACGCAG RVL6 TAATTGGCCCAGGCGGCCCAGTTTGTGCTGACTCAG RJL-b AGATGGTGCAGCCACAGTTCGGCCTGTGACGGTCAGCTGGGT Lead-VH GCC CAA CCA GCC ATG GCC (continued) Construction of Rabbit Immune Antibody Libraries 137

Table 1 (continued)

Name Sequences

RSC-SF TAATTGGCCCAGGCGGCC pC3X-f GCACGACAGGTTTCCCGAC pC3X-b AACCATCGATAGCAGCACCG pelseq ACCTATTGCCTACGGCAGCCG

and 150 mm diameter polystyrene Petri dishes (10 and 20 mL, respectively, per dish). Keep the plates at room temperature until agar solidiﬁes, and store at 4 C. 16. Electroporation cuvette: 2 mm gap (Bio-Rad, Hercules, CA, USA). 17. Electroporator (Micropulser™, Bio-Rab). 18. VCSM13 helper phage (Agilent Technologies, Santa Clara, CA, USA). Helper phage can be prepared according to [15]. 19. Phosphate-buffered saline (PBS): Dissolve 80 g NaCl, 2.0 g KCl, 17 g Na2HPO4, and 1.63 g KH2PO4 in 0.95 L deionized water. Set pH to 7.4 with HCl, and add water to 1 L. 20. 5Â PEG precipitation buffer: 20% (w/v) polyethyleneglycol- 8000, 15% (w/v) sodium chloride in deionized water. 21. Protease inhibitor cocktail: cOmplete™ EDTA-free (Roche).

3 Method

3.1 Rabbit 1. Prepare antigen (e.g., 350 μg of protein or 500 μg of keyhole Immunization limpet hemocyanin [KLH]-conjugated peptide/hapten) (see Note 1) in 500 μL of sterile PBS, and mix with 500 μLof Freund’s complete adjuvant, for a total of 1 mL antigen mixture per rabbit. 2. Subcutaneously inject the antigen mixture to New Zealand White rabbits (see Note 2). It is recommended that two rabbits are immunized in order to improve the probability of success of antibody production. 3. Four weeks after the initial immunization, perform boost immunization. Subcutaneously inject the antigen (300 μgof protein or 400 μg of KLH-conjugated peptide/hapten) mixed 1:1 with Freund’s incomplete adjuvant (1 mL/animal). 4. One week after the ﬁrst boost, take 1 mL blood from central auricular artery. Test the serum for the antigen-binding activity by indirect ELISA [16]. 138 Thi Thu Ha Nguyen et al.

5. Repeat two additional boost immunizations at two-week interval. Perform bleeding and ELISA test as above. 6. One week after the ﬁnal (third) boost, collect the whole blood (>50 mL) by cardiac puncture (see Note 2). Place the blood stationary for 1 h at room temperature, centrifuge (500 Â g, 15 min), and store the serum at À20 C for future use. 7. Surgically extract spleen from the rabbit, and proceed directly to Subheading 3.2. Alternatively, spleens can be frozen with liquid nitrogen and stored at À80 C, or kept submerged in ® RNA stabilization solution (e.g., RNAlater ) until use.

3.2 Preparation 1. Put the spleen removed from the immunized rabbit in a 50 mL of Total RNA from conical polypropylene tube, and add 1 mL of TRI reagent per the Spleens 100 mg of tissue. Homogenize the spleens using a tissue of Immunized Rabbits homogenizer. 2. Allow the homogenized sample to stand for 5 min at room temperature. Centrifuge at 3000 Â g for 15 min, and transfer the supernatant to a fresh tube. 3. Add 0.1 mL of 1-bromo-3-chloropropane per mL of TRI reagent used, and shake vigorously for 20 s. Allow the sample to stand for 5 min at room temperature. 4. Centrifuge the sample at 12,000 Â g for 15 min at 4 C. The mixture separates into three phases, of which the aqueous top layer contains RNA. 5. Carefully transfer only the top layer to a new centrifuge tube, and precipitate RNA by adding 0.5 mL of 2-propanol per 1 mL of TRI reagent used. Mix well and allow the mixture to stand for 5 min at room temperature. 6. Centrifuge the sample at 12,000 Â g for 10 min at 4 C. Carefully remove the supernatant and wash the RNA pellet with ice-cold 75% ethanol (1 mL per mL of TRI reagent used). Vortex the tube brieﬂy, and centrifuge at 12,000 Â g for 10 min at 4 C. 7. Remove the supernatant and air-dry the pellet for about 5 min, taking care not to allow it to over-dry. Dissolve the pellet in 0.5 mL of nuclease-free water. The RNA solution can be stored at À80 C for several weeks.

3.3 Synthesis 1. To 5 μg RNA from Subheading 3.2, add 1 μL of 100 μM oligo of First-Strand cDNA (dT)18 primer, 1 μL of 10 mM dNTP mix, 4 μLof5Â RT buffer, and 1 μL Maxima H minus enzyme mix. Add nuclease- free water to a ﬁnal volume of 20 μL. 2. Incubate the reaction mixture for 30 min at 50 C and terminated the reaction by heating the mixture at 85 C for 5 min. Construction of Rabbit Immune Antibody Libraries 139

3.4 Assembly VH and VL repertoires were ampliﬁed from the rabbit spleen of Rabbit/Human cDNA, and assembled with the human constant domains (Cκ and Chimeric Fab CH1) into a Fab library in three steps. Repertoire by PCR

3.4.1 First Round of PCR In the first round of PCR, individual domains (variable and constant) are amplified. Variable domains are amplified from the cDNA, and constant domains are amplified from the pComb3X- TT vector harboring genes for human Fab. Perform a separate reaction for each primer pair to maximally retain the diversity of the immunized repertoire. 1. For the amplification of rabbit VH and VL repertoires, add the following to nuclease-free water to make final volume of 100 μL in a PCR tube on ice: 1–3 μL cDNA (about 0.5 μg), dNTP mixture (0.2 mM final concentration of each dNTP), 10 μLof10Â standard Taq buffer, 0.5 μL (2.5 units) of Taq polymerase, and 0.6 μM of forward and reverse primers. Primer sequences are shown in Table 1 and primer pairs for the amplification of rabbit VH and VL are shown in Table 2. 2. Perform PCR with the following thermal cycle: initial melting at 94 C for 2 min; 30 cycles of 94 for 30 s, 56 C for 30 s, and 72 C for 30 s; final extension at 72 C for 7 min. 3. Electrophorese PCR products on 1% agarose gel with ethidium bromide, and inspect the gel under UV light. Excise gel bands for VH and VL at about 350 bp lengths and extract DNA from the excised band using the gel extraction kit, according to the manufacturer’s protocol. 4. Add the following to nuclease-free water to make final volume of 100 μL in PCR tube on ice: 1 μg of pComb3X-TT vector DNA, dNTP mixture (0.2 mM final concentration of each dNTP), 10 μLof10Â standard Taq buffer, 0.5 μL (2.5 units) of Taq polymerase, and 0.6 μM of forward and reverse primers. Sequences of primers are shown in Table 1 and primer pairs for the amplification of human CH1, CK are shown in Table 2. 5. Perform PCR and purify the amplified DNA by agarose gel extraction using the same protocol as step 3.

3.4.2 Second Round In the second round of PCR, Fd fragment (VH-CH1) and the light of PCR chain (LC; VL-CL) are generated by an overlap extension assembly of the rabbit variable domains and human constant domains from Subheading 3.4.1. VH, Vλ, and Vκ genes separately ampliﬁed in Subheading 3.4.1 can be combined at this stage, and only three PCRs (one each for VH, Vλ, and Vκ repertoires) are performed. 1. Use the following PCR mixture: 500 ng each of template DNA, dNTP mixture (0.2 mM ﬁnal concentration of each dNTP), 140 Thi Thu Ha Nguyen et al.

Table 2 Primer pairs for the amplification of VH, Vλ,Vκ, CH1, Cκ

Variable domain Forward primer Backward primer

VH1 RVH1 RJH1 VH2 RVH2 VH3 RVH3 VH4 RVH4 VH5 RVH5 VH6 RVH6 VK1 RVK1 RJK1-b RJK2-b RJK3-b RJK4-b VK2 RVK2 RJK1-b RJK2-b RJK3-b RJK4-b VK3 RVK3 RJK1-b RJK2-b RJK3-b RJK4-b VK4 RVK4 RJK1-b RJK2-b RJK3-b RJK4-b VK5 RVK5 RJK1-b RJK2-b RJK3-b RJK4-b VK6 RVK6 RJK1-b RJK2-b RJK3-b RJK4-b VK7 RVK7 RJK1-b RJK2-b RJK3-b RJK4-b VK8 RVK8 RJK1-b RJK2-b RJK3-b RJK4-b (continued) Construction of Rabbit Immune Antibody Libraries 141

Table 2 (continued)

Variable domain Forward primer Backward primer

VL1 RVL1 RJL-b VL2 RVL2 VL3 RVL3 VL4 RVL4 VL5 RVL5 VL6 RVL6 Human CH1 HIgCH1-f dpseq Human Cκ HKC-f Leab-b

Table 3 PCR scheme for the amplification of Fd chains and light chains by overlap extension PCR

Template

Template 1 (pooled) Template 2 Forward primer Reverse primer Product name

VH1-VH6 CH1 Lead-VH dpseq Fd VK1-VK8 Cκ RSC-SF Lead-B LC (kappa) VL1-VL6 Cκ RSC-SF Lead-B LC (lambda)

10 μLof10Â standard Taq buffer, 0.5 μL (2.5 units) of Taq polymerase, 0.6 μM of forward and reverse primers, and nuclease-free water to bring the ﬁnal reaction volume to 100 μL. The primer sequences are shown in Table 1, and the primer pairs for the overlap-extension PCR are shown in Table 3. 2. Perform the PCR under the following thermal cycles: initial melting at 94 C for 2 min; 25 cycles of 94 C for 30 s, 56 C for 30 s, and 72 C for 1 min; ﬁnal extension at 72 C for 7 min. 3. Isolate the PCR products by 1% agarose gel electrophoresis as described above. The expected lengths for LC and Fd PCR products are around 750 bp.

3.4.3 Third Round of PCR In the third round of PCR, the chimeric light chain products (kappa and lambda class) and the heavy chain (Fd) fragment are joined by second overlap extension PCR. Kappa and lambda LCs can be combined before this PCR. 1. Use PCR mixture: 500 ng each of template DNA, dNTP mixture (0.2 mM ﬁnal concentration of each dNTP), 10 μL of 10Â standard Taq buffer, 0.5 μL (2.5 units) of Taq 142 Thi Thu Ha Nguyen et al.

polymerase, 0.6 μM of RSC-SF (forward) and dpseq (reverse) primers, and nuclease-free water to bring the final reaction volume to 100 μL. Perform eight 100 μL reactions in parallel. The primer sequences are provided in Table 1. 2. Perform the PCR under the following thermal cycle: initial melting at 94 C for 2 min; 25 cycles of 94 C for 30 s, 56 C for 30 s, and 72 C for 1.5 min; final extension at 72 C for 7 min. 3. Pool the PCR products (800 μL), and add 0.1 volume (80 μL) of 3 M sodium acetate (pH 5.2) and 2.5 volume (2 mL) of ethanol solution (70% final ethanol concentration). Mix well and incubate at À20 C for over 1 h. Centrifuge the mixture at 14,000 Â g for 15 min, and wash the DNA pellet three times with ice-cold 70% ethanol. Air-dry the pellet briefly, and dissolve the DNA in 50 μL of nuclease-free water (see Note 3). Run the DNA on 1% agarose gel and extract Fab DNA from ~1500 bp band as described in Subheading 3.4.1, step 3.

3.5 Construction 1. Digest the purified PCR product and pComb3X-TT phagemid of Fab Library vector with SfiI restriction enzyme. Incubate 9 μg of DNA and 40 units of SfiI in 50 μL reaction volume at 50 C for 16 h (see Note 4). Separate DNA by electrophoresis and extract DNA bands (~1500 bp for Fab and ~3500 bp for the vector) from 1% agarose gel using the DNA gel extraction kit. 2. Mix 1 μg of the digested PCR product, 1.5 μg of the digested vector, 5 μLof10Â T4 ligase buffer, 2.5 μL of T4 DNA ligase (1000 units), and nuclease-free water to 50 μL final volume. Incubate the ligation reaction mixture overnight at room temperature. 3. Next morning, precipitate the ligated DNA as described in Subheading 3.4.3, step 3, and dissolve the pellet in 20 μLof 10% glycerol solution. 4. Mix the ligated DNA with 300 μL of ER2537 competent cells and add the mixture into an electroporation cuvette. Incubate the cuvette on ice for 1–2 min, and transform the bacteria by electroporation (a single 2.50 kV pulse). 5. After the electroporation, immediately add 1 mL of warm (37 C) SOC medium to the cuvette and resuspend the cells by pipetting up and down several times. Repeat the procedure twice again, and combine the bacterial suspensions (3 mL). Incubate the transformed cells at 37 C for 1 h with shaking at 200 rpm. À 6. For transformation titration, plate 10 μL and 100 μLof10 3 dilutions of the transformed cells on LBAG agar plates (see Note 5). Centrifuge the remaining cells (3500 Â g, 15 min), and resuspend the pellet in 200 μL SOC medium. Plate the Construction of Rabbit Immune Antibody Libraries 143

resuspended cell on a 150 mm diameter LBAG agar plate and incubate overnight at 37 C. 7. Next morning, add 5 mL of the SB medium to the agar plates, and scrape the bacterial growth using a ﬂame-sterilized glass spreader. Add 0.5 volume of sterile 50% glycerol to the collected bacteria, mix well, and snap-freeze several 1 mL aliquots in liquid nitrogen. Store the frozen stocks at À80 Corina liquid nitrogen tank.

3.6 Phage Antibody Phage-displayed antibody libraries can be rescued from the trans- Library Rescue formed E. coli by superinfection of the bacteria with helper phage. Helper phage is a derivative of M13 bacteriophage with a defective phage origin of replication that makes its production and packaging much slower than the phagemid DNA. As a result, when superinfected, the phage proteins inside their host E. coli preferentially package the phagemid DNA, ensuring the proper linkage of genotype with phenotype. Helper phages also have an antibiotic resistance gene for, e.g., kanamycin, for the selection of superinfected bacteria. After helper phage superinfection, E. coli cells secrete antibody-displaying phage particles to culture medium. These phages can then be precipitated, resuspended in a small volume, and kept frozen until use. Biopanning protocols for the identiﬁca- tion of target-speciﬁc clones are described elsewhere [15, 17]. 1. Thaw one 1 mL aliquot from Subheading 3.5, step 7, and add to 400 mL of SB medium supplemented with 100 μg/mL ampicillin and 2% (w/v) glucose (see Note 6). Grow at 37 C with shaking at 200 rpm for 2–3 h, until OD600 reaches ~0.7. 2. Centrifuge the culture (3500 Â g, 15 min, 4 C), and resuspend the pellet in 400 mL of SB medium with 100 μg/mL ampicillin and without glucose. Add VCSM13 helper phage (1012 pfu) (see Note 7), and superinfect the bacteria at 37 C for 1 h with slow shaking (80 rpm). Add kanamycin to 70 μg/ mL and incubate overnight at 30 C with shaking at 200 rpm. 3. Next morning, centrifuge the culture (3500 Â g, 15 min, 4 C). Transfer the supernatant to a clean centrifugation bottle, and add and dissolve 4% (w/v) PEG-8000 and 3% (w/v) NaCl. Keep the bottle in ice for >30 min to precipitate the phage. 4. Centrifuge the precipitated phage (3500 Â g, 15 min, 4 C). Remove the supernatant, and dissolve the phage pellet in 10 mL of PBS. 5. Centrifuge again (12,000 Â g, 20 min, 4 C) to remove insoluble cell debris, and transfer to a clean centrifugation tube. Add 0.25 volume of 5Â PEG precipitation buffer, mix well, and keep the mixture in ice for >30 min. 144 Thi Thu Ha Nguyen et al.

6. Centrifuge the precipitated phage (12,000 Â g, 20 min, 4 C), remove the supernatant, and dissolve the phage pellet in 2 mL PBS with protease inhibitor cocktail. The resulting phage solution is highly viscous. 7. Add 0.5 volume of 50% (v/v) glycerol, and mix well. Keep the mixture at 4 C. À À 8. To measure phage titer, make 10 7 and 10 8 dilutions of the phage in SB medium, add 1 μL of the diluted phage to 50 μLof mid-log phase ER2537 E. coli cells and incubate for 30 min, and plate the infected bacteria on an LBAG plate. Incubate the plate overnight at 37 C, and count the number of colonies next day to calculate the phage titer (see Note 9). 9. Aliquot 1012–1013 cfu of phage glycerol stock from step 7 into microcentrifuge tubes, and keep the aliquots frozen at À80 C until use. Phage antibody library can be stored frozen for at least several months without affecting the antibody isolation capacity.

4 Notes

1. Peptides are typically conjugated to KLH through amino, car- boxylate, or thiol groups [18]. Conjugation site and chemistry need to be carefully chosen because linkage via a side chain within a desired epitope may interfere with antibody binding. For sequences that do not contain cysteines, thiol conjugation using maleimide chemistry may give best results. For haptens, functional groups may need to be introduced synthetically in order to enable the conjugation to a carrier protein away from the antigenic determinant. 2. For immunization, use 0.6 mL intramuscular injection of 1:1 mixture of Zoletil™ and Rompun™ for anesthesia. For terminal cardiac puncture, use 1 mL of the same mixture intramus- cularly. Wait 15 min post-injection before beginning the procedure to ensure proper anesthesia of the animal. 3. Be careful not to over-dry the pellet, since completely dry DNA pellet is difﬁcult to redissolve. After removing the supernatant, the tube can be brieﬂy centrifuged, and the remaining liquid in the bottom of the microcentrifuge tube can be removed by pipetting. The remaining DNA pellet can then be immediately dissolved in water without brief air-drying. 4. When incubating the reaction mixture in a water bath at 50 C, water evaporates and condenses underneath the cap of the microcentrifuge tube, effectively increasing the concentration of the mixture. To minimize this, the microcentrifuge tube containing the reaction mixture can be put into a 50 mL Construction of Rabbit Immune Antibody Libraries 145

conical tube, which is then submerged in a water bath. By doing this the temperature is kept homogeneous around the microcentrifuge tube and condensation can be prevented. 5. Transformation titer can be: [No. of colonies Â 2 (mL recovery medium culture) Â 1000 À (μL/mL)]/[10 or 100 (μL plated) Â dilution fold (10 3)]. 6. Glucose is added to suppress lac promoter which controls the transcription of Fab-pIII fusion gene in pComb3X vector. Because pComb3X phagemid vector has the N-terminally truncated form of pIII minor coat protein (pIII C-terminal domains) which does not inhibit the bacteriophage superinfection, glucose is not required for the helper phage superinfection. However, the cells harboring different antibody genes under the control of lac promoter grow more evenly in the presence of 2% glucose.

7. In 400 mL of E. coli culture at OD600 ¼ 0.7, there are estimated to be ~1011 E. coli cells [19]. 1012 pfu of helper phage is added to ensure near-complete superinfection of bacteria. 8. The phage titer is calculated as: [No. of colonies Â (total volume of phage solution in μL)]/ À À [dilution fold (10 7 or 10 8) Â 1(μL of diluted phage added to ER2537)].

Acknowledgment

This work was supported in parts by the National Research Foun- dation of Korea (NRF) grant for Medical Bioconvergence Research Center (NRF-2013M3A6A4044991), the Bio & Medical Technol- ogy Development Program of the NRF funded by the Korean government, MSIP (NRF-2015M3A9B6029138), and the small and medium business convergence R&D program funded by the Small and Medium Business Administration (S2373145) to H.S.

References

1. Kohler G, Milstein C (1975) Continuous cul- hybridomas and monoclonal antibodies. Ple- tures of fused cells secreting antibody of pre- num Press, New York, N.Y., U.S.A. deﬁned speciﬁcity. Nature 256:495–497 4. Spieker-Polet H, Sethupathi P, Yam PC, 2. Little M, Kipriyanov SM, Le Gall F, Molden- Knight KL (1995) Rabbit monoclonal antibo- hauer G (2000) Of mice and men: hybridoma dies: generating a fusion partner to produce and recombinant antibodies. Immunol Today rabbit-rabbit hybridomas. Proc Natl Acad Sci 21:364–370 U S A 92:9348–9352 3. Shay JW (1985) Human hybridomas and 5. Liguori MJ, Hoff-Velk JA, Ostrow DH (2001) monoclonal antibodies: the biology of cell Recombinant human interleukin-6 enhances fusion. In: Engleman EG, Foung SKH, the immunoglobulin secretion of a rabbit- RLarrick J, Raubitschek A (eds) Human rabbit hybridoma. Hybridoma 20:189–198 146 Thi Thu Ha Nguyen et al.

6. Keen MJ (1995) The culture of rat myeloma 13. Steinwand M, Droste P, Frenzel A, Hust M, and rat hybridoma cells in a protein-free Dubel S, Schirrmann T (2014) The influence medium. Cytotechnology 17:193–202 of antibody fragment format on phage display 7. Digneffe C, Cormont F, Platteau B, Bazin H based affinity maturation of IgG. MAbs (1990) Fusion cell lines. In: Bazin H (ed) Rat 6:204–218 hybridomas and rat monoclonal antibodies. 14. Menzel C, Schirrmann T, Konthur Z, CRC Press, Boca Raton, Florida Jostock T, Dubel S (2008) Human antibody 8. Bystryn JC, Jacobsen JS, Liu P, Heaney-Kieras RNase fusion protein targeting CD30þ lym- J (1982) Comparison of cell-surface human phomas. Blood 111:3830–3837 melanoma-associated antigens identified by 15. Rader C, Steinberger P, Barbas CF 3rd (2001) rabbit and murine antibodies. Hybridoma Selection from antibody libraries. In: Barbas 1:465–472 CF 3rd, Burton DR, Scott JK, Silverman GJ 9. Shim H (2016) Therapeutic antibodies by (eds) Phage display: a laboratory manual. Cold phage display. Curr Pharm Des 22 Spring Harbor Laboratory Press, Cold Spring (43):6538–6559. https://doi.org/10.2174/ Harbor, N.Y., U.S.A 1381612822666160923113714 16. Andris-Widhopf J, Rader C, Barbas CF 3rd 10. Huse WD, Sastry L, Iverson SA, Kang AS, (2001) Generation of antibody libraries: Alting-Mees M, Burton DR, Benkovic SJ, Ler- immunization, RNA preparation, and cDNA ner RA (1989) Generation of a large combina- synthesis. In: Barbas CF 3rd, Burton DR, torial library of the immunoglobulin repertoire Scott JK, Silverman GJ (eds) Phage display: a in phage lambda. Science 246:1275–1281 laboratory manual. Cold Spring Harbot Labo- 11. Barbas CF 3rd, Kang AS, Lerner RA, Benkovic ratory Press, Cold Spring Harbor, New York SJ (1991) Assembly of combinatorial antibody 17. Yang HY, Kang KJ, Chung JE, Shim H (2009) libraries on phage surfaces: the gene III site. Construction of a large synthetic human scFv Proc Natl Acad Sci U S A 88:7978–7982 library with six diversified CDRs and high func- 12. Andris-Widhopf J, Steinberger P, Fuller R, tional diversity. Mol Cells 27:225–235 Rader C, Barbas CF 3rd (2001) Generation of 18. Angeletti RH (1999) Design of useful peptide antibody libraries: PCR amplification and antigens. J Biomol Tech 10:2–10 assembly of light- and heavy-chain coding 19. Tang X, Nakata Y, Li HO, Zhang M, Gao H, sequences. In: Barbas CF 3rd, Burton DR, Fujita A, Sakatsume O, Ohta T, Yokoyama K Scott JK, Silverman GJ (eds) Phage display: a (1994) The optimization of preparations of laboratory manual. Cold Spring Harbot Labo- competent cells for transformation of E. coli. ratory Press, Cold Spring Harbor, New York Nucleic Acids Res 22:2857–2858 Chapter 8

Generation of Semi-Synthetic Shark IgNAR Single-Domain Antibody Libraries

Julius Grzeschik, Doreen Ko¨nning, Steffen C. Hinz, Simon Krah, Christian Schro¨ter, Martin Empting, Harald Kolmar, and Stefan Zielonka

Abstract

Besides classical antibodies with the composition of heavy and light chains, sharks produce a unique heavy chain only isotype, termed Immunoglobulin New Antigen Receptor (IgNAR), in which antigen binding is solely mediated by a single domain, referred to as vNAR. Owing to their high affinity and specificity combined with their small size and high stability, vNAR domains emerged as promising target-binding scaffolds that can be tailor-made for biotechnological and biomedical applications. Herein, we describe protocols for the construction of semi-synthetic, CDR3-randomized vNAR libraries for the isolation of target-specific antibodies using yeast surface display or phage display as platform technology. Additionally, we provide information for affinity maturation of target-specific molecules through CDR1 diversification and sublibrary establishment.

Key words Shark, IgNAR, vNAR, Yeast surface display, Phage display, Antibody engineering, Protein engineering, Library generation, Afﬁnity maturation, Semi-synthetic antibody library, Single-domain antibody

1 Introduction

In addition to conventional hetero-tetrameric immunoglobulins composed of heavy and light chains, sharks produce a unique heavy chain only isotype that does not associate with light chains, referred to as IgNAR [1, 2]. Each chain of this homodimer consists of an N-terminal variable domain (vNAR, IgNAR V), which acts as an independent binding moiety, followed by ﬁve constant domains. vNAR domains display several unique features, clearly distinguishing them from camelid VHH domains. Due to a deletion in the framework2-CDR2-region IgNAR V domains only have two complementarity determining regions, CDR1 and CDR3 (Fig. 1) [3]. However, at the CDR2 truncation site, the remaining surface exposed loop forms a “belt-like” structure and it was shown that after antigen contact, somatic hypermutation also occurs in this

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_8, © Springer Science+Business Media LLC 2018 147 148 Julius Grzeschik et al.

Fig. 1 3D depiction of vNAR antibody fragment (pdb entry 4HGK) showing a transparent surface combined with a ribbon representation. CDR1 (red) and CDR 2(blue) are highlighted. A disulﬁde bond is shown in yellow. Picture rendered with POV-Ray (www.povray.org)

loop and in a loop which corresponds to HV4 in T-cell receptors, to which vNAR domains share structural similarity. Hence, these regions have been termed HV2 and HV4, respectively [4]. Owing to their inherent favorable attributes, shark vNAR domains emerged as promising tools for biotechnological and biomedical applications. The IgNAR V domain can be divided into four different types, categorized based on the number and pattern of noncanonical disulfide bonds that are typically not found in mammalian antibody domains [5–9]. Hence, the different types of vNAR domains form a very diverse set of additional disulfide bridges, resulting in an unprecedented repertoire of different loop structures, from which antigen-specific clones can be selected [10]. Moreover, the architecture of the paratope of those shark antibody domains seems to be predisposed to target also clefts of the antigen. Such recessed epitopes are usually not antigenic to conventional antibodies [9, 11–13]. More comprehensive reviews, addressing the different types of IgNAR V domains as well as the generation of the tremendous diversity found at the sequence level of the vNAR domain, were recently published [2, 3, 14]. Further beneficial attributes of the vNAR domain comprise its small size, superior stability compared to conventional antibody domains, as well as tolerance to irrevers- ible denaturation [15–17]. Additionally, there are multiple oppor- tunities to reformat and functionalize the vNAR domain, including Generation of Shark vNAR Libraries 149

pH-dependent binding behavior, bispecific binding, multimeric constructs, as well as Fc-based formats, clearly demonstrating the possibility of utilizing those molecules for a plethora of different applications [6, 18–24]. Accordingly, target-specific vNAR molecules have been isolated against a wide range of disease-related antigens [15–17, 25–30]. In this chapter, we provide a protocol for library establishment of semi-synthetic CDR3-randomized shark vNAR antibody domain libraries for yeast surface display and phage display, based on the natural IgNAR repertoire of the bamboo shark (Chiloscyl- lium plagiosum). Furthermore, we also detail our methods for affinity maturation of target-specific clones based on sublibrary establishment using second-generation randomization of CDR1 in order to isolate molecules with significantly enhanced affinities.

2 Materials

2.1 Shark Handling 1. 0.1% (w/v) Tricaine methanesulfonate in artiﬁcial seawater. and Blood Isolation 2. 4% (w/v) Trisodium citrate. 3. 23-gauge needle and 2 ml syringe. ® 4. TRI Reagent BD (Sigma-Aldrich, Taufkirchen, Germany). 5. 5 N Acetic Acid.

2.2 Preparation of 1. 1-Bromo-3-chloropropane. Total RNA from Whole 2. Isopropanol. Blood 3. 75% (v/v) ethanol. 4. RNAse-free water or DEPC-treated water.

® 2.3 cDNA Synthesis 1. Omniscript Reverse Transcriptase Kit (Qiagen, Hilden, and Gene-Speciﬁc Germany). Ampliﬁcation of vNAR 2. Oligo(dT)18 Primer. Regions as Template 3. RNase Inhibitor, murine (New England Biolabs, Frankfurt am for Library Main, Germany). Construction 4. Taq DNA polymerase (New England Biolabs). 5. 10Â Taq buffer (New England Biolabs). 6. dNTPs. 7. Nuclease-free water. 8. Thermocycler. 9. Device and reagents for agarose gel electrophoresis. 10. PCR Clean-Up System. 11. BioSpec (VWR) Nano or equivalent instrumentation. 150 Julius Grzeschik et al.

2.4 Library 1. BamHI-HF (New England Biolabs). Construction Yeast 2. NheI-HF (New England Biolabs). Surface Display 3. 10Â CutSmart buffer (New England Biolabs). 4. pCT plasmid [31]. 5. yeast strain: EBY100. 6. YPD media: 20 g/L D(+)-glucose, 20 g/L tryptone, 10 g/L yeast extract.

7. Electroporation buffer 1 M Sorbitol, 1 mM CaCl2. 8. LiAc buffer: 0.1 M LiAc, 10 mM DTT. 9. 1 M Sorbitol. 10. Electroporator GenePulser Xcell™ (Bio-Rad, Dreieich, Germany). 11. 0.2 cm Electroporation cuvettes (Bio-Rad). 12. Bacto™ Casamino acids (BD Biosciences, San Jose, USA).

13. SD-CAA media: 8.6 g/L NaH2PO4 Â H2O, 5.4 g/L Na2HPO4, 20 g/L D(+)-glucose, 6.7 g/L Yeast nitrogen base without amino acids and 5 g/L Bacto™ casamino acids.

14. SD-CAA agar plates: 8.6 g/L NaH2PO4 Â H2O, 5.4 g/L Na2HPO4, 20 g/L D(+)-galactose, 6.7 g/L Yeast nitrogen base without amino acids, 5 g/L Bacto™ casamino acids and 100 g/L polyethylene glycol 8000. 15. 9 cm Petri dishes.

2.5 Library 1. pHAL14 phagemid [32]. Construction Phage 2. NcoI-HF (New England Biolabs). Display 3. NotI-HF (New England Biolabs). 4. rSAP (New England Biolabs). 5. T4 DNA Ligase (New England Biolabs). 6. 10Â T4 DNA Ligase buffer (New England Biolabs). 7. 3 M sodium acetate. 8. Ethanol. 9. 0.1 cm Electroporation cuvettes (Bio-Rad). 10. Electrocompetent E. coli cells. 11. SOC medium (ThermoFisher Scientiﬁc, Darmstadt, Germany). 12. dYT medium: 1.6% (w/v) Tryptone, 1% (w/v) yeast extract, 0.5% (w/v) NaCl. 13. dYT-GAT medium: dYT medium, 100 mM Glucose, 100 μg/ ml ampicillin, 20 μg/ml tetracycline. Generation of Shark vNAR Libraries 151

14. dYT-GAT agar plates: dYT-GAT, 1.5% (w/v) agar-agar. 15. 25 cm petri dishes. 16. Glycerol.

3 Methods

The following section describes the protocol for library establishment of a semi-synthetic CDR3-randomized vNAR library based on the natural repertoire of the bamboo shark using yeast surface display technology as platform technology. To this end, blood samples need to be harvested followed by the isolation of total RNA from whole blood samples, cDNA preparation, and library generation using a generic 3-step PCR methodology (Fig. 2). Alternatively, the library can also be constructed based on a single-bamboo shark vNAR sequence as a template (see Note 1). However, we recommend starting with a template pool of different vNAR scaffolds. It is known that subtle sequence variations in the scaffold sequence such as in the framework regions may have a large impact on folding stability and protein solubility [33, 34]. These small sequence variations in the framework might contribute to the isolation of a large set of stable binders, as previously shown by our group [15, 23]. Additionally, the herein-described procedure can be slightly modiﬁed to establish semi-synthetic vNAR phage-display libraries. The generation of shark IgNAR V domain libraries was elegantly described by Flajnik and Dooley [35] as well as by the group of Barelle and coworkers [36]. Notwithstanding, we detail a cloning protocol for the generation of vNAR libraries using phage-display technology. Information on that will be given in Subheading 3.7.

3.1 Blood Collection All the procedures need to be conducted in accordance with national laws. For example, the generation of CDR3-randomized vNAR libraries as described by our group was in accordance with the national laws } 4 Abs. 3 of the German Tierschutzgesetz (TierSchG, animal welfare act). Permission number: V 54–19 c 20 15 [1] Gl 18/19 Nr. A 35/2011, Regierungspr€asidium Gies- sen, Germany (Regional council Giessen). Make sure that an expe- rienced veterinarian performs blood collection. 1. Transfer the bamboo shark from its original tank into a smaller container preﬁlled with MS-222 working solution. Anesthetize by submersion. 2. Collect 1–2 ml blood from the caudal vein using a 23-gauge needle (size depends in weight of the animal, adjust when needed). Syringe and needle should be preﬁlled with 152 Julius Grzeschik et al.

Fig. 2 Schematic representation of PCR-based library design. PCR amplified vNAR fragments from blood circulating lymphocytes of a bamboo shark or alternatively a DNA-fragment encoding a defined vNAR domain are used as template for randomization. In a first PCR the vNAR framework is amplified until CDR3. Cys is replaced in CDR1 by Tyr and a marginal diversity is introduced using the forward primer. In the second PCR CDR3 is totally randomized and in the final PCR overhangs are added for gap repair cloning or alternatively restriction sites are added for cloning into a phagemid (not shown)

approximately 100 μl Trisodium citrate solution in order to prevent coagulation. 3. Add approximately 200 μl collected blood to 750 μl TRI Reagent BD supplemented with 20 μl 5 N acetic acid. Vortex or shake thoroughly. Samples can be stored at À80 C. (Cau- tion: TRI Reagent BD is a mixture of phenol and guanidine thiocyanate. Take appropriate safety precautions). Generation of Shark vNAR Libraries 153

3.2 Total RNA 1. Incubate blood samples (in TRI Reagent BD) for at least 5 min Preparation at room temperature. 2. Add 100 μl 1-Bromo-3-chloropropane to each sample (i.e., per 200 μl blood). Shake or vortex for approx. 15 s and incubate at room temperature for 5 min. 3. Centrifuge for 15 min at 4 C at minimum 12,000 Â g. 4. Transfer the upper aqueous phase to a fresh tube preﬁlled with 500 μl isopropanol per 200 μl blood. Incubate for 10 min at room temperature. 5. Centrifuge for 10 min at 4 C at minimum 12,000 Â g. 6. Remove the supernatant carefully and wash RNA by adding 1 ml 75% ethanol per 200 μl blood. Vortex samples and centrifuge for 10 min at 4 C at minimum 12,000 Â g. 7. Remove the supernatant carefully and air-dry the RNA pellet for 5–10 min at room temperature. Dissolve RNA in RNAse- free water (approx. 100 μl per 200 μl blood sample. Volume of RNAse-free water depends on size of the pellet, see Note 2). 8. Calculate the concentration of isolated RNA by using a BioS- pec Nano or equivalent instrumentation. An OD260 of one corresponds to 40 μg/ml total RNA. 9. Analyze RNA integrity by running a 1% (w/v) gel. Two clear and distinct bands representing the 28 s and 18 s rRNA should be visible. Otherwise, the isolated RNA might already be partially degraded (see Note 3). Keep RNA exclusively on ice or freeze at À20 CorÀ80 C.

3.3 cDNA Synthesis The protocol below describes cDNA synthesis for one reaction. If more reactions should be performed, prepare a master mix. Reverse ® transcription is described based on the Omniscript Reverse Tran- scription Kit (Qiagen). The total volume per reaction is set to 20 μl. 1. Place a nuclease-free PCR tube on ice and add approximately 2 μg of isolated total RNA. Add 2 μl of Buffer RT (component of Omniscript RT Kit), 2 μl of dNTP Mix (component of Omniscript RT Kit), 1 μl Oligo-dT-primer, 1 μl RNase inhibitor, and 1 μl Omniscript Reverse Transcriptase. Add RNase- free water to a final volume of 20 μl. 2. Mix by vortexing and centrifuge briefly. 3. Incubate for 60 min at 37 C in a thermocycler. Use 5 μlof cDNA as a template for the subsequent amplification (PCR) of the natural vNAR repertoire. 154 Julius Grzeschik et al.

3.4 Amplification of In order to construct semi-synthetic CDR3-randomized vNAR the Natural vNAR libraries, the natural framework repertoire of the bamboo shark is Repertoire used as a template. To this end, based on the synthesized cDNA a PCR is performed. As described above, alternatively, the randomization can be performed based on a single vNAR template (see Note 1). From each cDNA generation reaction, 5 μl were used as template for subsequent PCR in a final volume of 50 μl. 1. Place the PCR tube on ice and add 36.75 μl Nuclease-free water. Add 5 μl of the cDNA reaction as a template as well as 5 μl10Â Standard Taq Buffer, 1 μl bamboo/nat_up and 1 μl bamboo/nat_lo (out of a 10 μM stock, primer sequences are listed in Table 1). Add 1 μl dNTP mixture (10 mM each) and 0.25 μl Taq DNA polymerase. (We recommend scaling up a master mix for at least five reactions). 2. Carry out PCR using the following parameters: Initial denaturation 95 C for 2 min. 30 cycles of 30 s at 95 C, 30 s at 55 C, and 40 s at 68 C, followed by 72 C for 7 min. 3. Analyze PCR products by 1–1.5% (w/v) agarose gel electrophoresis. Amplified vNAR genes should give a distinct band at approx. 330–350 bp. Pool PCR products and purify using a PCR clean-up kit according to the manufacturer’s instruction. PCR products might be stored at À20 C.

3.5 Generation of the The initial CDR3-randomized library is constructed in three con- CDR3-Randomized secutive PCR steps (Fig. 2). In the first PCR the forward primer PCR Insert for Library FR1/CDR1/Tyr_up replaces Cys in CDR1 by Tyr and introduces Establishment Using a marginal diversity within CDR1 that mimics the diversity found in Yeast Surface Display the natural vNAR repertoire. Cys is replaced in CDR1 to avoid as Platform mispairing of disulfide bonds that might lead to a significant frac- Technology tion of non-functional vNAR molecules in the final library that would drastically complicate the selection of favored library candidates. The second PCR is performed to fully randomize CDR3 using the degenerated primer CDR3rand/Fr4_lo. To this end, we recommend using trinucleotide phosphoramidite primers (see Note 4). In the last PCR overlaps up- and downstream of the NheI and BamHI restriction sites of the pCT plasmid [31] are added to the vNAR fragment in order to establish a library using gap repair cloning. Exchange of primer for the last PCR reaction facilitates cloning into a phage-display vector, such as pHAL14 [32]. Infor- mation on this will be given in Subheading 3.7. For all PCRs, conditions are as follows (Table 2): Generation of Shark vNAR Libraries 155

Table 1 Primers used for the construction of semi-synthetic CDR3-randomized shark vNAR antibody libraries and for the generation of CDR1-diversified sublibraries for affinity enhancement. Randomized trinucleotide primers were purchased at Ella Biotech (Martinsried, Germany)

Name Sequence (50!30)

Gene-specific amplification of vNAR regions as template for library construction Bamboo/nat_up ATGGCCSMACGGSTTGAAC AAACACC Bamboo/nat_lo WTTCACAGTCASARKGGTSCC Generation of CDR3-randomized vNAR library using yeast surface display FR1/CDR1/Tyr_up ACCATCAATTGCGTCCTAAAA GGTTCCRNMTATGBATTGGGTANMAC GTACTGGT FR3_lo CGCTTCACAGTGATATGTACC FR1_up ATGGCCGCACGGCTTGAACAAA CACCGACAACGACAACAAAGGAGGCA GGCGAATCACTGACCATCAAT TGCGTCCTAA CDR3rand/Fr4_lo WTTCACAGTCASARKGGTSCCSCC NCCTTCAAT(X)12CGCTTCACAGTGATATGTACC GR_up GTGGTGGTGGTTCTGCTAGCAT GGCCGCACGGCTTGAACA GR_lo ATAAGCTTTTGTTCGGATCCWTT CACAGTCASARKGGTSCCSCCNCC pCT_Seq_up GCGGCGGTTCCAGACTACGCTC TGCAGGCT pCT_Seq_lo GCGCGCTAACGGAACGAAAAA TAGAAA Generation of CDR1-diversified vNAR library using yeast surface display CDR1rand_up ACCATCAATTGCGTCCTAAAA (X)5TTGGGTAGCACGTACTGGTATT TCACAAAGAAG Generation of vNAR library using phage display (restriction sites are underlined) pHAL14_vNAR_NcoI_up CAGCTCAGCCGGCCATGGCCATGGCCGCACGGCTTGAACA pHAL14_vNAR_NotI_lo TGATGATGATGTGCGGCCGCWTTCACAGTCASARKGGTSCC X: triplet codon for all natural amino acids w/o Cys 156 Julius Grzeschik et al.

Table 2 PCR conditions for the generation of the CDR3-randomized insert

95 C 2 min 95 C30s 55 C30s 35 cycles 68 C40s 68 C 7 min

3.5.1 First PCR 1. Carry out approx. five reactions in parallel. Prepare a master mix. Reagents per reaction in a final volume of 50 μl: Approx. 100 ng vNAR PCR product of the natural vNAR repertoire (see Subheading 3.4). Add 1 μl FR1/CDR1/Tyr_up and 1 μl FR3_lo (out of a 10 μM stock, primer sequences are listed in Table 1). Add 1 μl dNTP mixture (10 mM each), 5 μl10Â Standard Taq Buffer, and 0.25 μl Taq DNA polymerase. Add nuclease-free water to a final volume of 50 μl. 2. Perform PCR in a thermocycler and separate PCR products on a 1–1.5% agarose gel. The amplified PCR product should be visible as a distinct band on the gel at a size of approx. 200 bp. Purify PCR products using a PCR clean up kit according to the manufacturer’s instructions (see Note 5). Determine the DNA concentration. PCR products can be stored at À20 C.

3.5.2 Second PCR 1. Carry out approx. ten reactions in parallel. Prepare a master mix. Reagents per reaction in a final volume of 50 μl: Approx. 100 ng first PCR product (see Subheading 3.5.1). Add 1 μl FR1_up and 1 μl CDR3rand/Fr4_lo (out of 10 μM stocks). Add 1 μl dNTP mixture (10 mM each), 5 μl10Â Standard Taq Buffer and 0.25 μl Taq DNA polymerase. Add nuclease-free water to a final volume of 50 μl. 2. Perform PCR in a thermocycler and separate PCR products on a 1–1.5% agarose gel. The amplified PCR product should be visible as a distinct band on the gel at a size of 330 bp. Purify PCR products using a PCR clean-up kit according to the manufacturer’s instructions. Determine the DNA concentration. Primer CDR3rand/FR4_lo introduces a totally randomized CDR3 (without Cys) in a length of 12 amino acids. When a longer CDR3 is needed, the corresponding sequence of the oligonucleotide can be adjusted accordingly.

3.5.3 Third PCR 1. Carry out as many reactions as needed to achieve an adequate library size. In general, we perform about 10 transformation reactions for a yeast surface library with an estimated Generation of Shark vNAR Libraries 157

complexity of more than 108 unique clones. For each electroporation we use approx. 6–8 μg of PCR product, consequently, approx. 80 μg of insert DNA are needed (approx. 96 PCRs). Prepare a master mix. Reagents per reaction in a final volume of 50 μl: Approx. 100–200 ng second PCR product (see Subhead- ing 3.5.2). Add 1 μl GR_up and 1 μl GR_lo. Add 1 μl dNTP mixture (10 mM each), 5 μl10Â Standard Taq Buffer, and 0.25 μl Taq DNA polymerase. Add nuclease-free water to a final volume of 50 μl. 2. Perform PCR in a thermocycler and analyze PCR products on a 1–1.5% agarose gel. The amplified PCR product should be visible as a distinct band on the gel at a size of 370 bp. Purify PCR products using a PCR clean-up kit according to the manufacturer’s instructions. Determine the DNA concentration.

3.6 Shark vNAR The following protocol for the library establishment in S. cerevisiae Library Generation for is a modiﬁed version of the improved yeast transformation protocol Yeast Surface Display of Benatuil and colleagues [37].

3.6.1 Digestion of pCT Libraries for yeast surface display are typically constructed by trans- Plasmid formation of yeast in a homologous recombination-based process referred to as gap repair. For this, the display vector needs to be digested ﬁrst. As mentioned above, in general, we perform ten transformation reactions. For each electroporation, 1–2 μg NheI and BamHI digested plasmid are used. Hence, digestion is performed with 50 μg plasmid DNA. 1. The volume for the restriction enzyme double digest is set to ® 100 μl. Add 50 μg pCT plasmid, 60 U of NheI-HF ,60U ® BamHI-HF , and 10 μl CutSmart buffer. Add nuclease-free water to a ﬁnal volume of 100 μl. 2. Digest overnight at 37 C. Analyze an aliquot of the digestion on a 1% agarose gel. Make sure that double digest is complete. Purify digested pCT plasmid using a PCR clean-up kit according to the manufacturer’s instructions. Since no ligation reaction is performed, there is no need for gel excision, because no re-ligation can occur. Determine the DNA concentration. PCR products might be stored at À20 C.

3.6.2 Yeast The protocol for the improved yeast electroporation can be found Transformation elsewhere [37]. In brief: 1. Incubate EBY100 overnight to stationary phase in YPD media at 180 rpm and 30 C. 2. Inoculate 100 ml fresh YPD media with the overnight culture to an OD600 of about 0.3. 158 Julius Grzeschik et al.

3. Incubate the cells at 30 C and 180 rpm until OD600 reaches about 1.6. 4. Centrifuge the cells at 4000 Â g for 3 min, remove the supernatant. 5. Wash the cells twice (by resuspending) using 50 ml ice-cold water followed by a wash step using 50 ml ice-cold electroporation buffer. 6. Incubate the cells (after resuspending) in 20 ml LiAc-buffer for 30 min at 30 C and 180 rpm. 7. Centrifuge the cells, wash once with 50 ml ice-cold electroporation buffer. 8. Resuspend cell pellet in approx. 200 μl electroporation buffer to a final volume of approx. 1 ml. This gives two electroporation reactions with 400 μl electrocompetent EBY100 each. 9. Combine 1–2 μg digested pCT plasmid with 3–6 μg insert DNA (volume should not exceed 50 μl) and add mixture to 400 μl electrocompetent cells. 10. Transfer cell-DNA mix to ice-cold electroporation cuvette (0.2 cm). Electroporate at 2.500 V. Time constant should range from 3.0 to 4.5 ms. Transfer cells from each “shot” into 8–10 ml of a 1:1 mixture of YPD and 1 M sorbitol. Incubate for 1 h at 30 C and 180 rpm. 11. Centrifuge cells and resuspend in 10 ml SD-CAA media. Cal- culate complexity of library by dilution plating (SD-CAA plates, estimate number of transformants after 72 h). Incubate library for at least 2 days at 30 C and 180 rpm. 12. For long time storage, centrifuge library and resuspend cells in 5% (v/v) glycerol and 0.67% (w/v) yeast nitrogen base. The final library is now ready to be screened via fluorescence- activated cell sorting. To induce surface expression, the cells need to get transferred into the SG-CAA medium. Protocols for screening yeast surface display libraries can be found elsewhere [38, 39]. To evaluate the quality of the final library, the authors recommend testing at least 10 colonies for the presence of an insert with correct length and sequence (from dilution plating). For this, plasmid DNA from overnight cultures of single clones is extracted using a commercially available yeast plasmid miniprep kit or yeast DNA extraction kit according to the manufacturer’s instructions (see Note 6). PCR can be performed using the conditions found in 3.5 with pCT_Seq_up and pCT_Seq_lo as primer combination. This should result in a distinct band on a 1%–1.5% agarose gel with a size of approx. 600 bp. Send out PCR positive clones for sequencing. Generation of Shark vNAR Libraries 159

3.7 Generation of For the shark vNAR phage-display library generation, the reader is CDR3-Randomized referred to Flajnik and Dooley as well as Barelle et al. vNAR Libraries for [35, 36]. Detailed information for cloning of antibody derivatives Phage Display into phagemid pHAL14 can be found at Schirrmann and Hust [40]. A general cloning scheme for pHAL14 is outlined below. To construct a CDR3-randomized phage-display library, the ﬁrst two steps of the three-step randomization scheme (Fig. 2) are essentially identical to the library build for yeast surface display (Subheadings 3.5.1 and 3.5.2). However, in the last PCR step, the primer pair pHAL14_vNAR_NcoI_up and pHAL14_vNAR_NotI_lo is employed to facilitate restriction enzyme-based cloning into phage- midpHAL14(see Note 7).

3.7.1 Insert Generation 1. Carry out as many reactions as needed to achieve an adequate (Third PCR) for Phage library size. Reagents per reaction in a final volume of 50 μl: Display Approx. 100–200 ng second PCR product (see Subheading 3.5.2). Add 1 μl pHAL14_vNAR_NcoI_up and 1 μl pHAL14_vNAR_NotI_lo (out of 10 μMstocks).Add1μl dNTP mixture (10 mM each), 5 μl10Â Standard Taq Buffer, and 0.25 μl Taq DNA polymerase. Add nuclease-free water to a final volume of 50 μl. The PCR conditions are shown in Table 2. 2. Perform PCR in a thermocycler and separate PCR products on a 1–1.5% agarose gel. The amplified PCR product should be visible as a distinct band on the gel at a size of 370 bp. Purify PCR products using a gel clean-up kit according to the manufacturer’s instructions. Determine the DNA concentration using BioSpec Nano.

3.7.2 Cloning into The phagemid pHAL14, established in the group of Dubel€ and pHAL14 coworkers, was originally constructed for easy two-step cloning of scFv libraries [32]. Here, the antibody fragments are fused to the minor coat protein III gene of bacteriophage M13, facilitating expression on the surface of the phage. The outlined cloning protocol describes a simpliﬁed procedure that was originally established by Schirrmann and Hust [40], since only one domain needs to be cloned into the vector. Digestion of pHAL14 We recommend performing single-restriction enzyme digestions of the vector in parallel to the double digest, in order to analyze whether cleavage is complete. 1. The volume is set to 100 μl. Add approx. 5 μg pHAL14, 40 U NcoI-HF, 40 U NotI-HF, and 10 μl CutSmart buffer. Add nuclease-free water to a ﬁnal volume of 100 μl. 2. Incubate for at least 2 h (or overnight) at 37 C. Analyze an aliquot on a 1% agarose gel. When digestion is not complete, extend incubation time or add up to 20 U of each enzyme. 160 Julius Grzeschik et al.

3. Inactivate at 80 C for 20 min. 4. Add 1 μl rSAP and incubate at 37 C for 1 h. 5. Purify vector using a PCR clean-up kit according to the manufacturer’s instructions. Elute in 50 μl nuclease-free water. Determine DNA concentration (see Note 8). Digestion of vNAR insert 1. The volume is set to 100 μl. Add approx. 2 μg vNAR insert (Subheading 3.7.1), 40 U NcoI-HF, 40 U NotI-HF, and 10 μl CutSmart buffer. Add nuclease-free water to a final volume of 100 μl. 2. Incubate for at least 2 h (or overnight) at 37 C. Analyze an aliquot on a 1% agarose gel. When digestion is not complete, extend incubation time or add up to 20 U of each enzyme. 3. Inactivate at 80 C for 20 min. 4. Purify insert using a PCR clean-up kit according to the manufacturer’s instructions. Elute in 50 μl nuclease-free water. Determine DNA concentration. Ligation and library build The following procedure is for one electroporation reaction. Parallelize when more electroporations are needed in order to obtain a sufficient library size. Up to 108 unique clones can be expected from one reaction. 1. Set up ligation with 1000 ng pHAL14 (approx. 4200 bp) and 270 ng vNAR insert (approx. 370 bp) in a final reaction volume of 100 μl. Add 10 μl T4 ligase buffer and 400 U T4 ligase. 2. Incubate at 16 C overnight. 3. Inactivate at 65 C for 10 min. 4. Precipitate the ligation mixture with 10 μl 3 M sodium acetate, pH 5.2 and 250 μl ethanol, incubate for 5 min at RT, and centrifuge for 2 min at 16,000 Â g at 4 C. 5. Wash the pellet with 500 μl 70% (v/v) ethanol and centrifuge for 2 min at 16,000 Â g at 4 C. Repeat this step and resolve the DNA pellet in 35 μl nuclease-free water (see Note 9). 6. Thaw 25 μl “high-efficiency” electrocompetent E. coli (such as XL1-Blue MRF’) on ice and mix with ligation mixture (see Note 10). 7. Transfer the mix to an ice-cold electroporation cuvette (0.1 cm) and pulse at 1700 V. Immediately, add 1 ml 37 C pre-warmed SOC medium and incubate for 1 h at 37 C and 600 rpm. Generation of Shark vNAR Libraries 161

8. Remove 10 μl for dilution plating in order to calculate library complexity. Perform dilution plating on dYT-GAT plates. Incubate overnight at 37 C. 9. Plate out remaining 990 μl on two 25 cm square petri dishes of dYT-GAT agar. Incubate overnight at 37 C. 10. Calculate library size. 11. Float off the colonies on all “library” plates with dYT media, add 20% glycerol, and store at À80 C. Procedures for the analysis of library quality (colony PCR), library packaging, and vNAR production as well as selection, i.e., panning have been elegantly described elsewhere [35, 36, 40].

3.8 Affinity We previously established a generic two-step strategy for the isola- Maturation by CDR1 tion of high-affinity shark-derived antibody domains [15]. First, Diversification and CDR3 was randomized, as described above. This library was sub- Sublibrary jected to screening for target-specific molecules, albeit with mod- Establishment of erate affinities to their antigen. DNA of the target-specific Target-Enriched population was isolated after the last screening round and CDR1 Binders comprising five residues was totally diversified. Sublibraries were established and screened with significantly decreased target concentrations. This strategy proved to be useful to obtain vNARs with affinities in the nanomolar range. Interestingly, this in vitro method resembles the natural immune response in sharks to select clones from a primary nearly entirely CDR3-based IgNAR repertoire, followed by affinity maturation of CDR1 and hypervariable loops after antigen exposure [41]. As shown in Fig. 3, CDR1 of the target-specific population (see Note 11) is randomized in a consecutive 3-step PCR, similar to the initial diversification of CDR3. Starting point is isolated plasmid DNA from the last round of screening using yeast surface display or phage display in which target-specific clones have been significantly enriched. Plasmid DNA can be isolated using commercially available kits according to the manufacturer’s instructions. Akin to the initial library generation, the PCR protocol for insert preparation is very similar between yeast display and phage display, only differing in oligonucleotides used for final cloning into pCT or pHAL14, respectively. Consequently, the protocol below describes the CDR1-diversified insert generation for yeast display and phage display. Methodologies for gap-repair cloning into pCT or cloning into pHAL14 can be found in Subheadings 3.5, 3.6 and 3.7, respectively. For all PCRs, the conditions are shown in Table 3.

3.8.1 First PCR 1. Prepare a master mix. Reagents per reaction in a ﬁnal volume of 50 μl: Approx. 100 ng isolated plasmid DNA (pCT or pHAL14, respectively) isolated after last round of screening. Primer combination yeast display: add 1 μl CDR1rand_up and 162 Julius Grzeschik et al.

Fig. 3 Schematic representation of PCR-based sublibrary design for affinity maturation. Plasmid DNA of target-specific vNAR domains or alternatively of a single antigen-specific vNAR domain are used as template for sublibrary establishment. In a first PCR five residues of CDR1 are totally randomized. Subsequent PCRs are performed to generate a full vNAR domain with gap repair overhangs (as shown) or alternatively with restriction sites for phagemid cloning (not shown, see Subheading 3.8)

Table 3 PCR conditions for the generation of the CDR1-randomized insert

95 C 2 min 95 C30s 55 C30s 35 cycles 68 C40s 68 C 7 min Generation of Shark vNAR Libraries 163

1 μl GR_lo (out of 10 μM stocks). Primer combination phage display: add 1 μl CDR1rand_up and 1 μl pHAL14_vNAR_No- tI_lo (out of 10 μM stocks. Primer sequences are listed in Table 1). Add 1 μl dNTP mixture (10 mM each), 5 μl10Â Standard Taq Buffer, and 0.25 μl Taq DNA polymerase. Add nuclease-free water to a ﬁnal volume of 50 μl. 2. Perform PCR in a thermocycler and separate PCR products on a 1–1.5% agarose gel. The ampliﬁed PCR product should be visible as a distinct band on the gel at a size of approx. 290 bp. Purify PCR products using a gel clean-up kit according to the manufacturer’s instructions. Determine the DNA concentration. PCR products might be stored at À20 C.

3.8.2 Second PCR 1. Carry out approximately ten reactions in parallel. Prepare a master mix. Reagents per reaction in a final volume of 50 μl: Approx. 100 ng first PCR product (see Subheading 3.8.1). Primer combination yeast display: add 1 μl FR1_up and 1 μl GR_lo. Primer combination phage display: add 1 μl FR1_up and 1 μl pHAL14_vNAR_NotI_lo (out of 10 μM stocks). Add 1 μl dNTP mixture (10 mM each), 5 μl10Â Standard Taq Buffer and 0.25 μl Taq DNA polymerase. Add nuclease-free water to a final volume of 50 μl. 2. Perform PCR in a thermocycler and separate PCR products on a 1–1.5% agarose gel. The amplified PCR product should be visible as a distinct band on the gel at a size of 330 bp. Purify PCR products using a gel clean-up kit according to the manufacturer’s instructions. Determine the DNA concentration using Biospec Nano or equivalent equipment.

3.8.3 Third PCR 1. Carry out as many reactions as needed to achieve an adequate library size. Prepare a master mix. Reagents per reaction in a final volume of 50 μl: Approx. 100–200 ng second PCR product (see Subheading 3.8.2). Primer combination yeast display: add 1 μl GR_up and 1 μl GR_lo (out of 10 μM stocks). Primer combination phage display: add 1 μl pHAL14_vNAR_N- coI_up and 1 μl pHAL14_vNAR_NotI_lo (out of 10 μM stocks). Add 1 μl dNTP mixture (10 mM each), 5 μl10Â Standard Taq Buffer and 0.25 μl Taq DNA polymerase. Add nuclease-free water to a final volume of 50 μl. 2. Perform PCR in a thermocycler and separate PCR products on a 1–1.5% agarose gel. The amplified PCR product should be visible as a distinct band on the gel at a size of 370 bp. Purify PCR products using a gel clean-up kit according to the manufacturer’s instructions. Determine the DNA concentration. 164 Julius Grzeschik et al.

4 Notes

1. Instead of constructing a semi-synthetic vNAR library based on the natural repertoire of the bamboo shark, also the construction of a fully synthetic vNAR library based on a single-vNAR scaffold is possible. The following sequence can be used as a template and synthesized: vNAR template sequence: ATGGCCGCACGGCTTGAACAAACACCGACAACGAC AACAAAGGAGGCAGGCGAATCACTGACCATCAATTGC GTCCTAAAAGGTTCCAGATATGGATTGGGTACAACG TACTGGTATTTCACAAAAAAGGGCGCAACAAAGAAGG CGAGCTTATCAACTGGCGGACGATACTCGGACACAAA GAATACGGCATCAAAGTCCTTTTCCTTGCGAATTAGT GACCTAAGAGTTGAAGACAGTGGTACATATCACTGT GAAGCGATGCTGGGCATTAACCCATTTGGCTGGAA ACGGCTGATTGAAGGAGGGGGCACCACTGTGACTGT GAAA 2. For resuspension of the RNA pellet we recommend starting with a small volume of RNase-free water. Add small aliquots of water, until the pellet is completely dissolved. This should ensure that RNA is concentrated as much as possible. 3. Since it is not necessary to display the full diversity of the natural vNAR repertoire, working with partially degraded RNA might also work for semi-synthetic library establishment. 4. Other diversiﬁcation strategies, e.g., NNK or NNS randomization might also work. However, these technologies typically result in the incorporation of unwanted stop codons, clearly impairing the quality of the library. 5. When unwanted side-products appear on the gel, gel-excision using a commercial kit according to the manufacturer’s instructions is possible. However, at least in the third PCR we do not recommend gel-excision, since the yield of extracted DNA might be too low for library construction. Instead, try enhancing PCR stringency by increasing the annealing temperature. 6. Alternatively, a single clone can be picked with a sterile pipette- tip and transferred into 10 μl 0.02 M NaOH. After 10 min incubation at 99 C, use 1 μl as template for colony PCR using pCT_Seq_up and pCT_Seq_lo. 7. When another phage-display vector is used instead of pHAL14, primer sequences need to be adjusted accordingly. 8. When extensive religation occurs, we recommend to gel-excise the digested pHAL14 fragment using a commercial DNA gel extraction kit according to the manufacturer’s instructions. This will lower the yield of plasmid DNA signiﬁcantly. Scale Generation of Shark vNAR Libraries 165

up or parallelize digestions in order to obtain an adequate amount of plasmid DNA. 9. Instead of steps 4 and 5 of Subheading 3.7.2 Ligation and library build, ligation mixture might be purified using a PCR clean-up kit according to the manufacturer’s instructions. Elute in 35 μl. 10. An alternative to commercially available electrocompetent E. coli cells is the preparation of electrocompetent cells [42]. 11. Alternatively, affinity maturation using this methodology might be performed with a defined single clone, i.e., target- specific molecule. However, success of affinity maturation depends on the structure of the paratope of the initially isolated vNAR.

Acknowledgments

We thank Michael Hust for discussion and advice related to the page-display section of this chapter. Furthermore, we gratefully acknowledge funding from Merck Lab@Technische Universit€at Darmstadt.

References

1. Greenberg AS, Avila D, Hughes M, Hughes A, Mosyak L, Tchistiakova L (2013) Atypical anti- McKinney EC, Flajnik MF (1995) A new antigen recognition mode of a shark immunoglob- gen receptor gene family that undergoes rear- ulin new antigen receptor (IgNAR) variable rangement and extensive somatic domain characterized by humanization and diversification in sharks. Nature 374:168–173 structural analysis. J Biol Chem 2. Zielonka S, Empting M, Grzeschik J, 288:17408–17419 Ko¨nning D, Barelle CJ, Kolmar H (2015) 7. Stanfield RL, Dooley H, Flajnik MF, Wilson IA Structural insights and biomedical potential of (2004) Crystal structure of a shark single- IgNAR scaffolds from sharks. MAbs 7:15–25 domain antibody V region in complex with 3. Krah S, Schro¨ter C, Zielonka S, Empting M, lysozyme. Science 305:1770–1773 Valldorf B, Kolmar H (2016) Single-domain 8. Stanfield RL, Dooley H, Verdino P, Flajnik antibodies for biomedical applications. Immu- MF, Wilson IA (2007) Maturation of shark nopharmacol Immunotoxicol 38:21–28 single-domain (IgNAR) antibodies: evidence 4. Dooley H, Stanfield RL, Brady RA, Flajnik MF for induced-fit binding. J Mol Biol (2006) First molecular and biochemical analy- 367:358–372 sis of in vivo affinity maturation in an ectother- 9. Streltsov VA, Carmichael JA, Nuttall SD mic vertebrate. Proc Natl Acad Sci U S A (2005) Structure of a shark IgNAR antibody 103:1846–1851 variable domain and modeling of an early- 5. Diaz M, Stanfield RL, Greenberg AS, Flajnik developmental isotype. Protein Sci MF (2002) Structural analysis, selection, and 14:2901–2909 ontogeny of the shark new antigen receptor 10. Simmons DP, Streltsov VA, Dolezal O, Hud- (IgNAR): identification of a new locus prefer- son PJ, Coley AM, Foley M, Proll DF, Nuttall entially expressed in early development. Immu- SD (2008) Shark IgNAR antibody mimotopes nogenetics 54:501–512 target a murine immunoglobulin through 6. Kovalenko OV, Olland A, Piche-Nicholas N, extended CDR3 loop structures. Proteins Godbole A, King D, Svenson K, Calabro V, 71:119–130 Muller€ MR, Barelle CJ, Somers W, Gill DS, 166 Julius Grzeschik et al.

11. Flajnik MF, Deschacht N, Muyldermans S 21. Uth C, Zielonka S, Ho¨rner S, Rasche N, (2011) A case of convergence: why did a simple Plog A, Orelma H, Avrutina O, Zhang K, Kol- alternative to canonical antibodies arise in mar H (2014) A chemoenzymatic approach to sharks and camels? PLoS Biol 9:e1001120 protein immobilization onto crystalline cellu- 12. Streltsov VA, Varghese JN, Carmichael JA, lose nanoscaffolds. Angew Chem Int Ed Engl Irving RA, Hudson PJ, Nuttall SD (2004) 53:12618–12623 Structural evidence for evolution of shark Ig 22. Streltsov VA, Varghese JN, Masters CL, Nuttall new antigen receptor variable domain antibo- SD (2011) Crystal structure of the amyloid- dies from a cell-surface receptor. Proc Natl beta p3 fragment provides a model for oligo- Acad Sci U S A 101:12444–12449 mer formation in Alzheimer’s disease. J Neu- 13. Henderson KA, Streltsov VA, Coley AM, rosci 31:1419–1426 Dolezal O, Hudson PJ, Batchelor AH, 23. Ko¨nning D, Zielonka S, Sellmann C, Gupta A, Bai T, Murphy VJ, Anders RF, Schro¨ter C, Grzeschik J, Becker S, Kolmar H Foley M, Nuttall SD (2007) Structure of an (2016) Isolation of a pH-sensitive IgNAR vari- IgNAR-AMA1 complex: targeting a conserved able domain from a yeast-displayed, histidine- hydrophobic cleft broadens malarial strain rec- doped master library. Mar Biotechnol ognition. Structure 15:1452–1466 (NY) 18:161–167 14. Barelle C, Porter A (2015) VNARs: an ancient 24. Zielonka S, Empting M, Ko¨nning D, and unique repertoire of molecules that deliver Grzeschik J, Krah S, Becker S, Dickgiesser S, small, soluble, stable and high affinity binders Kolmar H (2015) The shark strikes twice: of proteins. Antibodies 4:240 hypervariable loop 2 of shark IgNAR antibody 15. Zielonka S, Weber N, Becker S, Doerner A, variable domains and its potential to function Christmann A, Christmann C, Uth C, Fritz J, as an autonomous paratope. Mar Biotechnol Sch€afer E, Steinmann B, Empting M, (NY) 17:386–392 Ockelmann P, Lierz M, Kolmar H (2014) 25. Walsh R, Nuttall S, Revill P, Colledge D, Shark attack: high affinity binding proteins Cabuang L, Soppe S, Dolezal O, Griffiths K, derived from shark vNAR domains by stepwise Bartholomeusz A, Locarnini S (2011) Target- in vitro affinity maturation. J Biotechnol ing the hepatitis B virus precore antigen with a 191:236–245 novel IgNAR single variable domain intrabody. 16. Liu JL, Anderson GP, Delehanty JB, Virology 411:132–141 Baumann R, Hayhurst A, Goldman ER 26. Liu JL, Anderson GP, Goldman ER (2007) (2007) Selection of cholera toxin specific Isolation of anti-toxin single domain antibo- IgNAR single-domain antibodies from a naive dies from a semi-synthetic spiny dogfish shark shark library. Mol Immunol 44:1775–1783 display library. BMC Biotechnol 7:78 17. Goodchild SA, Dooley H, Schoepp RJ, 27. Nuttall SD, Humberstone KS, Krishnan UV, Flajnik M, Lonsdale SG (2011) Isolation and Carmichael JA, Doughty L, Hattarki M, characterisation of Ebolavirus-specific recom- Coley AM, Casey JL, Anders RF, Foley M, binant antibody fragments from murine and Irving RA, Hudson PJ (2004) Selection and shark immune libraries. Mol Immunol affinity maturation of IgNAR variable domains 48:2027–2037 targeting plasmodium falciparum AMA1. Pro- 18. Kovaleva M, Ferguson L, Steven J, Porter A, teins 55:187–197 Barelle C (2014) Shark variable new antigen 28. Nuttall SD, Krishnan UV, Doughty L, receptor biologics–a novel technology platform Pearson K, Ryan MT, Hoogenraad NJ, for therapeutic drug development. Expert Hattarki M, Carmichael JA, Irving RA, Hud- Opin Biol Ther 14(10):1527–1539 son PJ (2003) Isolation and characterization of 19. Muller€ MR, Saunders K, Grace C, Jin M, an IgNAR variable domain specific for the Piche-Nicholas N, Steven J, O’Dwyer R, human mitochondrial translocase receptor Wu L, Khetemenee L, Vugmeyster Y, Hickling Tom70. Eur J Biochem 270:3543–3554 TP, Tchistiakova L, Olland S, Gill D, Jensen A, 29. Ohtani M, Hikima J, Jung TS, Kondo H, Barelle CJ (2012) Improving the pharmacoki- Hirono I, Takeyama H, Aoki T (2013) Variable netic properties of biologics by fusion to an domain antibodies specific for viral hemor- anti-HSA shark VNAR domain. MAbs rhagic septicemia virus (VHSV) selected from 4:673–685 a randomized IgNAR phage display library. 20. Simmons DP, Abregu FA, Krishnan UV, Proll Fish Shellfish Immunol 34:724–728 DF, Streltsov VA, Doughty L, Hattarki MK, 30. Bojalil R, Mata-Gonzalez MT, Sanchez- Nuttall SD (2006) Dimerisation strategies for Munoz F, Yee Y, Argueta I, Bolanos L, shark IgNAR single domain antibody frag- Amezcua-Guerra LM, Camacho-Villegas TA, ments. J Immunol Methods 315:171–184 Sanchez-Castrejon E, Garcia-Ubbelohde WJ, Generation of Shark vNAR Libraries 167

Licea-Navarro AF, Marquez-Velasco R, domain antibodies from shark immune reper- Paniagua-Solis JF (2013) Anti-tumor necrosis toires. Methods Mol Biol 907:177–194 factor VNAR single domains reduce lethality 37. Benatuil L, Perez JM, Belk J, Hsieh CM and regulate underlying inflammatory response (2010) An improved yeast transformation in a murine model of endotoxic shock. BMC method for the generation of very large Immunol 14:17 human antibody libraries. Protein Eng Des Sel 31. Boder ET, Wittrup KD (1997) Yeast surface 23:155–159 display for screening combinatorial polypeptide 38. Boder ET, Wittrup KD (2000) Yeast surface libraries. Nat Biotechnol 15:553–557 display for directed evolution of protein expres- 32. Hust M, Meyer T, Voedisch B, Rulker€ T, sion, affinity, and stability. Methods Enzymol Thie H, El-Ghezal A, Kirsch MI, Schutte€ M, 328:430–444 Helmsing S, Meier D, Schirrmann T, Dubel€ S 39. Angelini A, Chen TF, de Picciotto S, Yang NJ, (2011) A human scFv antibody generation Tzeng A, Santos MS, Van Deventer JA, Traxl- pipeline for proteome research. J Biotechnol mayr MW, Wittrup KD (2015) Protein engi- 152:159–170 neering and selection using yeast surface 33. Ewert S, Honegger A, Pluckthun€ A (2003) display. Methods Mol Biol 1319:3–36 Structure-based improvement of the biophysi- 40. Schirrmann T, Hust M (2010) Construction of cal properties of immunoglobulin VH domains human antibody gene libraries and selection of with a generalizable approach. Biochemistry antibodies by phage display. Methods Mol Biol 42:1517–1528 651:177–209 34. Ewert S, Huber T, Honegger A, Pluckthun€ A 41. Diaz M, Greenberg AS, Flajnik MF (1998) (2003) Biophysical properties of human anti- Somatic hypermutation of the new antigen body variable domains. J Mol Biol receptor gene (NAR) in the nurse shark does 325:531–553 not generate the repertoire: possible role in 35. Flajnik MF, Dooley H (2009) The generation antigen-driven reactions in the absence of ger- and selection of single-domain, v region minal centers. Proc Natl Acad Sci U S A libraries from nurse sharks. Methods Mol Biol 95:14343–14348 562:71–82 42. Sambrook J, Russell DW (2006) Transforma- 36. Muller€ MR, O’Dwyer R, Kovaleva M, tion of E. Coli by electroporation. CSH Protoc Rudkin F, Dooley H, Barelle CJ (2012) Gen- 2006(1). https://doi.org/10.1101/pdb. eration and isolation of target-specific single- prot3933 Chapter 9

Construction of High-Quality Camel Immune Antibody Libraries

Ema Roma˜o, Vianney Poignavent, Ce´cile Vincke, Christophe Ritzenthaler, Serge Muyldermans, and Baptiste Monsion

Abstract

Single-domain antibodies libraries of heavy-chain only immunoglobulins from camelids or shark are enriched for high-affinity antigen-specific binders by a short in vivo immunization. Thus, potent binders are readily retrieved from relatively small-sized libraries of 107–108 individual transformants, mostly after phage display and panning on a purified target. However, the remaining drawback of this strategy arises from the need to generate a dedicated library, for nearly every envisaged target. Therefore, all the procedures that shorten and facilitate the construction of an immune library of best possible quality are definitely a step forward. In this chapter, we provide the protocol to generate a high-quality immune VHH library using the Golden Gate Cloning strategy employing an adapted phage display vector where a lethal ccdB gene has to be substituted by the VHH gene. With this procedure, the construction of the library can be shortened to less than a week starting from bleeding the animal. Our libraries exceed 108 individual transformants and close to 100% of the clones harbor a phage display vector having an insert with the length of a VHH gene. These libraries are also more economic to make than previous standard approaches using classical restriction enzymes and ligations. The quality of the Nanobodies that are retrieved from immune libraries obtained by Golden Gate Cloning is identical to those from immune libraries made according to the classical procedure.

Key words Golden gate cloning, Immune libraries, Nanobodies, Single-domain antibodies, VHH

1 Introduction

From all libraries used to retrieve molecular recognition units, only antibody-based scaffolds offer the possibility of a natural enrichment within the repertoire of target-specific binders during an immunization step [1]. Although the immunization will elicit in vivo affinity-matured, target-specific antibodies, these are largely lost during the subsequent construction of the Fab or scFv library. Indeed, for practical reasons, VH and VL gene fragments are

Ema Roma˜o and Vianney Poignavent contributed equally to this work.

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_9, © Springer Science+Business Media LLC 2018 169 170 Ema Roma˜ o et al.

amplified separately, in bulk from collected lymphocytes and cloned afterward as VH-VL pairs in the library. Thereby, the original VH- VL pair as it was affinity-matured in the B cell during the immunization becomes scrambled in the library. This is not the case for single-domain antibody libraries derived from some shark species or camelids (llama, vicugna, dromedaries, and camels). These species express antigen-binding antibodies, comprising a homodimer of the heavy chain immunoglobulin lacking light chains [2, 3]. The heavy chain-only antibodies (HCAbs in camelids or Ig-NAR in shark) are raised and affinity matured during an immunization step. The antigen binding entity of Ig-NARs of HCAbs involves a single domain only, referred to as V-NAR when derived from Ig-NARs and VHH when derived from HCAbs. Since VHH sequences share a higher degree of sequence and structure identity with human VH domains, and since llama (Lama glama) and alpaca (Vicugna pacos) are easier to keep in our countries, we prefer to immunize these species to generate our immune single-domain libraries. The classical strategy to generate immune single-domain libraries from camelids requires five steps [4]: (1) An immunization step; (2) recovering blood peripheral blood lymphocytes; (3) amplifying the VHH gene fragments from the lymphocyte cDNA; (4) ligating these VHH fragments after restriction enzyme digestion in a phage display vector with compatible ends; and (5) transforming the ligated material in a suitable bacterial host. The efficiency of restriction enzyme digestion and ligation step is very critical. Cutting high amounts of PCR fragments at high concentration is tedious. In addition, the presence of a tiny fraction of uncut vector DNA within the ligation mixture will generate a significant number of transformants with “empty” vectors, which might complicate subsequent panning. In practice, for most of our libraries obtained by ligating restriction enzyme digested PCR amplicons, at best “only” 70–80% of the transformants within the library contain a phage display vector with an insert length of a VHH. Here, we describe an alternative cloning strategy, one that is based on the Golden Gate Cloning [5] and on the negative selection of transformants with an unmodified phage display vector by the presence of a lethal ccdB gene [6]. Thus, we adapted our phage display vector and our PCR primers to amplify the VHH genes. With this modified protocol we routinely obtain 108 transformants, and importantly, where 100% of the clones possess a phagemid carrying an insert with a length of a VHH. The immunization, collecting blood of the immunized animal, lymphocyte preparation, and cDNA synthesis are all performed as previously described in great detail [7, 8] and will not be repeated here. Also, the protocols for phage display and the panning on antigen immobilized on microtiter plates are identical to those previous publications. Golden Gate Cloning of VHH Libraries 171

2 Materials

2.1 Amplification 1. Thermocycler. of VHH Sequences 2. Agarose gel electrophoresis equipment and UV trans- by PCR illuminator. ® 3. Nanodrop ND-1000 (Thermo Scientific, Isogen Life science, PW De Meern, The Netherlands), or equivalent spectrophotometer. 4. PCR tubes. 5. Oligonucleotides CALL001, CALL002, VHH-BACK-SAPI, VHH-FORWARD-SAPI. Please refer to Table 1 for a list of primers used. All primers were prepared in 20 mM in H2O. The primer VHH-BACK-SAPI primer has degeneracy R is A or G; the recognition sequence of Sap I restriction enzyme is under- ∨ lined and ∧ indicate nicks introduced in top and bottom strand, respectively. VHH-FORWARD-SAPI primer has the recognition sequence of Sap I restriction enzyme that is under- ∨ lined, and ∧ indicate nick in top and bottom strands, respectively, the recognition sequence of BstEII restriction enzyme (GGTNACC) is also underlined 6. dNTP mix (Thermo Fischer Scientific, Haasrode, Belgium): 10 mM dATP, 10 mM dCTP, 10 mM dGTP, 10 mM dTTP. 7. FastStart Taq DNA polymerase (5 U/mL, Roche, Basel, Swit- zerland) supplied with 10Â PCR buffer containing 20 mM MgCl2. 8. Agarose. 9. TBE electrophoresis buffer for 1 L pH 8.3: 10.80 g Tris, 5.50 g boric acid, 0.93 g EDTA.

Table 1 Table showing list of primers and their respective sequences

Refer to No Primer name Primer sequence Notes

1. CALL001 50-GTCCTGGCTGCTCTTCTACAAGG-30 See Note 1 2. CALL002 50-GGTACGTGCTGTTGAACTGTTCC-30 See Note 2 0 ∨ 3. VHH-BACK-SAPI 5 -CTTGGCTCTTCT GTG∧ CAG CTG CAG GAG TCT See Note 3 GGR GGA GG-30 0 ∨ 4. VHH-FORWARD- 5 -TGATGCTCTTCC GC T∧GA GGA GAC GGT GAC CTG See Note 4 SAPI GGT-30 5. MP57 50-TTATGCTTCCGGCTCGTATG-30 6. GIII 50-CCACAGACAGCCCTCATAG-30 172 Ema Roma˜ o et al.

10. Ethidium bromide (stock solution at 1 mg/mL in H2O, stored at room temperature in the dark). 11. Analytical and preparative 1% (w/v) agarose gel: 1 g agarose in 100 mL TBE, boil in a microwave to solubilize agarose, cool down to 60 C, supplement with ethidium bromide (1 μg/ mL), and pour in a gel tray. 12. Analytical 2% agarose gel: as above but with 2 g agarose for 100 mL. 13. QIAquick Gel Extraction kit (Qiagen, Hilden, Germany). 14. QIAquick PCR Puriﬁcation kit (Qiagen). 15. DNA Smart ladder (Eurogentec, Seraing, Belgium) or an equivalent DNA molecular weight marker.

2.2 Home Made 1. Oven/incubator at 37 C. Electro-competent TG1 2. Shaker incubator at 37 C. Cells 3. Centrifuge cooled at 4 C. 4. Spectrophotometer and plastic disposable or glass cuvettes to measure turbidity of bacterial cell culture. 5. Bucket with ice. 6. Sterile Eppendorf tubes. 7. Falcon tubes (50-mL) and/or sterile centrifuge tubes. 8. 1 mL micropipette tips with end cut off, sterile and put on ice. 9. E. coli TG1 cells stored in 25% glycerol in freezer at À80 C. 10. D-glucose (stock 20% in water and ﬁltered through 0.22 μm ﬁlters (Millipore Express, Merck, Overijse, Belgium) for sterilization.

11. 10Â M9 salts for 1 L: 74.76 g Na2HPO4.2aq, 30.0 g KH2PO4, 10.0 g NH4Cl and 5.0 g NaCl. 12. Minimal Medium plates: autoclave 3.75 g micro agar in 225 mL Milli-Q H2O. Cool to 60 C and add 25 mL 10Â M9 salts, 250 μL 1 M MgSO4 (autoclaved), 250 μL 100 mM CaCl2 (autoclaved), 125 μL Vitamin B (10 mg/mL, filtered to sterilize) and 2.5 mL 20% D-glucose. Mix and pour 20 mL into petri dish (90 mm diameter). Dry plates in laminar flow and store at 4 C until use. 13. 2Â TY medium pH 7.4: 16.0 g Bacto-Tryptone (Duchefa Biochemie, Groot Bijgaarden, Belgium), 10.0 g Bacto-Yeast extract (Duchefa Biochemie, Groot Bijgaarden, Belgium), 5.0 g NaCl, bring to 1 L with water, adjust the pH to 7.4 with NaOH. Pour 330 mL per baffled shake flask of 1 L, autoclave and store at room temperature. 14. 1 mM HEPES, pH 7.0, autoclaved, and stored in cold room. Golden Gate Cloning of VHH Libraries 173

15. Glycerol 10% (v/v) in H2O, autoclaved, and stored in cold room.

2.3 Restriction 1. Eppendorf tubes/PCR tubes. Enzyme Digestion 2. Thermocycler. of VHHs and Ligation 3. Freezer at À80 C. to pMECS-GG 4. Vortex, microcentrifuge. 5. pMECS-GG vector (see Fig. 1)(see Note 5). 6. Restriction enzyme Sap I (10,000 U/mL; New England Bio- Labs Inc., USA) and corresponding 10Â CutSmart buffer.

Fig. 1 The pMECS vector is a pUC-derived phagemid with an F1 origin of replication, where expression is driven from a Plac promoter that can be induced by IPTG. The VHH sequences are ligated in frame and downstream of a pelB leader signal sequence, and upstream of a hemagglutinin (HA) tag, a His6 tag and the gene III from M13 bacteriophage. An amber stop codon is present in between the His6 codons and the gene III. The pelB leader sequence ensures directing the Nanobody to the periplasmic region of E. coli. The His6 and HA tags can be used to detect Nanobody protein with tag-speciﬁc antibodies (e.g., in ELISA). The His6 tag is used to purify the Nanobody by immobilized metal afﬁnity chromatography (IMAC). In amber stop codon suppressor strains (e.g., E. coli TG1), the gene III protein is occasionally fused to the Nanobody and therefore, when infected with helper phages, the VHH is displayed at the tip of phage particles. In non-suppressor strains (e.g., E. coli WK6), the translation of the Nanobody protein stops after the HA and His6 tags 174 Ema Roma˜ o et al.

7. T4 DNA ligase (5 U/mL, Thermo Fischer Scientiﬁc) and 10Â ligation buffer. 8. ATP solution (10 mM, Thermo Fischer Scientiﬁc). 9. TE pH 8.0: 1 mM EDTA, 10 mM Tris–HCl. 10. Chloroform/Isoamyl alcohol (24/1) saturated with TE. 11. TE-Saturated phenol (Sigma Aldrich, Overijse, Belgium). 12. Sodium acetate 3 M pH 5.2. 13. Absolute ethanol (stored at À20 C).

14. Sterile H2O.

2.4 Test Ligation, 1. Oven/incubator at 37 C. Test Transformation, 2. Shaker incubator at 37 C. and Initial Quality 3. Eppendorf tubes. Control 4. Bucket with ice. 5. E. coli Pulser (electroporation instrument, Bio-Rad, Nazareth, Belgium). 6. Electroporation cuvettes (0.1 cm, Eurogentec). 7. Thermocycler and agarose gel electrophoresis equipment and UV trans-illuminator. 8. Petri dishes (round, 90 mm diameter). 9. Glass beads (VWR, Haasrode, Belgium): 2.7–3.5 mm diameter (Catalogue number 201-0087), autoclaved. 10. Wooden toothpicks (autoclaved) or sterile micropipette tips. 11. DNA Smart ladder (Eurogentec) or an equivalent DNA molecular weight marker. 12. Agarose. 13. TBE electrophoresis buffer for 1 L pH 8.3: 10.80 g Tris, 5.50 g boric acid, 0.93 g EDTA.

14. Ethidium bromide (stock solution at 1 mg/mL in H2O, stored at room temperature in the dark). 15. Analytical 1% (w/v) agarose gel: 1 g agarose in 100 mL TBE, boil in microwave to solubilize agarose, cool down to 60 C, supplement with ethidium bromide (1 μg/mL), and pour in a gel tray. 16. Oligonucleotides MP57 and GIII. Please refer to Table 1 for sequences of primers used. Primers were prepared in 20 mM in H2O. 17. Ampicillin (Sigma Aldrich), stock solution 100 mg/mL in 70% ethanol. 18. D-Glucose (stock 20% in water and ﬁltered through 0.22 μm ﬁlters (Millipore Express, Merck) for sterilization. Golden Gate Cloning of VHH Libraries 175

19. SOC medium (per 100 mL): 2.0 g peptone, 0.5 g yeast extract, 0.5 mL of 2 M NaCl, 1 mL of 250 mM KCl. Autoclave and add, just before use, 2 mL of 20% (w/v) sterile glucose, 1 mL of 1 M MgSO4 (autoclaved), and 0.5 mL of 2 M MgCl2 (autoclaved). 20. LB medium: 25.0 g LB-Broth High Salt (Duchefa Biochemie) in 1LH2O, autoclave, and store at room temperature until needed. 21. LB/AMP-GLU agar plates: 15.0 g micro agar (Duchefa Bio- chemie),25.0gLBBrothHighSaltin900mLH2O. Autoclave and cool to 60 C, add 1 mL ampicillin (stock 100 mg/mL in 70% ethanol), 100 mL D-glucose solution (stock 20% w/v). Mix and pour in 90 mm Petri plates. Store for up to 1 month in cold room. 22. dNTP mix (Thermo Fischer Scientiﬁc): 10 mM dATP, 10 mM dCTP, 10 mM dGTP, 10 mM dTTP. 23. FastStart Taq DNA polymerase (5 U/mL, Roche) supplied with 5Â PCR buffer with 20 mM MgCl2.

2.5 Construction 1. Oven/incubator at 37 C. of VHH Library 2. Shaker incubator at 37 C. 3. Centrifuge (for 50 mL Falcon tubes). 4. Spectrophotometer (visible light). 5. E. coli Pulser (electroporation instrument, Bio-Rad). 6. Electroporation cuvettes, 0.1 cm (Eurogentec). 7. Thermocycler and agarose gel electrophoresis equipment. 8. Bucket with ice. 9. Sterile Eppendorf tubes. 10. Falcon tubes (50 mL). 11. Petri dishes (round, 90 mm diameter). 12. Petri dishes (square 234 Â 234 mm). 13. Cell scraper. 14. Glass beads (VWR): 2.7–3.5 mm diameter, autoclaved. 15. Wooden toothpicks (autoclaved) or sterile micropipette tips. 16. Phage display electro-competent TG1 cells (SS320, Lucigen, USA) . 17. Recovery medium delivered with the electro-competent TG1 cells (SS320, Lucigen). 18. DNA Smart ladder (Eurogentec) or an equivalent DNA molecular weight marker. 19. Oligonucleotides MP57 and GIII. Please refer to Table 1 for sequences of primers used. Primers were prepared in 20 mM in H2O. 20. Ampicillin (Sigma Aldrich), stock solution 100 mg/mL in 70% ethanol. 176 Ema Roma˜ o et al.

21. D-Glucose (stock 20% in water and filter through 0.22 μm filters (Millipore Express, Merck) for sterilization. 22. LB medium: 25.0 g LB-Broth High Salt (Duchefa Biochemie) in 1 L H2O, autoclave, and store at room temperature until needed. 23. 2Â TY medium pH 7.4: 16.0 g Bacto-Tryptone (Duchefa Biochemie10.0 g Bacto-Yeast extract (Duchefa Biochemie), 5.0 g NaCl, bring to 1 L with water, adjust the pH to 7.4 with NaOH. Pour 330 mL per baffled shake flask of 1 L, autoclave and store at room temperature. 24. LB/AMP-GLU agar plates: 15.0 g micro agar (Duchefa Bio- chemie), 25.0 g LB Broth High Salt in 900 mL H2O. Autoclave and cool to 60 C, add 1 mL ampicillin (stock 100 mg/mL in 70% ethanol), 100 mL D-glucose solution (stock 20% w/v). Mix and pour in 90 mm Petri plates. Store for up to 1 month in cold room. 25. Large square Petri dishes (243 Â 243 mm) with LB agar containing 100 mg/mL ampicillin and 2% (w/v) glucose added from 20% stock (5 plates/library). 26. Glycerol 100% (autoclave to sterilize). 27. Agarose. 28. TBE electrophoresis buffer for 1 L pH 8.3: 10.80 g Tris, 5.50 g boric acid, 0.93 g EDTA.

29. Ethidium bromide (stock solution at 1 mg/mL in H2O, stored at room temperature in the dark). 30. Analytical 1% (w/v) agarose gel: 1 g agarose in 100 mL TBE, boil in a microwave to solubilize agarose, cool down to 60 C, supplement with ethidium bromide (1 μg/mL), and pour in a gel tray. 31. dNTP mix (Thermo Fischer Scientiﬁc): 10 mM dATP, 10 mM dCTP, 10 mM dGTP, 10 mM dTTP. 32. FastStart Taq DNA polymerase (5 U/mL, Roche) supplied with 10Â PCR buffer containing 20 mM MgCl2.

3 Methods

3.1 Amplification In these steps we will amplify the VHH sequences from the cDNA of VHH Sequences pool. The first amplification generates two distinct PCR amplicons, by PCR one of about 0.7 kb and one of about 0.9 kb [7]. The 0.7 kb PCR amplicon that originates from the heavy chain of heavy chain-only antibody encoding mRNA, is purified from a preparative agarose gel (according to the protocol described in [7]), and is then used here as a template in a nested PCR to amplify the VHH only Golden Gate Cloning of VHH Libraries 177 sequences. This nested PCR also introduces Sap I restriction enzyme sites at the 50 and 30 ends of the VHH amplicons. 1. Prepare nine PCR tubes in parallel, each containing 1 μL dNTP mix (each nucleotide at a final concentration of 0.2 mM), 1 μL CALL001 primer (0.4 μM final concentration), 1 μL CALL002 primer (0.4 μM final concentration), 0.25 μL (1.25 U) FastStart Taq DNA polymerase, 5 μL10Â PCR buffer with 20 mM MgCl2, and either 0.5, 1, 2, or 4 μL first- strand cDNA material (each of these 4 cDNA volumes is used twice to fill eight tubes) and H2O to bring the total volume in each tube to 50 μL. The cDNA is prepared as described in Vincke et al. [7]. Tube “9” is without cDNA template and serves as a negative control to ensure that PCR components are not contaminated. Put the nine tubes in a thermocycler. 2. Incubate the PCR tubes for 7 min at 95 C to activate the polymerase and to remove secondary structures from the cDNA. This step is followed by 30–35 PCR cycles (see Note 6), each cycle consisting of 60 s at 94 C, 60 s at 55 C, and 60 s at 72 C. Include a final DNA extension step for 10 min at 72 C after the last PCR cycle. 3. Apply 8 μL from each PCR tube on a 1% (w/v) analytical agarose gel in TBE buffer (with ethidium bromide) to assess the PCR amplification. Use DNA smart ladder (or equivalent) as a DNA molecular weight marker in an adjacent lane. After electrophoresis, the DNA bands are visualized on the UV trans-illuminator (according to your house rules). One band or a few bands around 0.7 kb should be present (see Note 7). 4. Pool the contents of all tubes that contain an amplicon of around 0.7 kb (ethidium bromide stained band of 0.7 kb in length is seen upon UV exposure after analytical agarose electrophoresis), then load the pooled material of the first PCR (see step 2 in Subheading 3.1) on 1% (w/v) preparative agarose gel (with ethidium bromide) in TBE buffer (see Note 8); electrophorese until the PCR products are well separated. 5. Put the agarose gel on the (clean) UV trans-illuminator and cut out immediately the PCR band of about 700 bp with a sterile scalpel or razor blade (see Note 9). 6. Extract the DNA fragments from the agarose block using the QIAquick Gel Extraction Kit. Follow the protocol recommended by the manufacturer and elute the DNA in 100 μL H2O(see Note 10). 7. Use 2–3 μL of the agarose eluted DNA to measure the DNA concentration by UV absorption at 260 nm on a Nanodrop spectrophotometer. 178 Ema Roma˜ o et al.

8. Prepare in parallel, ten tubes for nested PCR, each tube containing 1 μL dNTP mix (each nucleotide at a final concentration of 0.2 mM), 1 μL VHH-BACK-SAPI primer (0.4 μM final concentration), 1 μL VHH-FORWARD-SAPI primer (0.4 μM final concentration), 0.25 μL (1.25 U) FastStart Taq DNA polymerase, 5 μL5Â PCR buffer with 20 mM MgCl2, 10–50 ng of the purified first PCR product (e.g., three tubes with 10 ng, 3 tubes with 20 ng, 3 tubes with 40 ng and 1 tube without PCR product as a control) and H2O to bring the total volume in each tube to 50 μL. The ten tubes are put in the thermocycler. 9. Incubate the PCR tubes for 7 min at 95 C to denature the DNA template and activate the polymerase. Then proceed the PCR by 17–20 PCR cycles (see Note 11) each cycle consisting of 45 s at 94 C, 45 s at 55 C, and 45 s at 72 C. Include a final DNA extension step for 10 min at 72 C after the last PCR cycle. 10. Apply an aliquot of 8 μL from each of the ten PCR tubes, separately on a 2% (w/v) analytical agarose gel (with ethidium bromide) in TBE buffer to assess the PCR amplification. Use DNA smart ladder (or equivalent) as DNA molecular weight marker in an adjacent lane. After electrophoresis, the DNA bands are visualized on the UV trans-illuminator (according to house rules). Confirm the presence of a PCR amplicon of about 400 bp corresponding to the size of VHHs including the surrounding restriction enzyme sites and clamp sequences. Only one single band should be visible and no band should be present in the negative control, the tube where the first PCR product was omitted. 11. Pool the samples from PCR tubes where a DNA fragment of proper size was amplified and purify the PCR amplicon using QIAquick PCR Purification Kit according to the guidelines of the manufacturer. Elute the DNA in 200 μLH2O. 12. Measure the DNA concentration by UVabsorption on a Nano- drop spectrophotometer. The purified nested PCR product is used immediately for the golden gate restriction and ligation (see Subheading 3.3) or can be stored for prolonged times at À20 C.

3.2 Home-Made 1. Streak the TG1 cells (from glycerol stock at À80 C freezer) on Electro-competent TG1 a Minimal medium plate. Incubate the plate overnight in an Cells oven at 37 C. 2. The next day, bring 5 mL 2Â TY in a 50 mL Falcon tube and inoculate with a single colony of E. coli TG1, from the fresh plate. Shake at 250 rpm, overnight at 37 C. Golden Gate Cloning of VHH Libraries 179

3. Take a baffled flask with 300 mL 2Â TY medium and inoculate with 2 mL of the overnight TG1 culture. Shake at 250 rpm at 37 C until OD600nm is between 0.8 and 1.0 (this takes about 3–4 h). 4. Put the culture on ice for 1 h in the cold room. Put six Falcon tubes (or sterile centrifuge tubes) on ice as well. Refrigerate the centrifuge at 4 C. 5. Bring the TG1 culture into the ice-cold centrifuge tubes. Pellet the cells in the refrigerated centrifuge at 4 C for 7 min at 2200 Â g. 6. Decant the supernatant and gently resuspend the cell pellet in the original culture-volume of ice-cold 1 mM HEPES, pH 7.0. 7. Centrifuge bacteria at 2200 Â g for 6 min at 4 C. 8. Decant the supernatant very carefully, gently resuspend the cell pellet in half of the original culture-volume of ice-cold 10% (v/v) glycerol, and centrifuge at 4 C for 6 min at 2200 Â g (see Note 12). 9. Decant the supernatant, gently resuspend the cell pellet of each tube in 10 mL of ice-cold 10% (v/v) glycerol, and pool the resuspended cells in one or two 50 mL Falcon tubes (see Note 12). 10. Centrifuge at 4 C for 5 min at 2200 Â g. 11. Decant the supernatant very carefully, and gently resuspend the cells in ice-cold 10% (v/v) glycerol to a final volume of 1 mL (see Note 12). Bring 50 μL aliquots of cell suspension in sterile Eppendorf tubes and use the cells immediately as described (see Subheading 3.4).

3.3 Restriction The pMECS-GG vector is a modification of the pMECS vector Enzyme Digestion where a “killer cassette” flanked by Sap I restriction enzyme sites of VHHs and Ligation was introduced within the multiple cloning site (Fig. 1). Apart from to pMECS-GG a chloramphenicol resistance gene for positive selection in CcdA containing Survival cells (Thermo Fischer Scientific), this killer cassette encodes the CcdB protein that will kill host cells such as TG1 cells that lack CcdA. The Survival cells are used to prepare the pMECS-GG phagemid DNA. Mixing the pMECS-GG phagemid and the PCR amplicon followed by an incubation with Sap I will remove the enzyme’s recognition sequence from the amplicon and create three nucleotide overhangs (Fig. 2). On the plasmid, Sap I cuts the stuffer fragment with the chloramphenicol and ccdB gene, the remaining phagemid will have three nucleotide overhangs that are complementary to the overhangs on the cut PCR amplicon of the VHH. Ligation of this chimeric pMECS to VHH gene will create stable 180 Ema Roma˜ o et al.

Fig. 2 Cloning strategy to generate a high-quality VHH library in phage-display phagemid. The pMECS-GG phagemid vector is mixed with the PCR fragment containing the VHH (¼Nanobody) and incubated with SapI restriction enzyme and T4 DNA ligase. The SapI has two recognition sequences on the phagemid (shown in red) and cuts with a three nucleotides overhang as indicated with the red arrowheads. This cutting removes the stuffer fragment containing the chloramphenicol resistance and ccdB genes. The PCR amplicon is also cut with SapI and generates three nucleotide overhangs compatible for ligation into the cut phagemid vector. Ligation of the pMECS and Nanobody generates a chimeric pMECS-Nb phagemid where the Nanobody gene is in frame with the pel B leader sequence and upstream the NotI restriction enzyme site and the HA-tag, the His6 tag and the gene III

chimeric phagemids without SapI restriction recognition sequence that will be transformed in TG1 bacteria. 1. Prepare three tubes for SapI cutting and ligation reaction. These tubes are filled with 1.0, 0.6, or 0.3 μg of the purified nested PCR product and 3.0 μg of pMECS-GG vector (see Note 13). Add 20 μL Sap I (200 U), 10 μL10Â CutSmart Buffer, 2.5 μL T4 DNA Ligase (15 U), 5 μL 10 mM ATP and bring to a final volume to 100 μL with H2O. Mix, spin briefly, and put the three tubes in a thermocycler. 2. Digest and ligate the nested PCR product to the pMECS-GG vector by 18 cycli of 30 min at 37 C and 30 min at 18 C, Golden Gate Cloning of VHH Libraries 181

followed by a ﬁnal ligation step of 60 min at 18 C and enzyme inactivation step by incubating the tubes for 10 min at 50 C and 10 min at 80 C. 3. Add an equal volume (100 μL) of TE-saturated phenol (lower phase), vortex and spin for 10 min at 18,000 Â g to separate the phases. 4. Transfer the upper aqueous phase to a fresh microcentrifuge tube. Add 100 μL chloroform/isoamyl alcohol (24/1 ratio). Vortex and spin for 10 min at 18,000 Â g. 5. Transfer the upper aqueous phase to a fresh microcentrifuge tube and add 20 μL 3 M sodium acetate (pH 5.2), mix by pipetting in and out. Add 250 μL absolute ethanol, close the lid well, and invert the tubes a few times to mix. Incubate for at least 30 min at À80 C(see Note 14). 6. Put the tube in the microcentrifuge and spin for 20 min at 18,000 Â g. Remove carefully all liquid with a micropipette (see Note 15). Air-dry the pellets and dissolve DNA in 50 μL H2O. The dissolved DNA is ready for electroporation (see Subheading 3.4) or can be stored at À20 C.

3.4 Test Ligation, A test transformation is optional but recommended for occasional Test Transformation, users to assess the quality of puriﬁed material of each of the three and Initial Quality conditions of pMECS-GG to VHH amplicon ligation. This is Control achieved by monitoring the number of transformants and a rapid quality control to verify the presence of an insert with the size of a VHH within the phagemid of the transformants. 1. Set the electroporation apparatus “E. coli Pulser” (Bio-Rad) at 1.8 kV, and put 20 electroporation cuvettes on ice. 2. Mix 1 μL of the previously prepared and puriﬁed ligation reaction (from step 6 in Subheading 3.3) for each molar ratio of insert to vector ligation with 50 μL of freshly made electrocompetent E. coli TG1 cells (from step 11 in Subheading 3.2). Place on ice for at least 1 min. 3. Remove the ice from the outside of the electroporation cuvette with tissue paper, place the cuvette in the electroporation chamber, add 50 μL of the electrocompetent cells mixed with ligation product, and apply a pulse of 1.8 kV. 4. Remove the cuvette immediately and add 0.5 mL of the SOC medium, and transfer the cell suspension to an Eppendorf tube. Rinse the cuvette once more with 0.5 mL SOC medium and pool with the cell suspension in the Eppendorf tube. 5. Repeat steps 3 and 4 for all the ligation reactions previously prepared. 6. Shake the electroporated cells at 200 rpm for 1 h at 37 C. 182 Ema Roma˜ o et al.

7. Plate 100 μL of each transformation from the incubated electro-transformed cells on (90 mm) LB/AMP-GLU agar plates. Add a few glass beads to the plates and shake to roll the glass beads over the agar, to spread the cells. Invert the Petri dishes (so that glass beads fall into the lid) and incubate overnight at 37 C. 8. The next day, monitor the number of colonies on the various plates (see Note 16). The ligation mixture that gives the highest number of colonies will be used to generate the library, complemented with a fraction of the ligation mixture that gives the second highest number of colonies. 9. Make a mastermix by combining 12.5 μL dNTP mix (final concentration of 0.2 mM for each nucleotide), 25 μL MP57 (0.4 μM final concentration) and 25 μL GIII primer (0.4 μM final concentration), 6.25 μL (1.25 U) FastStart Taq DNA polymerase, 250 μL5Â PCR buffer, and bring the total volume to 1250 μL with H2O(see Note 17). Dispense 25 μL of this mastermix in 49 PCR tubes. 10. Take a Petri dish (from step 8 in Subheading 3.4) with colonies that are well spread to perform a colony PCR. Touch a single colony with a sterile toothpick (or sterile micropipette tip) and stir in the reaction mix of a single PCR tube. Repeat this handling for another 48 tubes picking 16 colonies from each ligation molar ratio plated. Tube number “49” is a negative control (no colony added, to detect possible contamination of reagents by template DNA). 11. Put all 49 tubes in the thermocycler. Incubate for 5 min at 95 C to lyse bacterial cells, denature DNA, and activate the polymerase enzyme. This step is followed by 28 PCR cycli, each cycle consisting of 45 s at 94 C, 45 s at 55 C, and 45 s at 72 C. After 28 cycli, perform a final extension step for 10 min at 72 C. 12. Use 8 μL of each PCR tube to analyze the amplicons via electrophoresis on an analytical 1% (w/v) agarose gel with ethidium bromide in TBE buffer. Put the gel on a UV transilluminator to determine the number of colonies out of 48 with the right insert size for a VHH. Usually, all clones of all the ligation reactions yield a PCR fragment with the size of a VHH (700 bp) within the phagemid.

3.5 Construction After monitoring which VHH/phagemid mixture ligation mixture of VHH Library is giving highest number of transformants (in our hands, usually the 3–1 ratio) a larger amount of electrocompetent E. coli TG1 cells are transformed with the purified ligation mixture and plated on selective medium to generate a VHH library of about 107 to >109 individual transformants. For the final library we recently switched Golden Gate Cloning of VHH Libraries 183 to commercial TG1 E. coli cells as they give more consistent results and are nowadays of good quality as well. 1. Set the electroporation apparatus E. coli Pulser (Bio-Rad) at 1.8 kV, and put 20 electroporation cuvettes on ice. 2. Prepare 12 tubes each with 6 μL of the cleaned ligation reaction (step 6 and Subheading 3.3) from the insert to vector ligation ratio that is yielding the highest and second highest number of transformants (determined from step 8 in Subheading 3.4) and 25 μL of electrocompetent E. coli TG1cells (SS320, Lucigen). Place on ice for at least 1 min. 3. Remove the ice from the outside of an electroporation cuvette with tissue paper, place the cuvette in the electroporation chamber, add 25 μL of the electrocompetent cells mixed with ligation product, and apply a pulse of 1.8 kV. 4. Remove the cuvette immediately from the holder, add 0.5 mL Recovery medium and transfer the cell suspension to a 50 mL Falcon tube. Rinse the cuvette once more with 0.5 mL Recov- ery medium and pool with the cell suspension in the Falcon tube. 5. Repeat steps 3 and 4 until 12 electroporations are performed. All electroporated E. coli cells can be pooled in one (or two) 50 mL Falcon tube(s). 6. Shake the pooled electroporated cells at 200 rpm for 1 h at 37 C. 7. Put 100 μLofa103-, 104-, and 105-fold diluted aliquot of the incubated electro-transformed cells on (90 mm) LB/AMP- GLU agar plates. Add a few glass beads to the plates, shake to roll the beads over the agar and to spread the cells evenly. Invert the Petri dishes (so that glass beads fall into the lid) and incubate overnight at 37 C. These plates are used to calculate the transformation efficiency. 8. The remaining of the suspension of electroporated cells is poured on five large square LB-agar plates containing 100 μg/mL ampicillin and 2% (w/v) D-glucose (maximally 3 mL cell suspension/large plate). Add glass beads to the plates and shake to spread the cells over the agar plate. Invert the Petri dishes (so that glass beads fall into the lid) and incubate overnight at 37 C. 9. The next day, calculate the library size from the number of colonies on the 90 mm plates (from step 7 in Subheading 3.5), taking into account the dilution, etc. (see Note 18). 10. Scrape the cells from the 243 Â 243 mm square plates using 3 mL LB medium per plate and a sterile cell scraper. Transfer the cell suspension with a micropipette to a 50 mL Falcon tube. 184 Ema Roma˜ o et al.

Rinse the large plates with an extra 2 mL LB medium and combine the cell suspensions in the Falcon tubes. Pellet the cells by 10 min centrifugation at 2200 Â g. 11. Decant the supernatant and resuspend the cell pellet in the LB medium to a final volume of 20–30 mL. Add glycerol (from autoclaved 100% stock) to a final concentration of 15%, and mix well. Make five aliquots of 1 mL of the library in Eppen- dorf tubes, the rest is stored in the large Falcon tubes. Store the Eppendorf tubes and the Falcon tubes containing the library at À80 C. Before freezing, take a small aliquot of the cell suspension in the Falcon tube and dilute about 100 times in 2Â TY. Measure the OD600nm in a visible light spectrophotometer (not with Nanodrop) to estimate the total numbers of cells within the library scraped from the plates (see Note 19). 12. To confirm that all clones possess a VHH insert, we perform again a PCR with primers GIII end MP57 as explained in steps 9–12 in Subheading 3.4, except that the colonies are taken from the plates described in step 9 (Subheading 3.5)(see Notes 20 and 21).

4 Notes

1. The CALL001 primer anneals to the template strand of leader- signal sequences of camelid VH and VHH genes, homologous to human family-3 VH genes [9, 10]. 2. The CALL002 primer anneals to a conserved part within the coding strand of CH2-sequence of all camelid IgG heavy chains. This oligonucleotide can be used to anneal on mRNA of camelid IgG heavy chain and can be used as a primer in ﬁrst- strand cDNA synthesis. 3. The VHH-BACK-SAPI primer anneals with its 30 end at the template strand for codons 2–10 of VHH genes of family-3. The 50 end of this primer consists of a clamp sequence (CTTG) and the Sap I recognition sequence (GCTCTTC). 4. The VHH-FORWARD-SAPI primer hybridizes with the coding strand at the last seven codons of the VHH (encoded by the J-minigene). This part of the PCR primer also harbors the restriction enzyme recognition sequence of BstEII (GGTNACC). At its 50 end the primer contains a clamp site (TGAT) to facilitate restriction enzyme digestion and the SapI recognition sequence (GCTCTTC). 5. The pMECS-GG vector is a derivative of the pMECS phage display vector [7] to make it compatible for the Golden Gate cloning strategy. The multiple cloning site within the pMECS to clone the camelid VHH gene was substituted to include two Golden Gate Cloning of VHH Libraries 185

Sap I restriction enzyme sites, the gene to confer chloramphenicol resistance for positive selection in bacteria and a ccdB gene encoding the plasmid addiction protein, toxic for bacteria not expressing the antidote ccdA gene. The pMECS-GG plasmid can be purified from transformed ccdB Survival Cells (Thermo Fischer Scientific). The plasmid is purified with the Qiafilter Plasmid Midi kit (or an equivalent plasmid purification kit). The phagemid DNA is stored in H2O or in TE (10 mM Tris–HCl pH 8.0, 1 mM EDTA) at 1 ng/μLat4C for short periods or at À20 C for longer periods until further use. 6. Routinely we use 32 PCR cycli, however, if we fear that cDNA will not be of good quality we occasionally employ up to 35 cycli to obtain enough PCR amplicons. 7. The bands around 0.7 kb are all from the heavy chain-only antibodies. If a good separation is obtained on agarose gel then a few closely spaced bands around 0.7 kb might be visible, due to the length difference in the hinge region of the various heavy chain-only immunoglobulin isotypes. In any case the band around 0.9 kb is from the heavy chain of classical antibodies. Sometimes, this band appears to be absent or invisible, probably due to a bias during the PCR amplification. 8. The analytical and preparative agarose gels differ from each other in thickness of the gel (thicker gels and wider slots that can accommodate larger sample volumes are used for the preparative gels). 9. Try to keep the exposure of ethidium bromide stained DNA bands to UV light to a minimum to avoid DNA damaging that might subsequently resist PCR amplification and/or cloning. 10. The agarose gel-purified first PCR product can be used immediately as a template for the nested PCR, or it can be stored overnight in the cold room to continue the day after, or it can be stored at À20 C if you wish to continue after a longer period of time. 11. Usually, we perform 17 cycli to start and then test the amplification product on an analytical agarose gel. If the amplification yield is too low, we increase the number of cycli to 20. 12. We use disposable 1 mL micropipette tips of which the end is cut off to have a larger opening so as to reduce shearing forces during pipetting to resuspend the cells gently. 13. These amounts of vector and PCR amplicon correspond to a molar ratio of insert to vector of 3:1, 2:1, and 1:1, respectively. 14. To precipitate the DNA with ethanol, it is also possible to put the tube on dry ice for 5–10 min instead of >30 min at À80 C. 186 Ema Roma˜ o et al.

15. Normally, this little amount of DNA cannot be seen if it is clean. Remove the liquid water/ethanol with a 1 mL micropipette. Give the tube a brief spin to collect the liquid from the edge of the tube to bottom of the tube. Take away the last traces of ethanol/water mixture with a 20 μL pipette without touching or disturbing the invisible DNA pellet. 16. Since the amount of phagemid was kept constant in the three ligation mixtures with different insert to phagemid ratios, the number of colonies on the plate gives a direct indication of which ratio will give the library of highest number of transformants. In principle, all transformants should have an insert as original phagemid is toxic for TG1 cells. 17. The MP57 primer anneals upstream the pelB and VHH insertion site, whereas GIII anneals in the gene III of M13 within the phage display vector and downstream the VHH-HA-His tag. A PCR with these primers will yield an amplicon of about 700 bp when a VHH has been inserted in the phage display vector and replacing the chloramphenicol resistance and ccdB genes. 18. From the small Petri dishes we calculated the total number of transformants within our library.

19. From the OD600nm turbidity measurement of the suspension of cells scraped from the large plates, we can estimate the total number of cells within the library. An OD600 nm of 1.0 corresponds to 8 Â 108 TG1 bacteria/mL and the total volume of cells scraped from the plates is also known. The total number of cells in the glycerol stock divided by the number of transformants (calculated from the number of colonies on small plates) gives the library amplification number, i.e. the average times each individual transformed cell is represented within the library. 20. Again, cells can only have an (putative) VHH insert, as TG1 cells transformed with the original pMECS-GG vector are unable to grow due to the toxic effect of CcdB. However, some colonies could have their chloramphenicol resistance and ccdB genes replaced with a shorter or longer non-VHH insert. 21. A few clones can be sequenced to ensure that all VHH inserts are in frame and have different CDRs. However, this is not highly relevant as only a very, very minor fraction of the total library can be sequenced, which has barely a predictive value of the diversity of the library. For example sequencing 50–100 clones is not representative for a library of 107 individual transformants; in contrast, finding two identical clones within a sampled fraction of 50 clones from 107 individual transformants or so indicates that the cDNA preparation and/or PCR amplification have been severely biased toward very few sequences. However up to now, we never observed such a biased library. Golden Gate Cloning of VHH Libraries 187

References

1. Muyldermans S (2013) Nanobodies: natural 7. Vincke C, Gutie´rrez C, Wernery U, single-domain antibodies. Annu Rev Biochem Devoogdt N, Hassanzadeh-Ghassabeh G, 82:775–797 Muyldermans S (2012) Chapter 8: Generation 2. Hamers-Casterman C, Atarhouch T, of single domain antibody fragments derived Muyldermans S, Robinson G, Hamers C, from camelids and generation of manifold con- Songa EB, Bendahman N, Hamers R (1993) structs. In: Chames P (ed) Antibody engineer- Naturally-occurring antibodies devoid of light- ing: methods and protocols, second edition. chains. Nature 363:446–448 Methods in molecular biology, vol 917. 3. Greenberg AS, Avila D, Hughes M, Hughes A, Humana, Louisville, KY, pp 145–176 McKinney EC, Flajnik MF (1995) A new anti- 8. Pardon E, Laeremans T, Triest S, Rasmussen gen receptor gene family that undergoes rear- SGF, Wohlko¨nig A, Ruf A, Muyldermans S, rangement and extensive somatic Hol WGJ, Kobilka BK, Steyaert J (2014) A diversification in sharks. Nature 374:168–173 general protocol for the generation of Nano- 4. Dmitriev OY, Lutsenko S, Muyldermans S bodies for structural biology. Nat Protoc (2016) Nanobodies as probes for protein 9:674–693 dynamics in vitro and in cells. J Biol Chem 9. Nguyen VK, Hamers R, Wyns L, Muyldermans 291:3767–3775 S (2000) Camel heavy-chain antibodies: 5. Engler C, Gruetzner R, Kandzia R, Marillonet diverse germline V(H)H and specific mechan- S (2009) Golden gate shuffling: a one-pot isms enlarge the antigen-binding repertoire. DNA shuffling method based on type IIs Eur Mol Biol Organ J 19:921–930 restriction enzymes. PLoS One 4:e5553 10. Deschacht N, De Groeve K, Vincke C, Raes G, 6. Bernard P, Gabant P, Bahassi E, Couturier M De Baetselier P, Muyldermans S (2010) A novel (1994) Positive-selectionvectors using the F promiscuous class of camelid single-domain plasmid ccdB killer gene. Gene 148:71–74 antibody contributes to the antigen-binding repertoire. J Immunol 184:5696–5704 Chapter 10

Construction of Chicken Antibody Libraries

Jeanni Fehrsen, Susan Wemmer, and Wouter van Wyngaardt

Abstract

Recombinant antibody libraries based on chicken immunoglobulin genes are potentially valuable sources of phage displayed scFvs for use in veterinary diagnostics and research. The libraries described in this chapter are based on chicken variable heavy and light chain immunoglobulin genes joined by a short ﬂexible linker cloned in the phagemid vector pHEN1. The resulting phagemids produce either scFvs displayed on the surface of the fusion phage subsequent to co-infection with helper phage, or soluble scFvs following IPTG induction. This chapter provides detailed and proven methods for the construction of such libraries.

Key words Chicken antibody library, scFv, Naive, Immune, Phagemid, Recombinant antibodies, Phage display

1 Introduction

Recombinant antibody libraries derived from chicken immunoglobulin genes can be used as a source of diagnostic and research reagents. Accessing the chicken immunoglobulin repertoire is relatively easy since all the immunoglobulin variable heavy (VH) and all 0 0 the variable light (VL) chain sequences are identical at their 5 and 3 ends. This implies that they can be amplified using only two sets of PCR primers [1–3]. Immunoglobulin diversity is generated in the bursa of Fabricius by gene conversion using pseudogene variable regions. This occurs in the first 4 months; therefore, the naive immunoglobulin repertoire can be accessed from avian bursal cells. Alternatively, chickens can be immunized with the antigens of interest and their spleens, blood lymphocytes, and/or bone marrow used as source of immunoglobulin mRNA [4–7]. The single-chain fragment variable (scFv) form of recombinant antibodies is described here in which the immunoglobulin VH region and the VL region are joined via (Gly4Ser)3 flexible linker (see Fig. 1)[8]. The phagemid vector used enables the scFv to be either displayed as a fusion protein on the phage, or alternatively expressed as a soluble scFv protein [9]. Diversity of the naive

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_10, © Springer Science+Business Media LLC 2018 189 190 Jeanni Fehrsen et al.

Fig. 1 Flow diagram of the process to convert mRNA to the scFv gene construct using the two component method. Primers are shown by solid lines and the arrows indicate the direction of extension

repertoire could be increased by using primers to randomize the complementarity determining regions (CDRs) [10, 11]. Shortening of the ﬂexible linker between the VH and VL regions results in the formation of dimers or higher order multimers and consequently the production of higher avidity binding entities [5, 12]. Two strategies are described to link the variable domains of the heavy and light chains. In the ﬁrst strategy, portions of the (Gly4- 0 0 Ser)3 linker are added to the 3 end of the VH and the 5 end of the Construction of Chicken Antibody Libraries 191

Fig. 2 Expected translation of the scFv gene construct. Positions of primers (from Table 1 and Fig. 1) are indicated by the lines below. “X” indicates amino acids encoded by randomized codon (MNN). n ¼ number of (MNN) codons to encode the synthetic CDR3. Amino acids in gray are removed by the restriction enzymes before cloning into the vector

VL using appropriate PCR primers. Splicing by overlap extension (SOE) [13] is then used to assemble the gene construct coding for the scFv (Fig. 1). This method is usually used for naive and immune libraries. The second option incorporates synthetically randomized VH CDR3s. The linker is ampliﬁed as a third component using 0 0 primers complementary to the 3 end of the VH chain and the 5 end of the VL chain (Fig. 2). An extended primer complementary to 0 the 3 end of the VH chain adds the synthetic CDR3. The scFv gene construct is then assembled using a three-component SOE. Most of the methods described here, with some adjustments, are based on previously published work that has already produced scFvs that have proven useful in veterinary diagnostics and research [5, 6, 11, 12, 14–18].

2 Materials (See Note 1)

2.1 RNA Isolation 1. RNA as source of the VH and VL genes (see Note 3). (See Note 2) 2. Invitrogen™ RNAlater™ Stabilization Solution (Thermo- Fischer Scientiﬁc Inc., Waltham, USA). 3. DEPC (Diethylpyrocarbonate). 4. RNase-free plastic ware. Scalpels, syringes, petri dishes, ﬁlter tips, centrifuge tubes. 5. Fine stainless-steel sieve. ® 6. TRI Reagent (Sigma-Aldrich, St. Louis, USA). 7. BCP (1-bromo-3-chloro propane). 8. Isopropanol. 9. Ethanol (dilute to 75% v/v with nuclease-free water). 10. Nuclease-free water.

2.2 cDNA Synthesis, 1. Thermocycler. PCR, and Ligation 2. RT- PCR kit (TaKaRa RNA PCR Kit (AMV) Ver 3.0. TaKaRa Bio Inc., Shiga, Japan). 3. Primers (see Table 1). 192 Jeanni Fehrsen et al.

Table 1 Nucleotide sequence of primers

Primera Nucleotide sequence (50–30)

Variable heavy chain VHF GTCCTCGCAACTGCGGCCCAGCCGGCCCTGATGGCGGCCGTGACG VHR CCGCCTCCGGAGGAGACGATGACTTCG Variable light chain VLF GACTCAGCCGTCCTCGGTGTCAG VLR TGATGGTGGCGGCCGCATTGGGCTG

Primers to add (Gly4Ser)3 linker NewVarH CCGCCAGAGCCACCTCCACCTGAACCGCCTCCACCGGAGGAG ACGATGACTTCGG NewVarL TCAGGTGGAGGTGGCTCTGGCGGAGGCGGATCGGCGCTGACTC AGCCGTCCTCGG Primers for randomised synthetic heavy chain CDR3

RandVH GACTTCGGTCCCGTGGCCCCATGCGTCGAT[MNN]n TTTGGCGCAGTAGTAGGTGCCGGTGTCCTC Primers to amplify Gly-Ser linker GSfor GGGGCCACGGGACCGAAGTC GSrev CGCTGACACCGAGGAC Sequencing primers OP52 CCCTCATAGTTAGCGTAACG M13R CAGGAAACAGCTATGAC aPrimers were derived from [1–3] and as published by Wyngaardt et al. [11]. Restriction enzyme sites and extra bases to enable restriction digestion are underlined. M ¼ A/C, N ¼ A/C/G/T. n ¼ length of desired CDR3

4. DNA polymerase (TaKaRa Ex Taq HS, Clontech). 5. High-ﬁdelity DNA polymerase (Pfu DNA polymerase, Pro- mega, Madison, USA). 6. Agarose. 7. 50Â TAE buffer: 2 M Tris, 1 M acetate, 50 mM EDTA. Dissolve 242 g Tris in 700 ml ddH2O, add 57.1 ml glacial acetic acid and 100 ml 0.5 M Na2EDTA (pH 8) and make up to 1L. 8. Crystal violet: 10 mg/ml stock solution. 9. Loading buffer (6Â) for crystal violet gels: 2% Ficoll 400 and 0.002% xylene cyanol in ddH2O. 10. Gel extraction kit (Qiagen, Hilden, Germany). Construction of Chicken Antibody Libraries 193

11. PCR purification kit (Qiagen, Hilden, Germany). 12. Phage display vector; phagemid pHEN1 [9]. 13. Sfi and NotI (Roche, Penzberg, Germany). 14. 2.5 M NaCl. 15. 1 M Tris (pH 8). 16. Acetylated BSA (10 mg/ml). 17. Enzyme reaction purification kit (Qiagen, Hilden, Germany). 18. T4 DNA ligase (Roche, Penzberg, Germany).

2.3 Electroporation 1. Electroporator (Biorad Laboratories Gene Pulser II electro- and Growth of E. coli porator, Hercules, USA). and Bacteriophage 2. 0.1 cm electroporation cuvette. 3. Incubator at 37 and 30 C. 4. E. coli TG1 electroporation-competent cells (Agilent Technol- ogies, Santa Clara, USA, see Note 4). 5. SOC medium: 20 g tryptone, 5 g yeast extract, 0.5 g NaCl, 0.186 g KCl. In 970 ml ddH2O. Autoclave for 20 min. Add 10 ml 1 M MgCl2 and 20 ml 1 M glucose before use. 6. TYE agar: 10 g tryptone, 5 g yeast extract, 8 g NaCl, 15 g agar in 900 ml ddH2O. Autoclave for 20 min and cool to 50 C. Add 100 ml of 20% v/v glucose, 1 ml of 100 mg/ml ampicillin before pouring plates. 7. 2ÂTY: 16 g tryptone, 10 g yeast extract, 5 g NaCl. Dissolve in ddH2O to obtain a ﬁnal volume of 1 l. Autoclave for 20 min. 8. Helper phage M13 K07. 9. Petri dishes, 9 and 15 cm. 10. 60% v/v glycerol. 11. 20% w/v glucose: 20 g D-glucose, add glucose slowly to 70 ml ddH2O, when dissolved make up to 100 ml. Autoclave or ﬁlter sterilize. 12. PEG/NaCl: 20% w/v polyethylene glycol 6000 or 8000, 2.5 M NaCl. Autoclave for 20 min.

13. 10Â PBS: 1.37 M NaCl, 27 mM KCl, 80 mM Na2H- PO4∙2H2O, 18 mM KH2PO4. Dissolve 80 g NaCl, 2 g KCl, 14,4 g Na2HPO4.2H2O, 2.4 g KH2PO4 in 800 ml ddH2O, adjust to pH 7.4 and ﬁnal volume of 1 l. Autoclave for 20 min. Alternatively use PBS tablets. 194 Jeanni Fehrsen et al.

3 Methods

3.1 RNA Isolation This is the source of the immunoglobulin genes. Total RNA is isolated and the VH and VL genes are amplified using specific primers after the mRNA in the pool is transcribed to cDNA. 1. Harvest bursa or spleen from chicken, cut into 0.5 cm pieces and immediately place in RNAlater™ until ready to proceed. 2. Cut tissue into small pieces. 3. Place tissue fragments into 10 ml TRI Reagent® and homogenize into a petri dish by forcing through a fine stainless-steel sieve using the plunger of a 20 ml disposable syringe as a pestle. Add more TRI Reagent® if needed up to 30 ml. Transfer to a 50 ml disposable tube. 4. Centrifuge for 10 min at 1500 Â g to remove the cell debris. 5. Transfer supernatant fluid (SNF) to a clean 50 ml tube, add 3 ml BCP, vortex and incubate at room temperature for 15 min. 6. Centrifuge at 2000 Â g for 15 min at 4 C. 7. Transfer the clear aqueous phase to a clean centrifuge tube, add 15 ml isopropanol, vortex and incubate at room temperature for 10 min to precipitate the RNA. 8. Centrifuge at 10000 Â g for 30 min at 4 C to recover the RNA. 9. Discard the SNF, wash pellet with 75% ethanol and centrifuge at 10000 Â g for 30 min at 4 C. 10. Air-dry the pellet at room temperature for 10 min and dissolve in 500 μl nuclease-free water at 55–60 C. 11. Remove 5 μl for analysis. Determine the concentration spectrophotometrically where an OD of 1 at 260 nm ¼ 40 μg/ml RNA. Store aliquots at À80 C.

3.2 cDNA Synthesis 1. Synthesize cDNA by reverse transcription. Prepare 20 separate and Amplification 20 μl reactions to ensure a diverse mixture of mRNAs is by PCR amplified. 2. Each 20 μl reaction (using the TaKaRa RNA PCR Kit) consists of 5 mM MgCl2, 1Â RT buffer, 1 mM dNTPs, 20 units RNase inhibitor, 5 units Reverse transcriptase, 0.125 μM Oligo dT-Adapter primer and 1 mg total RNA. 3. Synthesize cDNA in a thermocycler using the following conditions: 30 C for 10 min, 42 C for 1 h, 95 C for 5 min, and 5 C for 5 min. 4. The cDNA is then converted to dsDNA using primers specific for the chicken VH and VL genes (see Table 1 and Note 5). Ten Construction of Chicken Antibody Libraries 195

Fig. 3 Agarose gel electrophoresis showing the expected sizes of chicken the VH chain, VL chain and the scFv gene construct. VH 400 bp, VL 300 bp and scFv 800 bp

reverse transcription reactions (step 2) are used in conjunction with each set of primers. An individual reaction (using the TaKaRa RNA PCR kit) consists of 20 μl cDNA, 1Â PCR buffer, 2.5 units TaKaRa Ex Taq HS, 0.4 μM of each primer (VHF and VHR or VLF and VLR) and ddH2O up to 100 μl (see Subheading 3.4 for primer options to generate synthetic randomized VH CDR3s). 5. Synthesize dsDNA in a thermocycler using the following conditions: 30 cycles: 94 C for 1 min, 60 C for 1 min, 72 C for 1 min, followed by a ﬁnal extension at 72 C for 3 min.

6. The VH and VL PCR products are analyzed by electrophoresis on a 1.5–2% agarose gel (Fig. 3). Usually, there are additional amplicons present. These are removed by selectively purifying the desired VH or VL amplicons from the agarose gel. Crystal violet stained gels are used for this purpose as described below (also see Subheading 3.11).

7. Precipitate the VH and VL PCR products with 1 volume of isopropanol to concentrate the DNA. Centrifuge at 10000 Â g for 10 min. Wash the pellet with 70% ethanol, spin at 10000 Â g for 10 min. Remove all the ethanol, air-dry the pellet for 10 min. Resuspend the DNA pellet in approximately one tenth volume ddH2O (this may take some time). 8. Load on a 2% agarose gel containing 10 μg/ml crystal violet (see Subheading 3.11). 9. Cut out the selected amplicons and purify the DNA fragments using an agarose gel extraction kit.

10. Quantify DNA spectrophotometrically at OD260nm. 196 Jeanni Fehrsen et al.

3.3 Assembly of scFv 1. Segments of the (Gly4Ser)3 linker are added to the VH and VL Genes Based on gene fragments with primers NewVarH and NewVarL. This 0 Natural VH and VL results in overlapping regions on the 3 end of the VH with 0 the 5 end of VL. 2. Each 100 μl reaction contains: 1Â PCR buffer, 0.8 mM dNTPs, VH DNA (500 ng), VL DNA (500 ng), 0.2 μMof each primer (NewVarH and NewVarL), 3 units Pfu polymerase enzyme (Promega), 2.5 units TaKaRa Ex taq HS. Prepare 2–4 reactions. 3. Use the following conditions: 94 C for 2 min, then 15 cycles: 94 C for 1 min, 60 C for 1 min, 72 C for 1 min and a ﬁnal extension at 72 C for 5 min.

4. This is followed by joining of the VH and VL chains via the overlapping regions by SOE. Primers VHF and VLR are added to complete and amplify the joined products. The amount of product from step 3 requires optimization for the SOE. Initi- ally use 2 and 4 μl and evaluate the product on an agarose gel. Select the conditions that result in the most prominent amplicon at around 800 bp (Fig. 3) and repeat using 40 reactions. 5. Each reaction (using TaKaRa Ex Taq HS) consists of: 1Â PCR buffer, 0.8 mM dNTPs, 2 or 4 μl DNA from step 3, 0.2 μMof each primer (VHF and VLR), 2.5 units Ex taq HS and ddH2O to a final volume of 100 μl. 6. Use the following conditions: 30 cycles of 94 C for 1 min, 60 C for 1 min, 72 C for 1 min. Followed by the final extension at 72 C for 5 min. 7. The reaction mixtures are pooled and an aliquot is analyzed on a 1% agarose gel. If a single amplicon at 800 bp is observed, purify up the product using a PCR purification kit. If multiple amplicons are present, concentrate the sample by precipitation and gel purify as described above (see Subheading 3.2, steps 7–10). The scFv gene constructs are now ready for digestion with SfiI and NotI. Concurrently digest the vector (pHENI) using the same enzymes.

3.4 Assembly of scFv 1. To construct a library containing synthetically randomized VH Genes Based on CDR3s, the linker, VH and VL genes are ampliﬁed separately

Natural VL and prior to joining all three components by overlap extension.

Synthetic, Randomized 2. The VL gene is prepared as described in Subheading 3.2. VH CDR3s 0 3. The primer on the 5 end of the VH gene (VHF) is combined with RandVH primers (see Note 6, Table 1). 4. The linker fragment is ampliﬁed with the primer set GSfor and GSrev (Table 1) using a clone containing the linker as a template (see Note 7). Construction of Chicken Antibody Libraries 197

5. All the methods and conditions as in Subheadings 3.2 and 3.3 are used except 3.3 step 2 where the reactions contain 40 ng VL DNA, 40 ng VH DNA, and 54 ng linker. Here the primers, NewVarH and NewVarL, are omitted.

3.5 Restriction 1. Digest a total of 6 μg scFv “joined” product and 20 μg pHEN. Digestion of the scFv Each 100 μl reaction (using enzymes from Roche) should Gene Construct and contain 2 μg scFv or vector DNA, 1Â buffer M, 0.1 mg/ml pHEN1 Vector acetylated BSA, 40 U SfiI. Incubate overnight at 50 C(see Note 8). 2. For the NotI digest, add to each 100 μl reaction from step 1: 3 μl of 2.5 M NaCl, 6 μl of 1 M Tris (pH 8), 0.5 μl of 10 mg/ml acetylated BSA, 40 U NotI and ddH2O to make a final volume of 150 μl. Incubate at 37 C for at least 3 h or overnight. 3. Analyze the integrity of both digested products on a 1% agarose gel (see Note 9). 4. Purify the digested scFv construct with an enzyme reaction purification kit. 5. A stuffer fragment (+/À50 bp) is present between the SfiI and NotI sites of pHEN1 and should be removed. The pHEN1 reactions are concentrated by precipitation, separated by crystal violet stained agarose gel electrophoresis, the vector excised from the gel and purified (see Subheadings 3.2 steps 7–10 and 3.11).

3.6 Ligation of the 1. The molar ratio of vector (4.5 kbp): insert (800 bp) should be scFv Gene and Vector 1:2 (approximately 100 ng vector: 40 ng insert). 2. Evaluate the reagents by preparing a small ligation reaction. This is followed by reactions containing a total of 3 μg digested vector and 1.2 μg digested scFv insert. Include a ligation reaction without insert as control to confirm the efficacy of the restriction digestions. 3. Each 50 μl ligation reaction contains: 1 μg pHEN1, 0.4 μg scFv gene, 1Â ligation buffer, 3 U T4 DNA ligase (when using enzymes from Roche) and ddH2O. 4. Incubate at 15 C overnight. 5. Desalt the ligation before electroporation. This can be done by diffusion [19], with a DNA purification kit or by ethanol precipitation [20] and resuspension in a small volume of ddH2O(see Note 10).

3.7 Electroporation 1. Electroporate a small aliquot to evaluate the efﬁcacy of the ligation reaction (see Subheading 3.6) and to determine the amount of re-ligated vector. This ensures that competent cells are not wasted. 198 Jeanni Fehrsen et al.

2. Use E. coli TG1 electroporation-competent cells. 3. Each electroporation reaction consists of 2–4 μl salt free ligation reaction and 40 μl freshly thawed electrocompetent cells. Premix and then transfer to a cold 0.1 cm electroporation cuvette. Pulse at 1700 V, 200 Ω and 25 μF. 4. Transfer the cells immediately to 1 ml pre-warmed SOC medium and incubate at 37 C with shaking for 1 h. À À 5. Plate tenfold dilutions (10 1–10 4) on TYE agar containing 100 μg/ml ampicillin and 2% glucose. 6. Pellet the rest of the cells at 2000 Â g, resuspend in a small volume of medium, and plate on a larger Petri dish (15 cm). 7. Based on the trial ligation, calculate the amount of plates required for the remainder of the ligation reaction, as well as ligation reactions required to reach the predetermined library size (see Note 11). 8. From the titer plates, analyze a few clones to confirm the presence of the correct size insert. PCR the colonies using primers OP52 and M13R specific for pHEN1 (Table 1). The product should be around 1000 bp since these external primers add approximately 200 bp to the scFv gene construct. 9. Sequence a few clones with the same primers to confirm that the scFv gene constructs are correct (Fig. 2). 10. Scrape all the cells from the plates into 2Â TY medium (see Note 12), add glycerol to a final concentration of 15% v/v, aliquot and freeze away at À70 C. Make the aliquots of 500 μl or less since a small volume of cells is required per rescue, and once thawed the stock should rather not be used again. 11. Before freezing away, dilute a small aliquot 200Â and determine OD600, in the preparation for phage rescue (see Subhead- ing 3.8).

3.8 Phage Rescue 1. For panning, phages are rescued from the bacterial stocks. An aliquot of the glycerol stock is taken and the number of cells should be at least 10Â the library size (see Note 13). Example volumes are given in brackets for a small library of 107clones. 8 OD600 of 1 ¼ 8 Â 10 bacteria/ml. 2. Inoculate the bacterial glycerol stock (80 μl) into 2ÂTY (200 ml) containing 100 μg/ml ampicillin and 2% glucose. The OD600 should be less than 0.05. Ensure that the number of cells is still 10Â larger than the library size. 3. Incubate at 37 C with shaking at 240 rpm for about 2 h until OD600 ¼ 0.5. 4. Take enough to represent the library (40 ml) and add helper phage (e.g., M13KO7) at a ratio of 1:20 (bacteria: phage). Construction of Chicken Antibody Libraries 199

Incubate at 37 C without shaking for 30 and 30 min with shaking at 100 rpm. Continue with step 6. 5. The remainder of the culture (from step 3) can be centrifuged, resuspended in 1/100 volume of 2ÂTY containing 15% v/v glycerol, and stored at À70 C. This serves as secondary glycerol stocks and is convenient if sublibraries were pooled. 6. Centrifuge for 15 min at 3300 Â g and resuspend the cell pellet in 2ÂTY (200 ml) containing 100 μg/ml ampicillin and 25 μg/ml kanamycin but no glucose. 7. Grow overnight at 30 C shaking at 240 rpm. 8. Centrifuge at 3300 Â g for 15 min and use the supernatant that contains the phages displaying the scFvs. The phages are concentrated by PEG precipitation (see Subheading 3.9).

3.9 PEG Precipitation 1. Add 1/5 volume PEG/NaCl to the supernatant containing the of Phages rescued phages (see Subheading 3.8). 2. Mix well and incubate on ice for at least 30 min. 3. Spin at 2000 Â g for 15 min. Discard the SNF. 4. Re-spin for 1 min to remove all of the PEG solution. 5. Resuspend in a small volume of PBS (5 ml). Titer the phage as described below (Subheading 3.10). 6. Freeze away aliquots in the presence of 15% v/v glycerol at À70 C. 7. The phages are ready for afﬁnity selection [21, 22]. Use 1000Â or more phages than the primary library size for each panning.

3.10 Determining 1. Prepare “midlog” E. coli TG1 cells by inoculating 0.5 ml over- Phage Titer night TG1 culture into 50 ml 2ÂTY. 2. Incubate at 37 C with shaking to an OD600 of 0.3–0.6. Use immediately. 3. Make tenfold dilutions of the phage stock in 2ÂTY or PBS À (up to 10 10). 4. Mix 0.5 ml midlog TG1 with 0.5 ml phage dilution and incubate at 37 C for 30 min without shaking. 5. Plate 100 μl on TYE plates containing 100 μg/ml ampicillin and 2% glucose. Incubate at 30 C overnight. 6. Plate 100 μl of the “midlog” TG1 cells on an extra plate to ensure the cells are free of any contamination. 7. Determine the titer of the phage stock solution as follows:

CFU=ml ¼ number of colonies on plate Â 1=dilution Â 1=fraction plated Â 2: 200 Jeanni Fehrsen et al.

3.11 Crystal Violet 1. Prepare agarose gel in 1Â TAE. Don’t add usual stain such as Stained Gels ([23], See ethidium bromide. Note 14) 2. While cooling, add 1/1000 volume of the 10 mg/ml crystal violet solution (ﬁnal concentration 10 μg/ml). 3. Add 6Â loading buffer for crystal violet gels to DNA samples. 4. Electrophorese in the presence of 1Â TAE containing 10 μg/ ml crystal violet. 5. The DNA bands can be observed as blue bands on a white light box.

4 Notes

1. Suppliers of chemicals and kits listed are used with success in our laboratory, these are naturally not the only options available. 2. When working with RNA use disposable plastic ware and RNase-free ﬁlter tips where possible. All other reagents and tools must be treated with DEPC in order to inactivate RNases. 3. To access the naive repertoire of chickens, isolate RNA from the bursae of 5-week old chickens. For immune libraries, immunize chickens [24] and when the chickens have serocon- verted, use the spleen. Lymphocytes from blood and bone marrow can also be used as source of RNA. Handling the chickens and harvesting the organs should be performed by a trained animal health technician and/or veterinarian according to your institute’s Animal Ethics Code. 4. TG1 is an amber mutation suppressor strain to enable the pIII- scFv fusion protein production and has the F0 for phage transfection, both of which are essential. Overexpression with IPTG induction and/or alternative strains (HB2151) allows soluble scFv production. 5. In order to clone inserts in the pHEN1 vector [9], primer VHF 0 incorporates an SﬁI site to the VH 5 end and VLR a NotI site to 0 the VL 3 end. 6. RandVH primers will vary according to the desired CDR3 length. Usually, a sublibrary is prepared for each CDR3 length. For the Nkuku® library the synthetic VH CDR3s ranged from six to 14 amino acids [11]. 7. Alternatively, the same strategy can be followed as Subheading 3.3 by adding the linker sequence with a primer, but NewVarH will have to be extended to create a longer overlap with the 30 end of VH. Construction of Chicken Antibody Libraries 201

8. The digestion at 50 C is performed in a thermocycler (with a heated lid) to prevent condensation in the lid of the tube. 9. The agarose gel will not show a difference after restriction digestion of the scFv construct product or that the vector was digested by both enzymes, but we have had cases where there was possibly star activity or DNase contamination and the DNA reduced to smears. 10. Select a method that works best for you and results in the highest yield. We prefer ethanol precipitation. 11. By plating the primary library instead of growing it in liquid medium, all clones have an equal chance to grow. It is essential to titer the electroporated cells to determine the primary library size. You need to know that the library is large enough. Aim for at least 107 clones for an immune library and 109 for a naive or synthetic library. 12. Scrape cells into a small as possible volume. Use about 10 ml medium or less for three 15 cm plates. Add some medium to the first plate, scrape the colonies and transfer to the second plate. Give the first plate a “wash” with fresh medium and transfer as before. Continue until all the plates are scraped, add medium as needed. 13. Try to start with 10Â the library size or more if possible. Do the calculations and determine the final volumes before you start to ensure that you have the capacity to process the calculated volumes. There are often multiple glycerol stocks from different ligations and these need to be mixed proportionally to allow equal representation of each sublibrary. 14. Due to the lower sensitivity of the crystal violet stain, the gels must be “overloaded” to visualize the product. This is a convenient way to purify a large amount of DNA in a small volume.

Acknowledgments

We thank Dr. Dion H du Plessis, our now retired research leader and mentor who started the phage display group. We are grateful to Dr. Marco Romito for the veterinary support, the good care of our chickens over the years and even taking them out for walks. The Medical Research Council (Cambridge, UK) for the gift of pHENI vector. Funders that made the work possible include Agricultural Research Council-Onderstepoort Veterinary Research, Innovation Fund of the Department of Science and Technology, and National Department of Agriculture both of South Africa. 202 Jeanni Fehrsen et al.

References

1. Davies EL, Smith JS, Birkett CR, Manser JM, display library as immunochemical reagents. Anderson-Dear DV, Young JR (1995) Selec- EMBO J 13:692–698 tion of specific phage-display antibodies using 11. Van Wyngaardt W, Malatji T, Mashau C, libraries derived from chicken immunoglobulin Fehrsen J, Jordaan F, Miltiadou DR, Du Plessis genes. J Immunol Methods 186:125–135 DH (2004) A large semi-synthetic single-chain 2. Yamanaka HI, Inoue T, Ikeda-Tanaka O Fv phage display library based on chicken (1996) Chicken monoclonal antibody isolated immunoglobulin genes. BMC Biotechnol 4:6 by a phage display system. J Immunol 12. Sixholo J, van Wyngaardt W, Mashau C, 157:1156–1162 Frischmuth J, Du Plessis DH, Fehrsen J 3. Andris-Widhopf J, Rader C, Steinberger P, (2011) Improving the characteristics of a Fuller R, Barbas CF III (2000) Methods for mycobacterial 16kDa-specific chicken scFv. the generation of chicken monoclonal antibody Biologicals 39:110–116 fragments by phage display. J Immunol Meth- 13. Clackson T, Hoogenboom HR, Griffiths A, ods 242:159–181 Winter G (1991) Making antibody fragments 4. Abi-Ghanem D, Waghela SD, David J, Cald- using phage display libraries. Nature well DJ, Danforth HD, Berghman LR (2008) 352:624–628 Phage display selection and characterization of 14. Fehrsen J, van Wyngaardt W, Mashau C, single-chain recombinant antibodies against Potgieter C, Chaudhary VK, Gupta A, Eimeria tenella sporozoites. Vet Immunol Jordaan F, du Plessis DH (2005) Serogroup- Immunopathol 121:58–67 reactive and type-specific detection of blue- 5. Chiliza TE, Van Wyngaardt W, Du Plessis DH tongue virus antibodies using chicken scFvs in (2008) Single-chain antibody fragments from a inhibition ELISAs. J Virol Methods 129 display library derived from chickens immu- (1):31–39 nized with a mixture of parasite and viral anti- 15. Wemmer S, Mashau C, Fehrsen J, gens. Hybridoma 27:412–421. https://doi. Wyngaardt W, du Plessis DH (2010) Chicken org/10.1089/hyb.2008.0051 scFvs and bivalent scFv-CH fusions directed 6. Abolnik CA, Fehrsen J, Olivier A, van against HSP65 of Mycobacterium bovis. Biolo- Wyngaardt W, Fosgate G, Ellis C (2013) Sero- gicals 38:407–414 logical investigation of highly pathogenic avian 16. Rakabe M, Van Wyngaardt W, Fehrsen J influenza (HPAI) H5N2 in ostriches (Struthio (2011) Chicken single-chain antibody frag- camelus). Avian Pathol. https://doi.org/10. ments directed against recombinant VP7 of 1080/03079457.2013.779637 bluetongue virus. Food Agric Immunol 22 7. Li B, Yea J, Lin Y, Wanga M, Zhua J (2014) (3):p283–p295 Preparation and identification of a single-chain 17. Opperman PA, Maree FF, Van Wyngaardt W, variable fragment antibody against Newcastle Vosloo W, Theron J (2012) Mapping of anti- diseases virus F48E9. Vet Immunol Immuno- genic determinants on a SAT2 foot-and-mouth pathol 161:258–264 disease virus using chicken single-chain anti- 8. Huston JS, Levinson D, Mudgett-Hunter M, body fragments. Virus Res 167:370–379 Tai M-S, Novotny´ J, Margolies MN, Ridge RJ, 18. Van Wyngaardt W, Mashau C, Wright I, Fehr- Bruccoleri RE, Haber E, Crea R, Oppermann sen J (2013) Serotype- and serogroup-specific H (1988) Protein engineering of antibody detection of African horsesickness virus using binding sites: Recovery of specific activity in phage displayed chicken scFvs for indirect dou- an anti-digoxin single-chain Fv analogue pro- ble antibody sandwich ELISAs. J Vet Sci duced in Escherichia coli. Proc Natl Acad Sci U 14:95–98 S A 85:5879–5883 19. Atrazhef AM, Elliot JF (1996) Simplified 9. Hoogenboom HR, Griffiths AD, Johnson KS, desalting of ligation reactions immediately David J, Chiswell DJ, Hudson P, Winter G prior to electroporation into E. coli. Biotechni- (1991) Multi-subunit proteins on the surface ques 21:1024 of filamentous phage: methodologies for dis- 20. Sambrook J, Russel DW (eds) (2001) Molecu- playing antibody (Fab) heavy and light chains. lar cloning. A laboratory manual, Standard eth- Nucleic Acids Res 19:4133–4137 anol precipitation of DNA in microfuge tubes, 10. Nissim A, Hoogenboom HR, Tomlinson IM, vol 3, 3rd edn. Cold Spring Harbour Labora- Flynn G, Midgley C, Lane D, Winter G (1994) tory Press, Cold Spring Harbour, NY, p A8.14 Antibody fragments from a ’single pot’ phage Construction of Chicken Antibody Libraries 203

21. Clackson T, Lowman HB (eds) (2004) Phage electrophoresis and to improve cloning efﬁ- display. A practical approach. Oxford Univer- ciency. Trends J Tech Tips Online 1:23–24 sity Press, New York, NY 24. Schade R, Staak C, Hendriksen C, Erhard M, 22. Barbas CF III, Burton DR, Scott JK, Silverman Hugl H, Koch G, Larsson A, Pollmann W, van GJ (eds) (2001) Phage display. A laboratory Regenmortel M, Rijke E, Spielmann H, Stein- manual. Cold Spring Harbour Laboratory busch SD (1996) The production of avian (egg Press, Cold Spring Harbour, NY yolk) antibodies: IgY. ATLA 24:925–934 23. Rand KN (1996) Crystal violet can be used to visualize DNA bands during gel Chapter 11

Construction and Selection of Afﬁlin® Phage Display Libraries

Florian Settele, Madlen Zwarg, Sebastian Fiedler, Daniel Koscheinz, and Eva Bosse-Doenecke

Abstract

® Affilin molecules represent a new class of so-called scaffold proteins. The concept of scaffold proteins is to use stable and versatile protein structures which can be endowed with de novo binding properties and specificities by introducing mutations in surface exposed amino acid residues. Complex variations and combinations are generated by genetic methods of randomization resulting in large cDNA libraries. The selection for candidates binding to a desired target can be executed by display methods, especially the very ® robust and flexible phage display. Here, we describe the construction of ubiquitin based Affilin phage display libraries and their use in biopanning experiments for the identification of novel protein ligands.

® Key words Afﬁlin , Library construction, Phagemid, TAT phage display, Biopanning, Selection, Maturation

1 Introduction

Randomization of scaffold proteins has become a well-established method for the generation of a new class of ligands in recent years. Examples range from small protein domains such as the protein A domain (Affibody [1]), PDZ domains [2], and ankyrin repeat proteins (Darpins [3]), through small full-length proteins, such as the commonly used thioredoxin scaffold [4, 5] to higher- molecular-weight beta-barrels and Ig-like structures such as lipo- calins (Anticalins [6]), green fluorescent protein (GFP [7]), and the T-cell receptor complex [8]. They complement classical antibody- based strategies and mitigate shortcomings of the latter. Different approaches are pursued, some of which already entered technical or ® clinical use. The ubiquitin molecule is the basis of the Affilin technology and a particularly well-suited scaffold for therapeutic

Florian Settele and Madlen Zwarg contributed equally to this work.

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_11, © Springer Science+Business Media LLC 2018 205 206 Florian Settele et al.

and other applications [9]. Ubiquitin naturally occurs intracellular as well as in serum and possesses unique features with respect to protein characteristics, production, and safety aspects. Favorable biochemical properties like a stable structure over a wide pH range, thermal shifts or proteolytic degradation [10–12], and low immunogenic potential in humans are its major hallmarks. Ubiqui- tin can be easily produced as soluble protein in high yields in the cytoplasm of E. coli [12] and modular formats with dimeric head- to-tail fusions of two or more ubiquitin molecules are feasible [9]. Using the available structural data several different concepts for ubiquitin-based libraries with randomizations in the surface exposed regions of the molecule were developed. We present methods for the generation of the corresponding cDNA and the construction of highly complex phagemide libraries. The phage display technique is a powerful tool originating in the revolutionary work of Smith and coworkers as early as 1985 [13]. Since then it has been constantly developed further and many different variations emerged (reviewed for example in [14]). The method basically comprises the genetic fusion of a polypeptide of choice to a surface protein of a phage. The former is then coex- pressed with and displayed on the particles and exposed to a target protein to enable protein-protein interactions. Non-interacting phages are washed away and remaining speciﬁc binders are eluted. Here we use a special version, the so called TAT phage display adapted to the quick cytoplasmic located folding nature of our scaffold protein ubiquitin [15]. The general procedure is introduced and several options for the selection of the desired outcome are outlined.

2 Materials

2.1 Library Synthesis 1. Template DNA: plasmid containing ubiquitin gene optimized for expression in E. coli [9]. ® 2.1.1 Generation 2. Phusion High-Fidelity DNA Polymerase and 5Â Buffer-HF of the cDNA Insert Using (New England Biolabs (NEB), Frankfurt am Main, Germany). Randomized 3. PCR Nucleotide Mix, 10 mM each dNTP (Roche, Mannheim, Oligonucleotide Libraries Germany). Constructed with Synthetic Trinucleotide 4. Oligonucleotide primer (see Table 1). Phosphoramidites 5. Nuclease-Free Water (Qiagen, Hilden, Germany).

6. 50Â TAE: 2 M Tris–HCl, 1 M acetic acid, 0.1 M Na2EDTA∙2 H2O. 7. Biozyme Plaque Agarose (Biozyme Scientific GmbH, Hessisch Oldendorf, Germany). ® 8. SYBR -Safe DNA Gel Stain (Thermo Fisher Scientific, Schwerte, Germany). Construction and Selection of Affilin® Phage Display Libraries 207

Table 1 Primers used for construction and analysis of native and maturation libraries

Primer Sequence

® Affilin -lib-fw 50-GAAGGAGATATACATATGCAGATCTTCGTGAAAACCCTGACC X01X01X01X01X01X01 GGCAAAACCATTACCCTG-30 a ® Affilin -lib-rev 50-GCACGGTCTCACCCCGGCCCCCGCGGCACGCAGACGCAG AACCAGATGCAGX02X02 X02X02X02GATGTTATAATCGCTC-30 b ® Affilin -fw 50-AGGAGGGTCTCTCATGCAGATCTTCGTGAAAACC-30 ® Affilin -rev 50-GCACGG TCTCACCCCGGCCCCCGCGGCACGCAGAC-30 mod1-rev 50-CTGATCCGGCGGAATGCCTTCTTTATC-30 mod2-fw 50-GATAAAGAAGGCATTCCGCCGGATCAG-30 TorA-fw 50-TCATTGTTAACGCCGCGACG-30 MyCut-rev 50-GCCCTGGAAATACAGATTCT-30 aX01 indicate randomized positions consisting of a mix of trimer codons for all amino acids except for Cys, Ile, Leu, Val, and Phe bX02 indicate randomized positions consisting of a mix of trimer codons for all amino acids except for Cys

9. ß-Agarase I and 10Â ß-Agarase I Reaction Buffer (NEB). 10. 3 M NaAcetate (NaOAc) pH 5.0 (adjust pH with NaOH). 11. Isopropanol. 12. 70% (v/v) ethanol. 13. Elution buffer: 10 mM Tris–HCl, pH 8.5, 0.1 mM EDTA. 14. BsaI-HF (20.000 U/mL) and 10Â CutSmart Buffer (NEB). 15. DNA puriﬁcation: DNA Clean&Concentrator™-5 Kit (Zymo Research, Freiburg, Germany).

2.1.2 Generation of Insert 1. Gene library. from a Randomized Gene 2. Oligonucleotide primer (see Table 1). Library

2.1.3 Generation of Insert 1. Oligonucleotide primer (see Table 1). by Module Shufﬂing 2. Gel extraction: Zymoclean™ Gel DNA Recovery Kit (Zymo for Afﬁnity Maturation Research).

2.1.4 Generation of Insert 1. dNTP analogs, Taq polymerase, dNTP mix (10 mM) and by Error-Prone PCR mutagenesis buffer: JBS dNTP-Mutagenesis Kit (Jena Biosci- for Afﬁnity Maturation ence GmbH, Jena, Germany).

2.1.5 pCD12 Vector 1. CIP (Calf Intestinal Phosphatase, NEB). Preparation 208 Florian Settele et al.

2.1.6 Ligation of Insert 1. T4 ligase and 10Â ligase buffer (Promega GmbH, Mannheim, with BsaI-Digested pCD12 Germany). 2. ER2738 Electrocompetent Cells þ Recovery Medium (Luci- gen, Middleton, Wisconsin, US). ® 3. Gene Pulser Cuvette, 0.1 cm (Bio-Rad, Munchen,€ Germany). 4. Gene Pulser Xcell™ Electroporation System (Bio-Rad, Munchen,€ Germany). 5. 2Â YT medium: 5 g/L NaCl, 10 g/L yeast extract, 17 g/L peptone from casein (autoclaved). 6. SOBCG agar plates: 15 g/L agar, 20 g/L peptone from casein, 5 g/L yeast extract, 0.5 g/L NaCl, ad 900 mL ddH2O, autoclave, cool down to 50–60 C, add 10 mL/L sterile 1 M MgCl2∙6H2O, 55.6 mL/L sterile 2 M glucose and chloramphenicol (ﬁnal concentration 30 μg/mL) and ddH2O to a ﬁnal volume of 1000 mL). 7. Chloramphenicol stock solution: 30 mg/mL in EtOH. ® ® 8. OneTaq QuickLoad DNA Polymerase (2Â Master Mix, NEB). 9. Oligonucleotide primer (see Table 1). 10. Gel Red Nucleic Acid Stain, 10,000Â (VWR, Darmstadt, Germany).

2.1.7 Preparative 1. VentedQ-Traywithcover(sterile,240mmÂ240mmÂ20mm) Transformation of Ligation (Molecular Devices GmbH, Biberach an der Riß, Germany). Reaction and Library DNA 2. Plasmid preparation: ZymoPURE™ Plasmid Gigaprep Kit Preparation (Zymo Research).

2.1.8 Preparative 1. 2Â YT/Cm: 2Â YT medium, 30 μg/mL chloramphenicol. Transformation of Library DNA

2.1.9 Preparation, 1. M13K07 helper phage (1 Â 1011 pfu/mL, Invitrogen by Puriﬁcation, and Storage Thermo Fisher Scientiﬁc). of Phage Library 2. 2Â YT/Cm/Kan/Tet: 2Â YT medium, 30 μg/mL chloramphenicol, 50 μg/mL kanamycin, 0.1 μg/mL tetracyclin.

3. Kanamycin stock solution: 50 mg/mL in ddH2O. 4. Tetracyclin stock solution: 12.5 mg/mL in EtOH. 5. PEG precipitation solution: 20% PEG6000, 2.5 M NaCl in ddH2O. 6. 10Â PBS (Phosphate-buffered saline): 1.37 M NaCl, 27 mM KCl, 80 mM Na2HPO4∙2H2O, 20 mM KH2PO4, pH 6.7 (sterile ﬁltered). Construction and Selection of Afﬁlin® Phage Display Libraries 209

7. Desalting column: Sephadex G-25 ﬁne Hiprep desalting column (26/10, GE Healthcare, Munchen,€ Germany). 8. Buffer A: 50 mM Tris–HCl, pH 7.5. 9. Anion exchange column: Q-Sepharose High Performance column (XK50/120, GE Healthcare). 10. Buffer B: 50 mM citrate, 1.5 M NaCl, pH 4.0. 11. Blocking solution: 10Â Casein Blocking Buffer (Sigma- Aldrich, Munchen,€ Germany).

2.2 Phage Display 1. PBST: 1Â PBS, 0.1% Tween-20. Selection 2. 96 deep well plate for King Fisher, V-Bottom (Thermo Fisher Scientiﬁc). 2.2.1 Selection of Phage 3. Plastic comb for KingFisher Duo (12-Tip Comb). Libraries 4. Dynabeads ProteinA/G (Invitrogen by Thermo Fisher Scientiﬁc). 5. Trypsin (Sigma Aldrich). 6. Target protein, for example Her2-Fc (R&D Systems). 7. Off-target protein IgG-Fc (Sino Biological, Peking, China).

2.2.2 Analysis 1. 96-well medium binding plate (Greiner Bio-One, Frickenhau- of Selected Phage Libraries sen, Germany). 2. Anti-M13-HRP antibody (GE Healthcare).

2.2.3 Afﬁnity Maturation 1. Plasmid Miniprep Kit (QIAGEN, Hilden, Germany). for Selected Phage Pools

3 Methods

3.1 Library Synthesis This part describes the construction of highly complex native pha- ® gemide libraries containing more than 1010 independent Affilin variants. Furthermore, two methods for the generation of matura- ® tion libraries are described. The Affilin library is cloned into phagemid pCD12, a derivative of pCD87SA [15], in which the multiple cloning site was modified to enable cloning using the type ® IIS restriction enzyme BsaI. As an example an Affilin library was chosen, which is based on a monomeric ubiquitin molecule (76 amino acids) comprising two randomized regions. The first region is representing a loop insertion of six randomized amino acids in the N-terminal part (between amino acids nine and ten) and the second is a randomized region of five amino acids at the C-terminus of the ubiquitin molecule (amino acids 62–66) [16, 17]. 210 Florian Settele et al.

Fig. 1 Schematic overview over the generation of randomized insert for native libraries. (a) Diversified insert is constructed by PCR using randomized oligonucleotides. (b) Diversified insert is constructed by PCR amplification of gene library. Primers are indicated by arrows. BsaI restriction sites are marked by triangles

3.1.1 Generation 1. Set up 20 Â 50 μL PCR reactions containing: of the cDNA Insert Using Component Volume, μL Randomized Oligonucleotide Libraries Template DNA (1 ng/μL) 1 Constructed with Synthetic Phusion DNA polymerase (2 U/μL) 0.5 Trinucleotide Phosphoramidites (Fig. 1a) 5Â HF Buffer 10 10 mM dNTPs 1 ® Afﬁlin -lib-fwa (100 μM) 0.25 ® Afﬁlin -lib-reva (100 μM) 0.25

H2O37 aSee Note 1 Perform the PCR under following conditionsa Step Temperature, ˚C Duration

1 Initial denaturation 98 30 s 2 Denaturation 98 10 s 3 Annealing 54 20 s 4 Extension 72 30 s 5 Final extension 72 5 min aRepeat steps 2–4 for 25 cycles Construction and Selection of Afﬁlin® Phage Display Libraries 211

2. Pool PCR reactions, add loading dye and separate products on a2%1Â TAE agarose gel. Perform agarose gel electrophoresis by standard methods and use a low molecular weight standard to verify the correct product length. Use low melting point agarose to enable subsequent β-Agarase digestion. ® 3. After staining with SYBR Safe (1:2500 in H2O, 30 min), excise the product band (294 bp) using UV light and a blue- light transilluminator (see Note 2). 4. Transfer gel slices to several 2 mL reaction tubes (500–600 mg solid gel per tube). Add 1 gel volume 1Â TAE buffer to adjust gel concentration to 1%. 5. Add 1/10 volume 10Â β-Agarase I reaction buffer and melt gel slices for 10 min at 65 C with vigorous shaking. 6. Aliquot the molten agarose to 900 μL per tube, cool to 42 C, and add 13.5 units of β-Agarase I per tube. Incubate for 1 h at 42 C and 300 rpm (see Note 3). 7. After β-Agarase I digestion, remove remaining carbohydrates by alcohol precipitation of DNA: Adjust the salt concentration of the β-Agarase I treated solution by the addition of 1/9 volume of 3 M NaOAc, incubate on ice for 15 min. 8. Centrifuge at >10,000 Â g for 15 min at 4 C. 9. Remove the DNA-containing supernatants and transfer to new 2 mL reaction tubes, discard pellets. 10. Add 1 volume of isopropanol, incubate overnight at À20 C. 11. Centrifuge at >10,000 Â g for 30 min at 4 C. 12. Remove the supernatant, wash each pellet with 500 μL ice-cold 70% (v/v) ethanol. 13. Centrifuge at >10,000 Â g for 15 min at 4 C. 14. Remove the supernatant and dry the pellets at room temperature (5–10 min).

15. Resuspend all pellets in a total of 300 μLH2O. 16. Centrifuge at >10,000 Â g for 15 min at 4 C to remove any residual non-digested agarose. 17. Determine the concentration of DNA (see Note 4). 18. Set up ﬁve 200 μL restriction digest reactions with following components:

Component Volume/amount

10Â CutSmartBuffer 20 μL PCR product 3.5 μg BsaI-HF 5 μL

H2O ad 200 μL 212 Florian Settele et al.

19. Incubate at 37 C for 2 h with gentle mixing. 20. Purify the PCR product using a DNA Clean&Concentrator Kit according to the manufacturer’s instructions. Use one column for each 200 μL restriction digest reaction (number of columns depends on binding capacity). Elute each column in 50 μL H2O. Pool eluates. 21. Determine the concentration of DNA. You should get at least 5 μg of digested PCR product. Otherwise, repeat PCR, β-Agarase I, and BsaI digestion of PCR product. 22. Continue with Subheading 3.1.5.

3.1.2 Generation of Insert 1. Set up 25 Â 50 μL PCR reactions containing: from a Randomized Gene Library (Fig. 1b) Component Volume/amount Gene librarya 2.5 fmol Phusion DNA polymerase (2 U/μL) 0.5 μL 5Â HF Buffer 10 μL 10 mM dNTPs 1 μL ® Afﬁlin -fw (100 μM) 0.25 μL ® Afﬁlin -rev (100 μM) 0.25 μL

H2Oad50μL aSee Note 5 Perform the PCR under following conditionsa: Step Temperature, ˚C Duration

1 Initial denaturation 98 30 s 2 Denaturation 98 10 s 3 Annealing 54 20 s 4 Extension 72 30 s 5 Final extension 72 5 min aRepeat steps 2–4 for 22 cycles (see Note 6) 2. Continue with step 2 of Subheading 3.1.1.

3.1.3 Generation of Insert 1. For generating sublibrary 1 two PCR reactions have to be by Module Shuffling performed in parallel. One PCR will amplify module for Affinity Maturation 1 (amino acids 1–40 of the ubiquitin sequence) of the selected ® Affilin variant(s) and the other PCR will amplify module 2 (amino acids 32–76 of the ubiquitin sequence) of a native gene library (Fig. 2a). Construction and Selection of Affilin® Phage Display Libraries 213

Fig. 2 Schematic overview over the generation of randomized inserts for maturation libraries. (a) Creation of sublibrary 1 by module shuffling. Module 1 of one (or several) Affilin® variant(s) is fixed and combined with a 214 Florian Settele et al.

For module 1 set up 4 Â 50 μL PCR reactions containing:

Component Volume, μL

Template DNAa (1 ng/μL) 1 Phusion DNA polymerase (2 U/μL) 0.5 5Â HF Buffer 10 10 mM dNTPs 1 ® Afﬁlin -fw (100 μM) 0.25 mod1-rev (100 μM) 0.25

H2O37 ® ® aTemplate DNA can be a single Afﬁlin variant, a pool of several Afﬁlin variants or even a variant pool obtained from phage display selection of the initial library For module 2 set up 4 Â 50 μL PCR reactions containing: Component Volume/amount

Gene librarya 2.5 fmol Phusion DNA polymerase (2 U/μL) 0.5 μL 5Â HF Buffer 10 μL 10 mM dNTPs 1 μL mod2-fw (100 μM) 0.25 μL ® Afﬁlin -rev (100 μM) 0.25 μL

H2Oad50μL ause either gene library from Subheading 3.1.1 generated out of synthetic trinucleotide phosphoramidites (see Note 4) or gene library from Subheading 3.1.2. Perform both PCRs under following conditionsa:

Step Temperature, ˚C Duration

1 Initial denaturation 98 30 s 2 Denaturation 98 10 s 3 Annealing 54 20 s 4 Extension 72 30 s 5 Final extension 72 5 min

aRepeat steps 2–4 for 22 cycles (see Note 6) ä

Fig. 2 (Continued) native repertoire of module 2. Both modules are amplified separately by PCR yielding an overlapping complementary region which is used for the assembly of both modules in a subsequent overlap extension PCR (oePCR). (b) Creation of sublibrary 2 by module shuffling. Module 2 of one (or several) Affilin® variant(s) is fixed and combined with a native repertoire of module 1. Both modules are amplified separately by PCR yielding an overlapping complementary region which is used for assembly of both modules in a subsequent overlap extension PCR (oePCR). (c) Generation of maturation library by introducing mutations randomly by error-prone PCR using mutagenic dNTP analogs and Taq polymerase. Primers are indicated by arrows. BsaI restriction sites are marked by triangles Construction and Selection of Affilin® Phage Display Libraries 215

2. Pool the four PCR reactions of module 1 and module 2 respectively, add loading dye and separate products on a 2% 1Â TAE agarose gel (perform agarose gel electrophoresis by standard methods and use a low molecular weight standard to verify the correct product length). ® 3. After staining with SYBR Safe (1:2500 in H2O, 30 min), excise the product bands (150 bp each) using UV light and a blue-light transilluminator (see Note 2). 4. Purify the PCR products using a gel extraction kit according to the manufacturer’s instructions. 5. Determine the concentration of DNA. ® 6. To fuse both modules to a full-size Afﬁlin library perform 8 Â 50 μL overlap extension PCR (oePCR) reactions containing:

Component Volume/amount

PCR product module 1 10 ng PCR product module 2 10 ng Phusion DNA polymerase (2 U/μL) 0.5 μL 5Â HF Buffer 10 μL 10 mM dNTPs 1 μL ® Afﬁlin -fw (100 μM) 0.25 μL ® Afﬁlin -rev (100 μM) 0.25 μL

H2Oad50μL Perform oePCR under following conditionsa:

Step Temperature, ˚C Duration

1 Initial denaturation 98 30 s 2 Denaturation 98 10 s 3 Annealing 54 20 s 4 Extension 72 30 s 5 Final extension 72 5 min aRepeat steps 2–4 for 25 cycles (see Note 7). 7. For generating sub-library 2 two PCR reactions have to be performed in parallel. One PCR will amplify module 1 (amino acids 1–40 of the ubiquitin sequence) of a native gene library and the other PCR will amplify module 2 (amino ® acids 32–76 of the ubiquitin sequence) of the selected Afﬁlin variant(s) (Fig. 2b). For module 1 set up 4 Â 50 μL PCR reactions containing: 216 Florian Settele et al.

Component Volume/amount

Gene librarya 2.5 fmol Phusion DNA polymerase (2 U/μL) 0.5 μL 5Â HF Buffer 10 μL 10 mM dNTPs 1 μL ® Afﬁlin -fw (100 μM) 0.25 μL mod1-rev (100 μM) 0.25 μL

H2Oad50μL aUse either gene library from Subheading 3.1.1 generated using synthetic trinucleotide phosphoramidites (see Note 4) or gene library from Subheading 3.1.2. For module 2 set up 4 Â 50 μL PCR reactions containing: Component Volume, μL

Template DNAa (1 ng/μL) 1 Phusion DNA polymerase (2 U/μL) 0.5 5Â HF Buffer 10 10 mM dNTPs 1 mod2-fw (100 μM) 0.25 ® Afﬁlin -rev (100 μM) 0.25

H2O37 ® ® aTemplate DNA can be a single Afﬁlin variant, a pool of several Afﬁlin variants or even a variant pool obtained from phage display selection of the initial library Perform both PCRs under following conditionsa: Step Temperature, ˚C Duration

1 Initial denaturation 98 30 s 2 Denaturation 98 10 s 3 Annealing 54 20 s 4 Extension 72 30 s 5 Final extension 72 5 min aRepeat steps 2–4 for 22 cycles (see Note 6). 8. Treat all PCR reactions as described in steps 2–5. Perform oePCR of sublibrary 2 as described in step 6. 9. Continue with steps 2–16 of Subheading 3.1.1 with PCR samples of sublibrary 1 and sublibrary 2 separately. 10. Determine the concentration of DNA. 11. For each sublibrary set up two 200 μL restriction digest reactions with following components: Construction and Selection of Afﬁlin® Phage Display Libraries 217

Component Volume/amount

10Â CutSmartBuffer 20 μL oePCR product 3 μg BsaI-HF 5 μL

H2O ad 200 μL

12. Incubate at 37 C for 2 h with gentle mixing. 13. Purify the digested oePCR products using a DNA Clean&- Concentrator Kit according to the manufacturer’s instructions. Use one column for each 200 μL restriction digest reaction (number of columns depends on binding capacity). Elute each column in 50 μLH2O. Pool eluates. 14. Determine the concentration of DNA. You should get at least 1.5 μg of digested oePCR product for each sublibrary. Other- wise repeat oePCR, β-Agarase I and BsaI digestion of PCR product. 15. Continue with Subheading 3.1.5.

3.1.4 Generation of Insert 1. Set up 4 Â 50 μL PCR reactions containing: by Error-Prone PCR for Affinity Maturation (Fig. 2c) Component Volume, μL Template DNAa (1 ng/μL) 1 Phusion DNA polymerase (2 U/μL) 0.5 5Â HF Buffer 10 10 mM dNTPs 1 ® Affilin -fw (100 μM) 0.25 ® Affilin -rev (100 μM) 0.25

H2O37 ® ® aTemplate DNA can be a single Afﬁlin variant, a pool of several Afﬁlin variants or even a variant pool obtained from phage display selection of the initial library Perform the PCR under following conditionsa: Step Temperature, ˚C Duration

1 Initial denaturation 98 30 s 2 Denaturation 98 10 s 3 Annealing 72 20 s 4 Extension 72 30 s 5 Final extension 72 5 min aRepeat steps 2–4 for 22 cycles (see Note 6) 218 Florian Settele et al.

2. Pool all four PCR reactions, add loading dye and separate products on a 2% 1Â TAE agarose gel (perform agarose gel electrophoresis by standard methods and use a low molecular weight standard to verify the correct product length). ® 3. After staining with SYBR Safe (1:2500 in H2O, 30 min), excise the product band (282 bp) using UV light and a blue- light transilluminator (see Note 2). 4. Purify the PCR product using a gel extraction kit according to the manufacturer’s instructions. 5. Determine the concentration of DNA. Use this linear DNA as a template for subsequent error-prone PCR (epPCR). 6. Perform epPCR using dNTP analogs dPTP and 8-oxo-dGTP. Set up 1Â 50 μL epPCR reaction for each dNTP analog containing following components:

Component Volume/amount

PCR product (from step 5) 25 fmol Taq polymerase (5 U/μL) 1 μL 10Â Mutagenesis Buffer 5 μL 10 mM dNTPs 2.5 μL 10 mM dNTP analog 150 μM dPTP or 750 μM 8-oxo-dGTPa ® Afﬁlin -fw (100 μM) 0.25 μL ® Afﬁlin -rev (100 μM) 0.25 μL

H2Oad50μL aAmount of dPTP and 8-oxo-dGTP was determined by test PCRs to yield a ® mutation rate at an average of 2–3 exchanges per Afﬁlin monomer Perform the epPCR under following conditionsa:

Step Temperature, ˚C Duration

1 Initial denaturation 92 2 min 2 Denaturation 92 60 s 3 Annealing 55 90 s 4 Extension 72 5 min aRepeat steps 2–4 for 30 cycles for 8-oxo-dGTP epPCR and for 4 cycles for dPTP epPCR 7. To eliminate mutagenic dNTPs perform a final PCR with standard dNTP mix. Use epPCR product without further purification as a template. Set up 6 Â 50 μL PCR reactions for each dNTP analog, respectively: Construction and Selection of Affilin® Phage Display Libraries 219

Component Volume, μL

epPCR product 1 Taq polymerase (5 U/μL) 1 10Â Mutagenesis Buffer 5 10 mM dNTPs 2.5 ® Afﬁlin -fw (100 μM) 0.25 ® Afﬁlin -rev (100 μM) 0.25

H2Oad50 Perform ﬁnal PCR under following conditionsa:

Step Temperature, ˚C Duration

1 Initial denaturation 92 2 min 2 Denaturation 92 60 s 3 Annealing 55 90 s 4 Extension 72 5 min aRepeat steps 2–4 for 25 cycles 8. Continue with steps 2–16 of Subheading 3.1.1 treating the PCR samples for each dNTP analog separately. 9. Determine the concentration of DNA. 10. Mix puriﬁed PCR products of each dNTP analog in equal amounts. 11. Set up two 200 μL restriction digest reactions with following components:

Component Volume/amount

10Â CutSmartBuffer 20 μL Mixed PCR product 3 μg BsaI-HF 5 μL

H2O ad 200 μL

12. Incubate at 37 C for 2 h with gentle mixing. 13. Purify the PCR product using a DNA Clean&Concentrator Kit according to the manufacturer’s instructions. Use one column for each 200 μL restriction digest reaction (number of columns depends on binding capacity). Elute each column in 50 μL H2O. Pool eluates. 14. Determine the concentration of DNA. You should get at least 1.5 μg of digested PCR product. Otherwise, repeat ﬁnal PCR, β-Agarase I and BsaI digestion of PCR product. 15. Continue with Subheading 3.1.5. 220 Florian Settele et al.

3.1.5 pCD12 Vector 1. Set up 4 Â 250 μL restriction digest reactions with following Preparation components:

Component Volume/amount

10Â CutSmartBuffer 25 μL pCD12 vector 17.5 μg BsaI-HF 5.25 μL

H2O ad 250 μL 2. Incubate at 37 C for 2 h with gentle mixing. 3. Add 8.75 μL CIP (10 U/μL) to each reaction and incubate for 30 min at 37 C with gentle mixing. 4. Pool all four reactions, add loading dye, and separate products on a 1% 1Â TAE agarose gel (perform agarose gel electrophoresis using low melting agarose by standard methods and use a molecular weight standard to verify the correct product length). ® 5. After staining with SYBR Safe (1:2500 in ddH2O, 30 min), excise the product band (4294 bp) using UV light and a blue- light transilluminator. 6. Transfer gel slices to several 2 mL reaction tube (500–600 mg solid gel per tube). Add 2 gel volumes 1Â β-Agarase I reaction buffer and incubate for 30 min on ice. Remove buffer, repeat washing step once. 7. Remove buffer and melt gel slices by incubation for 10 min at 65 C with vigorous shaking. 8. Aliquot the molten agarose to 700 μL per tube, cool to 42 C, and add 5 units β-Agarase I per tube. Incubate for 1 h at 42 C and 300 rpm (see Note 8). 9. After β-Agarase I digestion, remove remaining carbohydrates by alcohol precipitation of DNA: Adjust the salt concentration of the β-Agarase I treated solution by the addition of 1/9 volume of 3 M NaOAc, incubate on ice for 15 min. 10. Centrifuge at 16,000 Â g for 15 min at 4 C. 11. Remove the DNA-containing supernatants and transfer to new 2 mL reaction tubes, discard pellets. 12. Add 1 volume of isopropanol, incubate overnight at À20 C. 13. Centrifuge at 16,000 Â g for 30 min at 4 C. 14. Remove the supernatant, wash each pellet with 500 μL ice-cold 70% (v/v) ethanol. 15. Centrifuge at 16,000 Â g for 15 min at 4 C. 16. Remove the supernatant and dry the pellets at room temperature (5–10 min). Construction and Selection of Afﬁlin® Phage Display Libraries 221

17. Resuspend all pellets in a total of 300 μLH2O. 18. Centrifuge at 16,000 Â g for 15 min at 4 C to remove any residual non-digested agarose. 19. Determine the concentration of DNA.

3.1.6 Ligation of Insert 1. To verify the quality of randomized libraries and digested with BsaI-Digested pCD12 vector perform one test ligation (1 Â 200 μL): Mix 1.5 μg linearized vector and 265 ng digested PCR product (molar ratio 1:3) obtained in Subheadings 3.1.1–3.1.3 or 3.1.4. In parallel set up a control reaction without insert DNA. 2. Add 1/9 volume of 3 M NaOAc and 1 volume of isopropanol to each reaction (see Note 9). 3. Incubate for 2 h at À20 C. 4. Centrifuge at 16,000 Â g for 30 min at 4 C. 5. Discard the supernatant, wash each pellet with 500 μL ice-cold 70% (v/v) ethanol. 6. Centrifuge at 16,000 Â g for 15 min at 4 C. 7. Discard the supernatant, dry the pellets at room temperature (5–10 min).

8. Resuspend each pellet in 170 μLH2O, add 20 μL10Â ligase buffer and 10 μL T4 ligase (3 U/μL). 9. Aliquot ligation reaction and control reaction in 10 Â 20 μL (0.2 mL PCR tubes) each and incubate overnight at 16 Cina PCR thermocycler. 10. Pool ten 20 μL aliquots of ligation and control reaction, respectively and purify DNA using a DNA Clean&Concentra- tor Kit according to the manufacturer’s instructions. Use one column for each 200 μL reaction. Elute each column in 11 μL pre-warmed (37 C) H2O. 11. Thaw electrocompetent ER2738 on ice and mix 25 μL cells with 2 μL of puriﬁed ligation reaction and control reaction, respectively. 12. Transfer the mix to a prechilled electroporation cuvette, dry the outside of the cuvette thoroughly with a tissue paper, and perform a 1.8 kV pulse (600 Ω,10μF) using an electroporator. 13. Immediately add 975 μL of pre-warmed (38 C) recovery medium, resuspend cells several times, and transfer suspension to a pre-warmed (37 C) 5 mL culture tube (see Note 10). 14. Incubate for 1 h at 250 rpm at 37 C. 15. To analyze transformation efﬁciency take 10 μL (1:100 dilution) of each cell suspension and make tenfold serial dilutions À in 2Â YT medium down to 10 8. Plate dilutions ranging from 222 Florian Settele et al.

À À 10 5 to 10 8 on SOBCG agar plates and incubate overnight at 32 C. 16. Calculate the transformation efﬁciency by counting the number of clones of the ligation and control reaction and multiply by the dilution factor (see Note 11). 17. To verify the percentage of clones carrying a full-size insert analyze 96 clones from the transformation plates by single- colony PCR. 18. Prepare a master mix (100 Â 20 μL) containing following components:

Component Volume, μL

® OneTaq 2Â Master Mix 1000 TorA-fwa (100 μM) 4 MyCut-revb (100 μM) 4

H2O 992 aPrimer located in the vector backbone upstream of the coding sequence bPrimer located in the vector backbone downstream of the coding sequence 19. Aliquot 20 μL of master mix in each well of a 96-well PCR plate. Pick a single colony for each well using a sterile toothpick and dip it several times directly into the master mix. 20. Perform the PCR under following conditionsa:

Step Temperature, ˚C Duration

1 Initial denaturation 94 2 min 2 Denaturation 94 20 s 3 Annealing 60 30 s 4 Extension 68 30 s 5 Final extension 68 5 min aRepeat steps 2–4 for 25 cycles 21. Analyze a 10 μL aliquot of each well on a 1.5% 1Â TAE agarose gel. After electrophoresis stain with GelRed™ (1:5000, 30 min, see Note 12) and visualize bands using UV light. 22. Positive clones with full-size insert will result in PCR products of 318 bp. Product bands of 114 bp will represent empty vector (see Note 13). 23. In addition to single-colony PCR the quality of the library will be checked by sequencing of 96 individual clones (see Note 14). 24. If percentage of clones with correct insert DNA is satisfying, perform ligation reactions in preparative scale (native library: 16 Â 200 μL, afﬁnity maturation library: 4 Â 200 μL) Construction and Selection of Afﬁlin® Phage Display Libraries 223

according to steps 1–10. As a result you will receive 160 μL purified ligation reaction sufficient for 80 transformations or 40 μL purified ligation reaction sufficient for 20 transformations, respectively (see Note 15).

3.1.7 Preparative 1. Prepare 75 Q-trays with SOBCG agar. Transformation of Ligation Reaction and Library DNA 2. Perform up to 75 transformations as described in Subheading Preparation 3.1.6 steps 11–14 (see Note 16). 3. Exactly 1 h after transferring the pulsed cells to 37 C take 10 μL from each transformation reaction for the determination of transformation efficiency (see Note 17). The rest of each transformation reaction (990 μL) will be plated on a Q-tray. 4. Incubate Q-trays and agar plates for the determination of transformation efficiency overnight at 32 C. 5. Calculate the transformation efficiency by counting the number of clones on agar plates. The theoretical library size will be the average clone numbers determined from the seven independent serial dilutions minus the clones without insert identified by single-colony PCR. 6. Scrape off the colonies of each Q-tray with 20 mL of 2Â YT medium, respectively. Pool cell suspensions of all Q-trays and pellet cells for 15 min at 4000 Â g and 4 C. Determine pellet weight and resuspend in 200 mL of 2Â YT medium. Prepare aliquots of 10 g cells (wet weight) and centrifuge again for 15 min at 4000 Â g and 4 C to remove residual liquid. 7. Store pelleted cells at À20 C. 8. Thaw one 10 g cell pellet aliquot and perform a plasmid Giga preparation according to the manufacturer’s instructions. Determine the concentration of DNA, aliquot DNA at 30 μg, and store at À80 C. 9. Eliminate residual empty vector by restriction digestion of library DNA with BsaI (see Note 18). Set up 4 Â 250 μL restriction digest reactions with following components:

Component Volume/amount

10Â CutSmartBuffer 25 μL Library DNA 7.5 μg BsaI-HF 3.75 μL

H2O ad 250 μL

10. Incubate at 37 C for 2 h with gentle mixing. Add 2.5 μL CIP (10 U/μL) to each reaction and incubate for further 30 min at 37 C with gentle mixing. 224 Florian Settele et al.

11. Purify the library DNA using a DNA Clean&Concentrator Kit according to the manufacturer’s instructions. Use two columns for each 250 μL reaction. Elute each column in 15 μL H2O. Pool eluates. 12. Determine the DNA concentration. Dilute to 85 ng/μL and store BsaI-digested library DNA at À80 C. 13. Perform a test transformation using 85 ng of BsaI-digested library DNA. Thaw electrocompetent ER2738 on ice and mix 25 μL cells with 1 μL of library DNA. Perform transformation as described in Subheading 3.1.6 steps 12–14. 14. Analyze transformation efﬁciency as described in Subheading 3.1.6 step 15. 15. Calculate the transformation efﬁciency by counting the number of clones and multiply by the dilution factor. 16. Determine percentage of clones carrying empty vector by single-colony PCR of 96 individual clones, as described in Subheading 3.1.6 steps 18–21. 17. Positive clones with full-size insert will result in PCR products of 318 bp. Product bands of 114 bp will represent empty vector (see Note 19).

3.1.8 Preparative 1. Perform 20–30 transformations of BsaI-digested library DNA. Transformation The number depends on transformation efficiency determined of Library DNA by test transformation (see Subheading 3.1.7 steps 13–15). The library should be overrepresented two times. For a theoretical library size of 1 Â 1010 2 Â 1010 transformants should be generated. 2. Thaw electrocompetent ER2738 on ice and mix 25 μL cells with 85 ng of BsaI-digested library DNA. 3. Perform transformation as described in Subheading 3.1.6 steps 12–14. 4. Analyze transformation efficiency as described in Subheading 3.1.6 step 15 with four randomly selected transformations. On the next day calculate the transformation efficiency by counting the number of clones and multiply by the dilution factor. 5. Transfer the complete cell suspension to 200 mL pre-warmed (37 C) 2Â YT medium containing 15 μg/mL chloramphenicol. 6. One hour after transfer of the last transformation sample add 200 mL pre-warmed (37 C) 2Â YT medium containing 45 μg/mL chloramphenicol and incubate for 2 h at 220 rpm and 37 C. Construction and Selection of Affilin® Phage Display Libraries 225

7. Measure OD600nm of cell suspension and use this culture to inoculate 600 mL 2Â YT/Cm medium for phage library preparation.

3.1.9 Preparation, 1. Use cell suspension of transformation of Subheading 3.1.8, Purification, and Storage step 7 to inoculate 600 mL of 2Â YT/Cm in a 5-L culture of Phage Library flask to an initial OD600nm of 0.1. 2. Incubate at 37 C and 220 rpm until the culture has reached an OD600nm of 0.5–0.6. 3. Infect the cell suspension with 3 Â 1011 cfu M13 K07 helper phages, incubate for 30 min at 37 C without shaking and additional 30 min at 26 C, 220 rpm. 4. Centrifuge cell suspension for 10 min at 4000 Â g,4C. Discard the supernatant and resuspend cells in 600 mL pre-warmed (26 C) 2Â YT/Cm/Kan/Tet medium. 5. Incubate overnight at 26 C, 220 rpm. 6. To separate produced phages from bacterial cells centrifuge culture for 10 min at 17,000 Â g,4C. Pass phage containing supernatant through a syringe filter of 0.45 μm pore size to remove any residual bacterial cells. Change filters every 60 mL. 7. Precipitate phages by adding 1/4 volume PEG solution and incubate on ice for 30 min. 8. Pellet phages by centrifugation for 10 min at 12,000 Â g,4C. 9. Discard the supernatant and place the centrifuge tubes upside down on a tissue paper to discard residual, viscous PEG solution. Resuspend phage pellet in 10 mL 1Â PBS and incubate for 5 min on ice. 10. Aliquot phage solution in 2 mL reaction tubes and centrifuge for 5 min at 16,000 Â g. 11. Pool all supernatants and precipitate phages again by adding 1/4 volume PEG solution and incubate on ice for 15 min. 12. Aliquot solution in 2 mL reaction tubes and pellet phages by centrifugation for 10 min at 16,000 Â g,4C. 13. Discard the supernatant and centrifuge again for one minute to remove all of the remaining PEG solution. Resuspend phage pellets in a total volume of 10 mL of 1Â PBS. 14. To remove all remaining insoluble debris centrifuge again for 5 min at 16,000 Â g,4C. Transfer the phage containing supernatant to a new 15 mL falcon tube. 15. Determine the phage amount by measuring the absorption at 269 and 320 nm and calculate the phage concentration with the following formula: 226 Florian Settele et al.

ðÞÂA À A 6 Â 1016 phages=mL ¼ 269 320 number of bases=phage

16. For long time storage purify phage library using an ion exchange chromatography [18]. Perform all chromatographic steps on an A¨KTA Avant FPLC system (GE Healthcare) at room temperature. 17. In order to remove NaCl and PEG apply the solubilized phages to a desalting column equilibrated with 1.5 CV of buffer A at a flow rate of 5 mL/min, followed by the injection of 10 mL of phage solution through a 50 mL superloop. Monitor desalting performance with respect to UV-signal (260 nm) and conduc- tivity (ms/cm) and collect 2 mL fractions. Pool phage containing fractions. 18. In a second step, apply the phages to an anion exchange column equilibrated with 2 CV of buffer A at a flow rate of 5 mL/ min, followed by injection of 20 mL desalted phages using a 50 mL superloop. Remove weakly bound phages by a washing step with 1.5 CV of buffer A. Elute bound phages in an isocratic mode by washing stepwise with 1.5 CV of 10, 20, 50, and 100% of buffer B. Collect eluted phages from the 20% buffer B step in 10 mL fractions (see Note 20). Pool all phage containing fractions. Determine the phage amount as described in step 15. The average recovery is around 75% of phage input. 19. Determine the phage titer by adding 10 μL phage solution to 90 μL1Â PBS and performing tenfold serial dilutions down to À 10 10. Add 810 μL of exponentially growing E. coli ER2738 (OD600nm ¼ 0.5–0.6) to each 90 μL phage dilution and incubate for 30 min at 37 C without shaking. Plate 100 μL of the À À infected bacterial culture from dilutions 10 7–10 10 on SOBCG agar plates and incubate overnight at 30 C. Calculate the phage titer by counting the colonies and multiply by the dilution factor (number of colonies Â dilution factor ¼ cfu/ 10 μL). 20. Add 10Â Blocking solution to an end concentration of 0.5Â and aliquot phage solution to 1–3 Â 1012 phages (see Note 21). 21. Freeze phage aliquots for several minutes in a mix of dry ice and ethanol and store phage library at À80 C(see Note 22). Construction and Selection of Affilin® Phage Display Libraries 227

® 3.2 Phage Display This part describes the selection of Affilin libraries to obtain Against Fc-Tagged binding molecules directed against Fc-tagged or biotinylated target Target Protein proteins (e.g., human membrane protein Her2). To achieve maxi- with Affinity mal flexibility but maintain high capacity, a magnetic bead handling Maturation Step device is used for all selections. Target proteins for selection are bound to functionalized beads (e.g., with Streptavidin or Protein A/G) and are incubated with phage preparations of libraries. The use of a KingFisher Duo device (Thermo Scientific) allows completion of twelve different selections in parallel and decreases manual handling times drastically. The binding molecules obtained in a primary selection do not always meet the desired criteria for, e.g., affinity or stability. To circumvent the time consuming search for a, maybe rare, single- lead candidate, a maturation of target binding positive whole phage pools by epPCR or module shuffling is a practical alternative. This shortens turnaround times and allows for straightforward selection of individual clones with high affinity.

3.2.1 Preparation 1. The day prior to infection of E. coli with the phage library and Puriﬁcation of Phages inoculate 50 mL 2Â YT medium with a single clone of E. coli from Cryo Stocks ER2738 strain and grow overnight at 37 C, 200 rpm.

2. Determine the OD600nm and dilute the ER2738 culture to OD600nm 0.1 in 800 mL 2Â YT medium without antibiotics. 3. Grow the culture at 37 C, 220 rpm until an OD600nm of 0.5–0.6 is reached (see Note 23). 4. Infect the ER2738 with one aliquot of the phage library cryo stock containing 1–3 Â 1012 phages from Subheading 3.1.9 for 30 min at 37 C without shaking. To start with freshly prepared phages see Subheading 3.1.8. 5. After infection with phages centrifuge the culture at 4000 Â g, 4 C for 10 min in a JLA-8.1000 rotor in a Beckman Avanti centrifuge. 6. Discard the resulting supernatant and resuspend the remaining bacteria pellet in 800 mL 2Â YT medium without antibiotics in a 5 L Erlenmeyer ﬂask. 7. After growth of the culture for 1 h at 37 C and 220 rpm add 200 μL of Chloramphenicol.

8. One additional hour later determine the OD600nm and dilute the culture to OD600nm 0.1 in 400 mL 2Â YT/Cm medium in a 3 L Erlenmeyer ﬂask. 9. Grow the bacterial culture at 37 C, 220 rpm until an OD600nm of 0.5 is reached. 10. Then infect the bacteria with 2 mL M13KO7 helper phages for 30 min at 37 C without shaking. 11. Transfer bacteria to a shaking incubator and cultivate for 30 min at 26 C, 220 rpm. 228 Florian Settele et al.

12. After infection with helper phages centrifuge the culture at 4000 Â g,4C for 10 min in a JLA-8.1000 rotor in a Beckman Avanti centrifuge. 13. Discard the resulting supernatant and resuspend the remaining bacteria pellet in 400 mL 2Â YT/Cm/Kan/Tet medium in a 3 L Erlenmeyer flask (see Note 24). 14. Incubate the final culture overnight at 26 C, 220 rpm for approximately 16 h. 15. Centrifuge the culture at 4000 Â g,4C for 10 min in a JLA-8.1000 rotor in a Beckman Avanti centrifuge. 16. Sterile filter the resulting supernatant through a syringe filter of 0.45 μm pore size, transfer to a fresh 1 L Erlenmeyer flask, and precipitate phages by adding 1/4 volume PEG solution. 17. Incubate on ice for 30 min. 18. Afterward, centrifuge the phages at 4 C, 12,000 Â g for 10 min in a JA10 rotor in a Beckman Avanti centrifuge. 19. Discard resulting supernatants and resuspend the remaining pellet in 10 mL ice-cold PBS. 20. Incubate phages for 5 min on ice and add 1/4 volume PEG precipitation solution. 21. Precipitate phages for 15 min on ice followed by a centrifugation step in an Eppendorf tabletop centrifuge (10 min, 4 C, 16,000 Â g). 22. Discard supernatants completely and resuspend the final phage pellet in 3 mL sterile PBS. 23. Quantify phages with the Nanodrop2000c instrument at OD269/320nm as described under Subheading 3.1.9, step 15. 24. Add 330 μL10Â blocking solution and store phages at 4 Cor use immediately for selection process.

3.2.2 Selection of Phage 1. For the selection of phage libraries in the KingFisher Duo Libraries 96 deep well plates are used. 2. In Table 2 selection conditions of a standard selection against a Fc-tagged target protein are shown (see Note 25). 3. To determine binding capacity of beads perform an immobilization test prior to selection start according to standard protocols. 4. The day before selection wash the desired volume of beads determined in the immobilization test with 1000 μL of PBST and incubate overnight with 10Â blocking solution. 5. Write a program for the KingFisher Duo using the BindIT software according to Fig. 3 and following steps 10–17 (see Note 26). Construction and Selection of Afﬁlin® Phage Display Libraries 229

Table 2 Selection scheme for four rounds of phage display and two rounds of maturation

Target Off target Wash steps Beads

Round selection 1 Fc-tagged target protein (200 nM) Empty beads 5Â PBST, 5Â PBS Protein A 2 Fc-tagged target protein (100 nM) IgG-Fc (300 nM) 5Â PBST, 5Â PBS Protein G 3 Fc-tagged target protein (50 nM) IgG-Fc (150 nM) 13Â PBST, 3Â PBS Protein G 4 Fc-tagged target protein (25 nM) IgG-Fc (75 nM) 17Â PBST, 3Â PBS Protein G Round maturation 1 Fc-tagged target protein (25 nM) IgG-Fc (75 nM) 13Â PBST, 3Â PBS Protein A 2 Fc-tagged target protein (5 nM) IgG-Fc (15 nM) 17Â PBST, 3Â PBS Protein G

6. On selection day before starting the KingFisher Duo device inoculate 35 mL 2Â YT medium with a single colony of ER2738 and grow at 37 C, 220 rpm for several hours. This culture will be needed later for re-infection with recovered phages from the selection process. 7. One hour prior to selection start, wash blocked beads three times with 1000 μL PBS and load pre-incubation beads (Dyna- beads ProteinA) with off-target (e.g., IgG-Fc) in an Eppendorf Thermomixer for 1 h at RT, 1200 rpm. Because empty beads are used as pre-incubation beads in round one the loading with off-target can be omitted (see Note 27). 8. Set up the required plates for selection round 1 according to the scheme in Fig. 3. For later rounds with additional washing steps include a fourth deep well plate. 9. Since the KingFisher Duo is a semi-automatic device, individual handling steps of the device will be described in more detail. 10. During the ﬁrst hour of the program, the selected Fc-tagged target protein is bound to Dynabeads ProteinA in parallel to pre-incubation of the phages with empty or off-target-beads: target beads from row A will be transferred by the magnetic ﬁngers using comb 1 to target solution in row D. Afterward the pre-incubation beads from row F will be transferred to phage solution in row H using comb 2. Target immobilization and phage pre-incubation will be performed in parallel by mixing in an alternating fashion with comb 1 and comb 2, respectively (time intervals are 3 min) (see Note 28). 11. After removal of pre-incubation beads from the phage solution, target beads are washed once for 3 min in PBS (row E) and then transferred to the phage solution (row H). 230 Florian Settele et al.

Fig. 3 Selection scheme in 96-well plates. Positions are marked in gray scale on the respective plate shown on the left. Reagents to be added in specified amounts are indicated beneath. This selection scheme is only an example for five selections in parallel with the KingFisher Duo. In theory, the flexibility of this device allows for 12 different selections in a 96 deep well plate. On the other hand, the KingFisher Duo is limited in that it can only handle rows and not wells, because the 12 magnetic fingers are one unit

12. Phages are incubated with the target beads for 2 h at RT with mixing of the solution every 3 min for 30 s (see Note 29). 13. Subsequently, the target beads are removed from the phage solution and transferred to a second 96 deep well plate. 14. Beads are washed five times with PBST and five times with PBS for 3 min each under continuous mixing to remove phages binding nonspecifically to the target beads (Plate 2, row A–H, plate 3, row D and E). 15. Target beads are then transferred to row A of the third 96 deep well plate and phages are eluted from the target beads by incubation for 30 min at 37 C with 200 μL of 100 μg/mL Trypsin (eluate 1) (see Note 30). Construction and Selection of Affilin® Phage Display Libraries 231

16. After trypsin elution target beads are washed 3 min in 200 μL PBS (eluate 2, plate 3, row B) and discarded afterward. 17. Plate three is moved to the front of the KingFisher Duo in order to take the plate with the eluate easily out of the device. 18. Take eluates 1 and 2 out of wells from rows A and B, respectively, and transfer to a 50 mL Falcon tube, mix and store shortly on ice (see Note 31). 19. One hour before the end of the KingFisher program determine the OD600nm of the ER2738 culture prepared in step 6 (see Note 32).

20. Dilute the bacterial culture to an OD600nm of 0.1 in a fresh Erlenmeyer ﬂask. The ﬁnal volume of the culture depends on the number of the selections performed. For each selection, 20 mL of bacterial culture are needed. 21. Incubate the bacterial culture at 37 C, 220 rpm until an OD600nm of 0.5–0.6 is reached.

22. As soon as the desired OD600nm is reached, transfer 20 mL of the culture to a 50 mL Falcon tube containing the mixed eluates of the respective selection. 23. Swirl the tubes gently and incubate for 30 min at 37 C without shaking. 24. Centrifuge the tubes for 10 min at >1700 Â g in an Eppendorf centrifuge and discard the supernatant. 25. Resuspend pellets in 1 mL 2Â YT medium and plate on a SOBCG-Q-Tray. 26. Incubate Q-Trays overnight at 30 C. 27. To perform subsequent selection rounds, scrape the bacteria off the respective Q-tray with 10–20 mL 2Â YT medium and determine the OD600nm of the resulting bacteria suspension (see Note 33). 28. Store glycerol stocks (20% end concentration of glycerol) for each selection round of the bacterial culture and bacterial pellets for DNA preparation (see Note 34).

29. Dilute bacteria in 30 mL 2Â YT/Cm medium to an OD600nm of 0.1 and use these cultures for subsequent infection with helper phages and phage preparation as described under Sub- heading 3.2.1, step 10 (see Note 35).

3.2.3 Analysis To determine the target binding of selection rounds and the best of Selected Phage Libraries suitable pool(s) for maturation by epPCR or module shufﬂing a phage pool ELISA is performed. Phage preparations of each round are analyzed for binding to the respective target or off-target used in the selection as well as controls. 232 Florian Settele et al.

1. The day prior to the ELISA, immobilize 50 μL of a 2.5 μg/mL solution of the respective targets on 96-well medium binding plates at 4 C overnight (see Note 36). 2. The next day, wash plates three times with 300 μL PBST. 3. Add 300 μL1Â blocking solution to each well and incubate for 2 h to block unspeciﬁc binding to the plastic surface. 4. After blocking, wash the plate three times with 300 μL PBST. 5. Dilute phage solutions until a concentration of 1 Â 1012 phages/mL is reached and add 100 μL to the respective wells. 6. Incubate phages for 1 h with the targets followed by three washing steps with 300 μL PBST. 7. To detect bound phages, add 50 μL of anti-M13-HRP antibody (1:5000 in 1Â blocking solution diluted in PBS) to each well and incubate for an additional hour. 8. Wash the plate three times with 300 μL PBST and 300 μL PBS, respectively. 9. After the removal of residual liquid, add 50 μL of TMB substrate and incubate for 5–30 min dependent on signal development. 10. Stop the enzymatic reaction of the HRP by addition of 50 μL of 0.2 M H2SO4. 11. Measure the 96-well plate in a 96-well plate reader at 450 and 620 nm. 12. Plot values for each pool against the selected targets and choose pools for maturation by epPCR or module shufﬂing (see Note 37). 13. Sequence 96 clones from bacterial colonies out of pools of interest to make sure that the pools still have a high diversity (see Note 38).

3.2.4 Affinity Maturation 1. Perform DNA preparation of bacterial pellets of selected pools for Selected Phage Pools with standard, commercial kits according to the manufacturer’s instructions (Subheading 3.2.2, step 28). 2. Prepared DNA is then used for the maturation of pools with module shuffling or epPCR as described in Subheadings 3.1.3 and 3.1.4. 3. Transform bacteria with the DNA of the maturation library and prepare phages as described in Subheadings 3.1.8 and 3.1.9. Adjust number of transformations according to theoretical library complexity and efficiency of test transformations. Use phages directly for the selection without chromatography purification for long-term storage at this point. 4. Use prepared phages for the selection as described in Subhead- ing 3.2.2. Construction and Selection of Affilin® Phage Display Libraries 233

5. Perform two rounds of maturation selection starting with the target amount of round 4 of the primary selection in round 1 and a decrease of target amount of at least ﬁvefold in round 2(see Table 2 and Note 39).

3.2.5 Screening 1. Perform DNA preparation of bacterial pellets of selected pools of Matured Pools with standard, commercial kits according to the manufacturer’s instructions. ® 2. Amplify the coding sequence (CDS) for Affilin molecules by PCR and add BsaI restriction sites via the respective fw- and rev-primer. 3. Purify the PCR product with the DNA Clean&Concentrator kit (Zymo Research) according to the manufacturer’s instructions. 4. Digest a bacterial expression vector (e.g., pET28a provided with a Strep-tag) with restriction enzymes of choice and additionally treat with calf intestinal phosphatase to minimize self- ligation of the vector. 5. Purify vector by gel electrophoresis and gel extraction. 6. Ligate vector and insert using a quick ligation protocol. 7. Purify the ligation reaction with the DNA Clean&Concentra- tor kit according to the manufacturer’s instructions. 8. Transform BL21(DE3) cells with a 1:10 dilution of the purified ligation reaction for bacterial protein expression. 9. Pick single colonies automatically using for instance a K3XL ® colony picker (K-Bio) and analyze expressed Affilin variants by high-throughput screening.

4 Notes

1. For preparative phage library construction it is recommended to use randomized oligonucleotides generated by synthetic trinucleotide phosphoramidites compared to conventional degenerated oligonucleotides (e.g., containing NNK codons). Synthesizing the random portion of the oligonucleotides codon-by-codon using the trimer codon synthesis method has the advantage to effectively control codon bias at randomized positions. It is possible to selectively exclude certain amino acids as well as to prevent unwanted stop codons and ® frame-shift mutations. For the construction of the Afﬁlin library the amino acid residues Cys, Ile, Leu, Val, and Phe have been omitted from randomized positions in the loop at the N-terminus and the residue Cys has been omitted from the randomized region at the C-terminus. To ﬁrst test new library 234 Florian Settele et al.

formats for solubility and temperature stability in an expression vector, it is possible to use cheaper degenerated oligonucleotides. ® 2. SYBR Safe DNA gel stain is specifically formulated to be a less hazardous alternative to ethidium bromide that can be used with either blue-light or UV excitation. Using UV light and a blue-light transilluminator will cause less DNA damage and therefore will give better cloning efficiencies compared to staining with ethidium bromide and excising directly under UV light. 3. β-Agarase I digestion will give better yields and quality of PCR product compared to other commercial gel extraction kits resulting in better transformation efficiency. 4. For long time storage repeat steps 1–17 several times to generate a randomized gene library stock. Resuspend DNA at step 15 in elution buffer and store DNA at À80 C. This library can be used as a template for affinity maturation (see Subheading 3.1.3). 5. For preparative phage library construction it is recommended to use a gene library in which the randomized sequences are synthesized codon-by-codon using the trimer codon synthesis method. This method has the advantage to effectively control codon bias at randomized positions, to exclude selectively certain amino acids as well as to prevent unwanted stop codons ® and frame-shift mutations. For the construction of Affilin libraries the amino acid residue cysteine is excluded from randomized positions. 6. Do not perform more than 22 cycles in order to minimize bias in PCR product. 7. Do not perform more than 25 cycles in order to minimize bias in PCR product. 8. β-Agarase I digestion will give better yields and quality of linearized vector DNA compared to other commercial gel extraction kits resulting in better transformation efficiency. 9. Precipitation of vector and insert prior ligation will increase ligation efficiency. 10. It is important to perform steps 12 and 13 very fast to ensure a quick resuspension of pulsed cells for optimal transformation efficiency. 11. At least 1.5 Â 108 transformants should be reached per transformation to get a theoretical library size of 1 Â 1010 clones with a reasonable amount of transformations. For libraries used for affinity maturation a transformation efficiency of 1 Â 108 is sufficient since the aspired theoretical library size will be Construction and Selection of Affilin® Phage Display Libraries 235

1 Â 109. The number of clones on the control plates should be below 5% of the clone number from the ligation reaction. 12. GelRed™ is used as an alternative compared to the supposed cancerogenic ethidium bromide. 13. Clones without insert should not exceed 10%. Otherwise, vector preparation should be repeated. Keep in mind to sub- tract clones without insert for calculating the theoretical library size. 14. Clones carrying deletions or insertions resulting in frameshifts should not exceed 15%. 15. If transformation efﬁciency is better than 1.5 Â 108 (for native library) or 1 Â 108 (for maturation library) the number of transformations (and correspondingly the number of ligation reactions) can be reduced accordingly to reach the desired complexity. 16. If necessary split the transformations in two runs of 35–40 transformations each. Note pulse time and time-point for each transformation when culturing at 37 C in 12 mL culture tube is started. Optimal pulse time should be between 5 and 6 ms per pulse. 17. Pool 10 μL aliquots of 10–11 transformations and take 10 μL out of this pool for analyzing transformation efﬁciency as described in Subheading 3.1.6 step 15. 18. As library cloning is performed via BsaI, a type IIS restriction enzyme, no BsaI sites will be present in vectors containing insert. An additional BsaI restriction digestion of library DNA will solely linearize residual vector without insert, the vector containing insert DNA will remain unaffected. 19. Percentage of clones carrying vector without insert should ideally be zero but must not exceed 2%. Otherwise, repeat BsaI restriction digestion of library DNA and analysis of empty vector content as described in Subheading 3.1.7 steps 9–17. 20. Perform Cleaning-in-Place of the A¨KTA system and columns extensively by incubation at the maximum possible NaOH concentrations. It is recommended to use one column for each phage library to prevent cross-contamination. 21. Additionally, prepare six aliquotes of 100 μL phage solution for monitoring infectivity during long time storage at À80 C. Use these samples to determine phage titer after 1 day, 1 week, 2 weeks, 1, 3, and 6 months. 22. Storage of the library as phage stock is recommended since storage as DNA (Subheading 3.1.7 step 12) would imply to perform transformations of E. coli prior to each selection. 236 Florian Settele et al.

Storage of the library as glycerol stock (of the E. coli suspension scraped from Q-Trays in Subheading 3.1.7 step 7) is also possible but may lead to a decrease in the number of variants or bias in variant occurrence over time.

23. Typically, the culture should reach the desired OD600nm within 1.5–2 h. 24. The ratio of culture volume to absolute volume of the Erlen- meyer flask seems to be critical for optimal phage yields. In our cases, a ratio of 1:8 works best. 25. Beads should be changed at least once during the selection process to avoid binders directed against the bead matrix. It is also possible to use biotinylated target protein. In that case M270-SA beads are used. For pre-selection use empty beads in rounds 1–4. 26. The selection can be set up with the target immobilized on beads before the incubation with phages or as a selection in solution (SIS), where the target is first incubated with phages followed by capturing of target/phage complexes by beads. In that case the layout of plate 1 and the KingFisher Duo program have to be adjusted accordingly. 27. The pre-selection is absolutely critical to remove binders against ProteinA, the bead matrix or the Fc-tag. 28. One comb of the KingFisher is used for the pre-incubation step. A second comb is then used for the target incubation step to avoid cross-contamination with phages binding to the plastic surface of the comb. 29. For competition experiments, different additives (e.g., untagged target for off-rate selection or known ligands to the target for epitope masking selection) can be added directly to the phage solution to guide the selection into the desired direction. We aim for an at least 100-fold molar excess of the additives in relation to the target amount. 30. Trypsin is used in our selection system for the efficient elution of phages from the target. Nevertheless, alternative elution buffers containing for example Glycine for pH-driven elution can be used depending on the design of the construct. 31. Due to the enzymatic activity of the trypsin, storage time on ice should be as short as possible. Good timing with regard to the preparation of the ER2738 culture for re-infection is crucial. 32. In our hands it proved to be extremely important for efficient infection of the bacteria with the phages that the ER2738 culture has at least reached an OD600nm of 2.5. Construction and Selection of Affilin® Phage Display Libraries 237

33. Typically, we perform 3–4 rounds of selection reducing target amounts and increasing number of washing steps in each round. 34. Glycerol stocks of individual rounds can be used as starting point for later variations of the performed selections once an initial, qualitative readout has been obtained. DNA can be used for cloning of maturation libraries. 35. All culture volumes are scalable, depending on the amount of phages needed. For typical rounds 2–4 we routinely use 30 mL culture volume. 36. Also indirect ELISA immobilization of biotinylated target on Streptavidin coated plates is compatible with the used detection system. 37. If comparable overall absorption values are obtained for different rounds of one selection arm, earlier rounds should be chosen for maturation to ensure pool diversity. In later rounds often enrichment of one or few variants is observed. 38. Sequence enrichment of a single binder in a phage pool for maturation should be below 20% to avoid limitation of diversity. 39. Variants of the primary selection with undesirable properties (e.g., binding of an unwanted epitope) can be added to the target solution prior to incubation with the phages to mask certain epitopes. Again, we aim for an at least 100-fold molar excess of the selected variants in relation to the target amount.

Acknowledgment

We thank Anja Kunert for early work on the protocols, Ulrich Haupts for helpful discussions, and Erik Fiedler for chromatographic puriﬁcation of the phage libraries.

References

1. Nord K, Nilsson J, Nilsson B, Uhlen M, Nyg- 4. LaVallie ER, DiBlasio EA, Kovacic S, Grant ren PA (1995) A combinatorial library of an KL, Schendel PF et al (1993) A thioredoxin alpha-helical bacterial receptor domain. Pro- gene fusion expression system that circumvents tein Eng 8:601–608 inclusion body formation in the E. coli cyto- 2. Schneider S, Buchert M, Georgiev O, plasm. Biotechnology (N Y) 11:187–193 Catimel B, Halford M et al (1999) Mutagene- 5. Colas P, Cohen B, Jessen T, Grishina I, McCoy sis and selection of PDZ domains that bind J et al (1996) Genetic selection of peptide new protein targets. Nat Biotechnol aptamers that recognize and inhibit cyclin- 17:170–175 dependent kinase 2. Nature 380:548–550 3. Binz HK, Amstutz P, Kohl A, Stumpp MT, 6. Beste G, Schmidt FS, Stibora T, Skerra A Briand C et al (2004) High-afﬁnity binders (1999) Small antibody-like proteins with pre- selected from designed ankyrin repeat protein scribed ligand speciﬁcities derived from the libraries. Nat Biotechnol 22:575–582 238 Florian Settele et al.

lipocalin fold. Proc Natl Acad Sci U S A charge-charge interactions to protein stability. 96:1898–1903 Biochemistry 38:8138–8149 7. Abedi MR, Caponigro G, Kamb A (1998) 12. Khorasanizadeh S, Peters ID, Butt TR, Roder Green fluorescent protein as a scaffold for H (1993) Folding and stability of a intracellular presentation of peptides. Nucleic tryptophan-containing mutant of ubiquitin. Acids Res 26:623–630 Biochemistry 32:7054–7063 8. Holler PD, Holman PO, Shusta EV, 13. Smith GP (1985) Filamentous fusion phage: O’Herrin S, Wittrup KD et al (2000) In vitro novel expression vectors that display cloned evolution of a T cell receptor with high affinity antigens on the virion surface. Science for peptide/MHC. Proc Natl Acad Sci U S A 228:1315–1317 97:5387–5392 14. Bratkovic T (2010) Progress in phage display: 9. Lorey S, Fiedler E, Kunert A, Nerkamp J, evolution of the technique and its application. Lange C et al (2014) Novel ubiquitin-derived Cell Mol Life Sci 67:749–767 high affinity binding proteins with tumor tar- 15. Paschke M, Ho¨hne W (2005) A twin-arginine geting properties. J Biol Chem 289 translocation (Tat)-mediated phage display sys- (12):8493–8507 tem. Gene 350:79–88 10. Vijay-Kumar S, Bugg CE, Cook WJ (1987) 16. European Patent 1,626,985 B1 Structure of ubiquitin refined at 1.8 A resolu- 17. US Patent 8,791,238 tion. J Mol Biol 194:531–544 18. Monjezi R, Tey BT, Sieo CC, Tan WS (2010) 11. Ibarra-Molero B, Loladze VV, Makhatadze GI, Purification of bacteriophage M13 by anion Sanchez-Ruiz JM (1999) Thermal versus exchange chromatography. J Chromatogr B guanidine-induced unfolding of ubiquitin. An Anal Technol Biomed Life Sci 878:1855–1859 analysis in terms of the contributions from Chapter 12

Construction of a Synthetic Antibody Gene Library for the Selection of Intrabodies and Antibodies

De´borah Caucheteur, Gautier Robin, Vincent Parez, and Pierre Martineau

Abstract

Libraries of antibody fragments displayed on ﬁlamentous phages have proved their value to generate human antibodies against virtually any target. We describe here a simple protocol to make large and diverse libraries based on a single or a limited number of frameworks. The approach is ﬂexible enough to be used with any antibody format, either single-chain (scFv, VHH) or multi-chain (Fv, Fab, (Fab0)2), and to target in a single step the six complementarity-determining regions—or any other part—of the antibody molecule. Using this protocol, libraries larger than 1010 can be easily constructed in a single week.

Key words Kunkel mutagenesis, Antibody fragment, Single-chain Fv, Phage-display, Synthetic library

1 Introduction

With more than 50 molecules already registered, human monoclonal antibodies (mAbs) have proved their value as therapeutic molecules in numerous pathologies [1–3]. These mAbs are usually obtained through three main technologies [4, 5]: mice immunization followed by hybridoma generation, and then in vitro humanization; transgenic mice genetically engineered for producing human antibodies [6]; in vitro methods such as phage, ribosome, or yeast-display [7]. The display-based methods present several advantages over the animal-based ones, in particular their low cost, high ﬂexibility, and high speed, and for therapeutic applications, their ability to directly generate human antibodies and human-mouse cross-reacting mAbs for the preclinical characteriza- tions in rodent models. Among the display methods, phage-display is currently the most widely used approach and has proved to be a cheap and robust technology. This is particularly the case when a naive library is used, since a single-antibody source is used for all the projects. For simple selections, antibodies against any target can be identiﬁed in about 2 weeks and the method can be automated and run in parallel with

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_12, © Springer Science+Business Media LLC 2018 239 240 De´ borah Caucheteur et al.

several antigens [8]. However, few naive and diverse phage-display libraries are publicly available and the first step is thus to develop your own antibody source. For this critical step, you can rely on the natural diversity of human-recombined antibodies by cloning them from a collection of healthy human donors, or you can design the library using structural and sequence information and build it in vitro using molecular biology techniques [9–12]. The former has the advantage of using natural genes but requires many steps and human donors, whereas the latter only uses in vitro techniques but requires a careful design to obtain efficient libraries. Most of the phage-displayed libraries use antibody fragments like scFv, Fab, or VHH fused to the g3p filamentous phage capsid protein. In addition, most authors use a phagemid since it is easier to manipulate than phages and essentially results in a monovalent display of the antibody fragment, allowing a more stringent selection than the multivalent display obtained when using phage vectors. Since the pioneering work of Dr. Kabat [13], it is well known that most of the diversity of the antibody molecule is found in the six complementarity-determining regions (CDRs), three being located in each chain. Because of their critical role in the antibody-antigen interaction, these six regions are the main target for randomization in most synthetic library designs. However, the average contribution of the six CDRs to the binding energy is very variable, the heavy chain CDR2 and CDR3 being the main contributors, and the light chain CDR2 only rarely interacting with the antigen. In addition, there is also a strong bias in the amino-acid distribution in the natural antibody paratopes with a high abundance of tyrosine residues. Using these informations may lead to more clever library designs but this requires a flexible mutagenesis method to construct them [11]. In this protocol, we will describe the construction of a synthetic library, based on a unique framework. We will add diversity in the 6 CDR loops and length variations in the VH CDR3. The protocol is based on the well-known site-directed mutagenesis method developed by Dr. Kunkel [14]. The main interest of this approach is that the library is constructed in a single step using a pool of randomized oligonucleotides, and can address any antibody format. We will introduce diversity in the six CDRs but the protocol can be easily adapted to also introduce diversity to framework regions, to a more restricted number of positions, or using a restrained amino-acid code.

2 Materials

All buffers must be prepared with ultra-pure water and ACS grade chemicals, and stored at room temperature unless otherwise indicated. Construction of a Synthetic scFv Library 241

2.1 Common 1. Ampicillin and/or Carbenicillin stock at 100 mg/mL (see Note 1). Materials Store at À20 C. 2. Chloramphenicol stock at 30 mg/mL in ethanol. Store at À20 C.

3. Kanamycin stock at 25 mg/mL in H2O. Sterilize by ﬁltration and store at À20 C.

4. 2ÂYT medium. In 900 mL of H2O, dissolve 16 g of Tryptone (Peptone), 10 g of yeast extract, and 5 g of NaCl, adjust pH to 7.0 with 5 N NaOH, and then the volume to 1 L. Autoclave and store at RT. 5. 40% glucose solution. Sterilize by autoclaving. Store at RT.

6. LB plates. In 900 mL of H2O, dissolve 10 g of Tryptone (Peptone), 5 g of yeast extract, and 10 g of NaCl, adjust pH to 7.0 with 5 N NaOH, and then the volume to 1 L. Add 15 g of agar and autoclave. Allow the solution to cool to 55–60 C before adding supplements, and then pour the plates. LB/GA plates, add 50 mL of 40% glucose solution and 1 mL of ampicillin (100 μg/mL); LB/C plates, add 500 μL of Chloram- phenicol (15 μg/mL); LB/GAC plates, add 50 mL of 40% glucose solution, 1 mL of ampicillin, and 500 μLof Chloramphenicol. 7. KM13 helper phage [15] stock at 1014 pfu/mL (see Note 2). Store at À70 C. 8. PEG/NaCl solution: 20% (w/v) PEG, 2.5 M NaCl in H2O. Autoclave and store at 4 C(see Note 3).

9. PBS Â10. In 900 mL of H2O, dissolve 80 g of NaCl (1370 mM), 2 g of KCl (27 mM), 11.5 g of Na2HPO4.7H2O (43 mM), 2 g of KH2PO4 (15 mM), adjust volume to 1 L, autoclave and store at RT. 10. Macherey Nagel NucleoSpin plasmid kit and NT2 buffer (see Note 4). 11. 50 mL conical centrifuge tubes (see Note 5) and a refrigerated centrifuge. 12. 1.5 mL microcentrifuge tubes and a refrigerated bench-top centrifuge (see Note 6).

2.2 Preparation 1. E. coli K12 CJ236: FΔ(HindIII)::cat (Tra+ Pil+ CamR)/ung-1 of Uracil-Containing relA1 dut-1 thi-1 spoT1 mcrA (see Note 7). Single-Stranded DNA 2. A phage or phagemid containing your antibody gene (see Note 8). Template In the protocol below we will use pHEN1 phagemid vector [16] containing the scFv13R4 [17].

3. Uridine stock. 6 mg/mL in H2O, sterilize by ﬁltration, and store at À20 C. 242 De´ borah Caucheteur et al.

2.3 Synthesis 1. Stock at 100 μM of the 50-phosphorylated mutagenenic pri- of the Mutagenized mers in Tris–HCl pH 8/EDTA 1 mM (Fig. 1)(see Note 9). Complementary Strand Store at À20 C. 2. NEB2 Â10 buffer (1Â: 10 mM Tris–HCl pH 7.9 at 25 C, 50 mM NaCl, 10 mM MgCl2, 1 mM DTT). Store at À20 C. 3. dNTPs (25 mM each). Store at À20 C. 4. T4 DNA Ligase (5 U/μL) and 10Â buffer (400 mM Tris–HCl, 100 mM MgCl2, 100 mM DTT, 5 mM ATP, pH 7.8 at 25 C; Thermoﬁsher). 5. T7 DNA Polymerase (NEB, 10 U/μL).

Fig. 1 Schematic view of the experiment. The main steps of the protocol are depicted in the figure. (a) Production of uracil-containing ssDNA template in a dut ung E. coli strain (Subheading 2.2); (b) Hybridization of the collection of mutagenic oligonucleotides and synthesis of the complementarity DNA strand using T7 polymerase and T4 DNA ligase (Subheading 2.3); (c) Transformation in a ung+ strain, elimination of the template ssDNA, and phage production originating from the newly synthesized mutant strand (Subheading 2.5). (d) The boxes at the center of the figure illustrate the design of the mutagenic oligonucleotides. For each CDR an oligonucleotide is designed that hybridizes perfectly to each side of the CDR and with a central degenerate sequence covering the CDR. Each perfect-match is 10–20 nucleotide-long with a Tm of around 45 C. The oligonucleotides are 50-phosphorylated to allow ligation and thus closing of the DNA by the T4 DNA ligase. Many different designs are possible, in particular the degenerate sequence can only partially cover the CDR and use optimized mixes of bases or trinucleotide precursors to avoid stop codons and precisely define the resulting degeneracy. Construction of a Synthetic scFv Library 243

2.4 Preparation 1. E. coli TG1 (see Note 10). of Electrocompetent 2. 500 mL centrifuge bottles. Bacteria 3. Sterile magnetic stir bars. 4. Magnetic stirrer.

5. HEPES 1 M: weight 2.38 g of HEPES, add 8 mL of H2O, adjust pH to 7.0 and the volume to 10 mL. Sterilize by ﬁltra- tion. Store at 4 C. 6. Glycerol/HEPES: Weigh 10 g of glycerol, make up to 1 L with water, autoclave. Add 1 mL of sterile HEPES 1 M, store at 4 C.

7. H2O/HEPES: Add 1 mL of sterile HEPES 1 M to 1 L of autoclaved ultra-pure water, store at 4 C.

2.5 Electroporation 1. SOC medium. In 950 mL of H2O, dissolve 20 g of tryptone and Phage Production (Peptone), 5 g of yeast extract, 0.5 g of NaCl, 10 mL of 250 mM KCl (18.6 g/L). Make up to 1 L with water, adjust pH to 7.0 with 5 N NaOH, autoclave. Before use, add 5 mL of sterile 2 M MgCl2 (190.4 g/L, autoclaved) and 9 mL of sterile 40% glucose (20 mM). 2. 14 mL sterile polypropylene round-bottom culture tubes (17 mm Â 100 mm). 3. Biorad GENE PULSER II and 0.2 cm gap cuvettes (see Note 11). 4. Sterile Pasteur pipettes.

3 Methods

3.1 Preparation In this protocol, we will ﬁrst infect a dut ung strain with our of Uracil-Containing phagemid, we then make a stock of phages using KM13 helper Single-Stranded DNA phage, and ﬁnally purify the single-stranded DNA (ssDNA) encap- Template sided in the phage particles. This ssDNA will contain uracil instead of thymine and will be used in the next protocol. 1. Pick a single fresh colony of CJ236 in 2 mL of 2ÂYT and grow overnight (ON) at 37 C with shaking. 2. Add 20 μL of the preculture in 2 mL of 2ÂYT. 3. Grow with vigorous agitation at 37 C until the OD600nm reaches 0.5. 4. Add 10 μL of a pHEN1-13R4 phage stock diluted to 106 cfu/ mL (see Note 12). 5. Incubate for 1 h at 37 C without or with a slow shaking (0–100 rpm). 244 De´ borah Caucheteur et al.

6. Plate 100 μL on LB/GAC plates and incubate ON at 37 C(see Note 13). 7. Pick a colony in 2 mL of 2ÂYT with 2% glucose and 100 μg/ mL Ampicillin and grow ON at 37 C with agitation. 8. Add 20 μL of the preculture in 2 mL of 2ÂYT with 2% glucose and 100 μg/mL Ampicillin and grow with vigorous agitation until OD600nm reaches 0.5. 9. Add 20 μL of a 1/100th dilution of the KM13 helper phage (1012 pfu/mL) (see Note 14). 10. Incubate without shaking for 30 min at 37 C(see Note 15). 11. Centrifuge at 4 C for 15 min at 3000 Â g. 12. Discard the supernatant. 13. Resuspend the pellet in 30 mL of 2ÂYT with 25 μg/mL of Kanamycin and 100 μg/mL of Ampicillin supplemented with 0.25 μg/mL of Uridine (1.25 μL of stock solution). 14. Grow ON with vigorous agitation (220–240 rpm) at 37 Cina 150 mL flask. 15. Transfer in a 50 mL conical centrifuge tube. 16. Centrifuge for 10 min at 12,000 Â g at 2 C(see Note 5). 17. Transfer the 30 mL of supernatant containing the phages into a 50 mL centrifuge tube. 18. Add 8 mL of cold PEG/NaCl solution and mix thoroughly by inverting the tube several times. 19. Incubate for 30 min at 4 C on ice with regular mixing. 20. Centrifuge for 10 min at 12,000 Â g at 2 C. 21. Remove the supernatant by inverting the tube with caution and put it gently upside down on absorbent paper to remove excess liquid (see Note 16). 22. Spin briefly and remove the remaining liquid with a pipette and using absorbent paper as in the previous step. 23. Resuspend the phage pellet in 0.5 mL of PBS using an aerosol-free tip (see Note 17) and transfer it in a 1.5 mL microcentrifuge tube. 24. Centrifuge at 4 C for 5 min at 16,000 Â g in a bench-top centrifuge to pellet any insoluble material (see Note 18). 25. Purify the ssDNA from the supernatant using Macherey Nagel NucleoSpin plasmid kit using the supplementary protocol for the isolation of M13 DNA (see Note 4). 26. Elute the ssDNA in 100 μL of 5 mM Tris–HCl, pH 8.5 (kit AE buffer). 27. Quantify the ssDNA in a spectrophotometer (see Note 19). 28. Store the purified ssDNA at À20 C. Construction of a Synthetic scFv Library 245

3.2 Synthesis Diversity in the library is introduced using degenerate oligonucleo- of the Mutagenized tides. The pool of phosphorylated oligonucleotides is ﬁrst hybri- Complementary Strand dized to the single-stranded template prepared in Subheading 3.1, then these oligonucleotides are used as primers and elongated using T7 polymerase, and ﬁnally the gaps closed using T4 DNA ligase. This results in a double-stranded circular DNA with a uracil- containing strand coding a wild-type scFv gene and a thymine- containing strand coding for the scFv library. Each reaction prepared in this section will be used in a single electroporation experiment (Subheading 3.4) and should generate between 5 Â 108 and 5 Â 109 clones. We usually make the following reaction independently for each VH-CDR3 loop length (5–10 lengths) to generate a large library of at least 5 Â 109 clones. 1. Annealing of the mutagenic primers to the ssDNA template.

l Prepare a primer mix in H2O, each primer at a 10 μM ﬁnal concentration (see Note 20): 10 μL of each primer in a ﬁnal volume of 100 μLofH2O. l Prepare an oligonucleotide:template mix with a 6:1 molar ratio: 12.5 pmol of uracil-containing ssDNA (26 μgto 25 μL), 7.5 μL of the primer mix (75 pmol of each primer), 25 μL of NEB2 10Â, and H2O up to 250 μL(see Note 21). l Transfer to a thermal cycler at 90 C for 2 min, 4 C10s, 45 C for 20 min, and 20 C for 10 min (see Note 22). Store at 4 C. 2. Add in the following order to the 250 μL of annealed oligonucleotide:template mix.

l 13 μLH2O (final volume of 350 μL). l 40 μL dNTPs (2.9 mM final) (see Note 23). l 35 μL10Â Ligase buffer (1Â final). l 8 μL T4 DNA ligase (40 Weiss units). l 4 μL T7 DNA Polymerase (40 units). 3. Incubate at 20 C for 5 h. 4. Purify DNA on a NucleoSpin Plasmid column using the “Plas- mid DNA clean-up” procedure (see Note 24). 5. Elute the DNA in 40 μL of AE buffer (5 mM Tris/HCl, pH 8.5) heated to 70 C to maximize yield (see Note 25). 6. Optional. Run an agarose gel to check the efficiency of the second-strand synthesis. l Pour a 1% agarose gel without any intercalating agent using TAE buffer (see Note 26). l Analyze 2 μL of your mutagenesis in parallel with the same amount of ssDNA (~0.5 μL). 246 De´ borah Caucheteur et al.

Fig. 2 Quality control of the dsDNA synthesis. Efﬁcacy of second-strand synthesis is analyzed by agarose gel electrophoresis. The ssDNA template (lane 1) migrates much faster in the absence of intercaling agent than the dsDNA synthetized in Subheading 2.3 (lanes 2–9). In this example, eight mutagenesis experiments were performed in parallel with eight VH-CDR3 loop lengths (indicated above the lanes). Because the analyzed DNAs are circular, their migration cannot be compared with the linear MW marker (1 kb Generuler, Fermentas)

l Run the gel for 1 h at 5 V/cm. l Incubate the gel for 30–60 min in a solution of TAE with 10 μg/mL of SYBR Safe DNA Gel Stain (see Note 27). A typical result is shown in Fig. 2.

3.3 Preparation Use freshly prepared electrocompetent cells following the protocol of Electrocompetent below in order to obtain the high transformation efficiency (typi- Bacteria cally 5 Â 109–2Â1010 transformants/μg of supercoiled pUC18 plasmid) required for the final library transformation (see Note 28). 1. All material must be precooled and kept as close to 4 Cas possible in an ice/water bath throughout the preparation (see Note 29). If possible, work in a cold room. The centrifuge and the rotor must be precooled to 4 C. 2. Pick a fresh colony of TG1 in a 50 mL flask containing 10 mL of 2ÂYT, and grow ON at 37 C with vigorous shaking (220 rpm) (see Note 10). 3. Pour the flask content in a 5 L flask containing 1 L of 2ÂYT, and grow at 37 C with vigorous shaking (220–240 rpm) until OD600nm reaches 0.7. 4. Pour the flask content in two 500 mL centrifuge bottles and cool down in an ice/water bath for 30 min. Mix regularly and gently the bottles. Construction of a Synthetic scFv Library 247

5. Centrifuge at 5000 Â g for 5 min at 4 C and discard the supernatant. 6. Add a cold and sterile magnetic stir bar and 500 mL of cold H2O/HEPES (1 mM) to each bottle. Resuspend the pellet using a magnetic stirrer. Start with a vigorous stirring until the pellet detaches from the bottle; continue with a slower rotation rate until all the bacteria are completely resuspended. You may also gently mix the bottle by turning it upside down several times. 7. Centrifuge at 5000 Â g for 10 min at 4 C and discard the supernatant gently, carefully avoiding disturbing the pellet containing the stir bar. 8. Repeat steps 6 and 7. 9. Resuspend, as in step 6, in 50 mL of cold glycerol/HEPES. Pool the two bottles in a new centrifuge bottle. Do not transfer the stir bars. 10. Centrifuge at 5000 Â g for 15 min at 2 C and discard the supernatant. 11. Resuspend the pellet in 1 mL of cold glycerol/HEPES using a cold 10 mL pipette. The ﬁnal volume should be around 2 mL (see Note 30).

3.4 E. coli If you prepared your own electrocompetent cells in Subheading Electroporation 3.3, you must immediately proceed and electroporate your DNA and Phage Production since transformation efﬁciency will decrease if cells are frozen. Each mutagenesis prepared in step 5 of Subheading 5 will generate, in a single electroporation experiment, between 5 Â 108 and 5 Â 109 clones. In this protocol, we directly make the stock of phages that can be then used in phage display experiments. With the volumes used below, we typically obtain enough aliquots for 1500 selections. 1. Prepare one sterile 50 mL centrifuge tube for each DNA preparation (Subheading 5) containing 12 mL of SOC and two 14 mL sterile polypropylene culture tubes containing 0.95 mL of SOC. 2. Warm these tubes to 37 C for at least 1 h. 3. Cool on ice: one electroporation cuvette for each DNA preparation, and one for the positive control; the same number of sterile microcentrifuge tubes; and the slide that holds the cuvette in the electroporator (see Note 31). 4. In a prechilled microcentrifuge tube, mix 350 μL of competent cells and the puriﬁed ligation (35–40 μL, prepared in step 5 of Subheading 5). Do not pipet up and down to mix since this will warm the cells. 248 De´ borah Caucheteur et al.

5. Transfer the mix in a prechilled electroporation cuvette. Be sure to put the sample at the bottom of the cuvette by gently taping the bottom of the cuvette on a flat surface, and avoid introducing bubbles. Quickly wet the cuvette and the cuvette slide with absorbent paper, then assemble them in the electroporator. 6. Apply an electric pulse using the following conditions: 2500 V, 25 μF, 200 Ω. 7. Immediately transfer the cells to one of the pre-warmed sterile 50 mL centrifuge tube containing 12 mL of SOC by washing the sample with 1 mL of outgrowth medium using a Pasteur pipette (see Note 32). 8. Immediately transfer the tube to a 37 C incubator and shake vigorously (220 rpm) for 1 h. 9. Repeat steps 4–8 with the other synthesized DNA. 10. Negative control: Add 40 μL of competent cells to one of the pre-warmed 14 mL tubes of SOC. 11. Positive control: Add 1 μL of a highly purified supercoiled pUC18 (10 pg/μL) plasmid to 40 μL of competent cells in one of the prechilled microcentrifuge tubes. Follow steps 5–8 but resuspend in 0.95 mL of SOC using the second 14 mL pre-warmed tube. 12. Plate on LB/GA plates: 100 μL of the negative control; 100 μL À À of 10 1 and 10 2 dilutions of the positive control; 100 μLof À À À À 10 2, 10 3,10 4 and 10 5 dilutions of each 50 mL conical tube (containing 12 mL of SOC and transformed bacteria). 13. Transfer the content of each 50 mL centrifuge tube (12 mL of SOC with transformed cells) in a 1 L flask containing 200 mL of 2ÂTY with 2% glucose and 100 μg/mL Carbenicillin (see Note 1). 14. Incubate at 37 C with shaking (220 rpm) until OD600nm reaches 0.5. 15. Add 20 μL of KM13 helper phage at 1014 pfu/mL (20-fold excess). 16. Incubate for 30 min at 37 C without or with a slow shaking (0–100 rpm). 17. Centrifuge at 3000 Â g for 20 min at 4 C. 18. Resuspend each pellet in 500 mL of 2ÂTY with 25 μg/mL Kanamycin and 100 μg/mL Carbenicillin. 19. Incubate ON in a 2 L flask at 37 C with agitation (220–240 rpm). Construction of a Synthetic scFv Library 249

20. Calculate the size of the library and the transformation efficiency using the series of dilutions plated in step 12 (see Note 33). 21. Pool all the flasks and centrifuge for 30 min at 10,000 Â g at 4 C. 22. Recover the supernatant and add 1/5th of the volume of PEG/NaCl (200 mL/L of supernatant). 23. Incubate on ice in a cold room with regular mixing for at least 1h. 24. Centrifuge for 30 min at 10,000 Â g at 4 C and discard the supernatant. 25. Spin briefly and eliminate any remaining drop of PEG/NaCl. 26. Resuspend all the pellets in a total volume of 300 mL of cold PBS with 15% glycerol. 27. Add 75 mL of PEG/NaCl for a second precipitation and proceed as before (steps 23–26); resuspend the pellet in 80 mL of cold PBS with 15% glycerol (see Note 34). 28. Centrifuge for 30 min at 10,000 Â g at 4 C and recover the supernatant containing the phages. 29. Estimate phage concentration using UV absorbance with the 13 formula: phages/mL ¼ (A269nm À A320nm) Â 10 (see Note 35). 30. Aliquot in 50 μL and store at À70 C(see Note 36).

4 Notes

1. Carbenicillin and Ampicillin can be alternatively used. How- ever, since it is more stable, we prefer to use Carbenicillin for the last step of the library production (Subheading 3.4). To make a stock solution of Carbenicillin or Ampicillin at 100 mg/ mL ﬁnal: dissolve 1 g of powder in 9 mL of H2O, adjust to 10 mL with H2O, sterilize by ﬁltration, and store in aliquots at À20 C. 2. We use here KM13 helper phage that confers resistance to kanamycin. M13KO7 or another helper phage can be alternatively used. 3. The quality of the PEG is critical (e.g., PEG 8000 for molecular biology from Sigma #81268). 4. Nucleospin plasmid kit can be used to purify single-stranded M13 DNA. Macherey Nagel provides an additional protocol for this application that can be downloaded from their web site [18] or obtained on request. This protocol requires a buffer 250 De´ borah Caucheteur et al.

not present in the kit (NT2, #740597). Alternatively, phenol extraction and ethanol precipitation can be used [14]. 5. Be sure that the tubes are resistant enough. Falcon (#352070) and Corning (#430290) branded 50 mL polypropylene conical centrifuge tubes are resistant to 16,000 Â g. 6. If not refrigerated, put the centrifuge in a cold room. Phages are very stable even at high temperature but the expressed scFv are heat-sensitive. 7. E. coli K12 CJ236 can be obtained from NEB. Streak out the strain on LB agar containing chloramphenicol (15 μg/mL) to ensure that you start with an F+ host, but do not include chloramphenicol in liquid media. 8. The protocol is ﬂexible enough to work with any antibody 0 format (scFv, Fv, Fab, (Fab )2, VHH, etc.), but requires a phage or phagemid vector. The mutagenic oligonucleotides must be complementary to the encapsided (þ) strand that can be either the coding or the noncoding strand depending on the cloning orientation. 9. High-quality oligonucleotides must be used. The best is to order cloning-quality 50-phosphorylated oligonucleotides. À À 0 10. TG1: supE thi-1 Δ(lac-proAB) Δ(mcrB-hsdSM)5 (rk mk )[F traD36 proAB lacIqZΔM15]. For phage display it is critical to check for F0 presence. For this reason you must keep TG1 on a synthetic plate without proline (proAB), for instance M9 plates with glucose and thiamine (thi-1). Use a recently streaked plate of less than 1 week. 11. BioRad 0.2 cm gap cuvettes allow the electroporation of 400 μL of cells. Other brands are possible but check the size of the cuvette. 12. cfu: colony-forming unit. We use a large excess of bacteria to ensure that all the phages can infect a bacterium. 13. We use glucose in all the plates because the scFv is under the control of the lac promoter in pHEN1. This ensures a strong repression of the gene and avoids toxicity. You should get 100–1000 colonies. 8 14. At an OD600nm of 0.5, you should have around 5 Â 10 bacteria/mL. To infect all the bacteria, a 20-fold excess of phages is used, that is 1010 KM13 per mL of culture. 15. Avoid vigorous shaking but a slow shaking (100 rpm) is also possible. 16. Empty the supernatant into a liquid trash by inverting the tube, then, without reverting it, place the tube open upside down on a piece of paper towel to absorb the remaining liquid. Construction of a Synthetic scFv Library 251

17. Clean the hood with phagospray and use filtered pipette tips to prevent contaminations with filamentous phages. 18. The pellet may be absent since it is essentially due to bacteria that were not fully eliminated by the first centrifugation step. 19. When using a 1 cm path length, a 33 μg/mL solution of single- stranded DNA has an absorbance of 1 at 260 nm. The yield should be around 75 μg. This is enough for three large-scale mutagenesis experiments. 20. We usually use a single mutagenic oligonucleotide for the CDR1s, CDR2s and the VL-CDR3, and a series of oligonucleotides of different lengths for the VH CDR3. We prepare one mix for each VH-CDR3 length (10 μL of each oligonucleotide) to ensure an equal representation of the CDR3 lengths in the library. You can however mix together all the oligonucleotides if you need a library of a more limited size. 21. pHEN1-13R4 vector is 5229 bases long. If your phagemid contains N bases, 15 pmol of ssDNA represents À (15 Â 330 Â N Â 10 6)~N/200 μg. 22. It is extremely important to cool down very quickly from 90 to 4 C to avoid a hybridization bias due to a partial matching between the degenerated oligonucleotide and the original CDR sequence. You can use a thermal cycler or simply boil your sample and transfer it directly into an ice bucket. The third step at 45 C is to remove wrongly hybridized oligonucleotides with a low Tm. 23. This is a very high dNTP concentration, much higher that what is used in most protocols (0.1–0.6 mM). With the classical dNTP concentrations, we only get few dsDNA. For the ssDNA given in this protocol, at least 1.25 mM is required. 24. This is page 23 of the current manual. You can also heat- inactivate the reaction and purify by precipitation. 25. See Subheading 2.5 “Elution procedures” in the Macherey- Nagel manual. 26. Intercalating agents change the DNA supercoiling state and the migration speed. Resolution of single and double-stranded DNA is much better in their absence. For 1 L of TAE Â50: Tris 242 g, 57.1 mL acetic acid, 100 mL EDTA 0.5 M, pH 8.0 (NaOH). 27. SYBR Safe DNA Gel Stain (Thermofisher). Any DNA stain can be used, e.g., Ethidium Bromide. 28. Alternatively you can use electrocompetent TG1 from Luci- gen. You can contact them to get bulk quantities (12 Â 500 μL). 252 De´ borah Caucheteur et al.

29. Do not use ice but an ice/water mix to optimize temperature exchange. 30. We use a 10 mL pipette to avoid stressing the bacteria by shearing. You can use a 1 mL micropipette by cutting the tip at around 5 mm from the extremity. With the volume of competent cells prepared here you can perform up to six large electroporation experiments. 31. See fig. 7 in the Biorad technical note MC1652101C (http:// www.bio-rad.com/cmc_upload/Literature/12864/ M1652101C.pdf). 32. The period between applying the pulse and transferring the cells to the outgrowth medium is critical for efficient recovering of E. coli transformants. Delaying this transfer by even one minute causes a threefold drop in transformation efficiency. 33. If the library is not large enough and the transformation efficiency lower than 5 Â 109 you must improve electrocompetent cell preparation or use commercial ones. If the cells are competent enough, the problem comes presumably from the dsDNA preparation (Subheading 2.3). Check the efficiency of this step by analyzing your sample on an agarose gel: no ssDNA but a strong dsDNA band should be visible (Fig. 2). 34. This volume is suitable if you have pooled 5–10 transformations. You can scale it down or up depending on the anticipated results (see Notes 35 and 36). 35. The formula depends on the phage/phagemid size 13 [19]. Phages/mL ¼ (A269nm À A320nm) Â 6 Â 10 /(number of bases of the phage ssDNA). It is better to also titer the infectious phages using serial dilutions, infection of mid-log TG1, then plating on LB/GA (cfu/mL). In general, infectious phages represent 10–50% of the UV-determined particles; e.g., in one of the libraries made in the group, we obtained at this step 2.2 Â 1014 phages/mL using UVand 3 Â 1013 cfu/mL by titration. 36. Each aliquot should be a hundred times larger than the library size measured by titration. For most libraries this means around 1012 phages. It is convenient to aliquot into strips of 8 Â 0.2 μL PCR tubes.

References

1. Scolnik PA (2009) mAbs: a business perspec- 3. Reichert JM (2016) Antibodies to watch in tive. MAbs 1:179–184 2017. MAbs 9:167–181 2. Reichert JM (2012) Marketed therapeutic anti- 4. Strohl WR (2014) Antibody discovery: sourc- bodies compendium. MAbs 4:19–21 ing of monoclonal antibody variable domains. Curr Drug Discov Technol 11:3–19 Construction of a Synthetic scFv Library 253

5. Nelson AL, Dhimolea E, Reichert JM (2010) 12. Sidhu SS, Fellouse FA (2006) Synthetic thera- Development trends for human monoclonal peutic antibodies. Nat Chem Biol 2:682–688 antibody therapeutics. Nat Rev Drug Discov 13. Johnson G, Wu TT (2000) Kabat Database and 9:767–774 its applications: 30 years after the first variabil- 6. Jakobovits A, Amado RG, Yang X et al (2007) ity plot. Nucleic Acids Res 28:214–218 From XenoMouse technology to panitumu- 14. Kunkel TA (1985) Rapid and efficient site- mab, the first fully human antibody product specific mutagenesis without phenotypic selec- from transgenic mice. Nat Biotechnol tion. Proc Natl Acad Sci U S A 82:488–492 25:1134–1143 15. Kristensen P, Winter G (1998) Proteolytic 7. Frenzel A, Schirrmann T, Hust M (2016) selection for protein folding using filamentous Phage display-derived human antibodies in bacteriophages. Fold Des 3:321–328 clinical development and therapy. MAbs 16. Hoogenboom HR, Griffiths AD, Johnson KS 8:1177–1194 et al (1991) Multi-subunit proteins on the sur- 8. Schofield DJ, Pope AR, Clementel V et al face of filamentous phage: methodologies for (2007) Application of phage display to high displaying antibody (Fab) heavy and light throughput antibody generation and character- chains. Nucleic Acids Res 19:4133–4137 ization. Genome Biol 8:R254 17. Martineau P, Jones P, Winter G (1998) Expres- 9. Philibert P, Stoessel A, Wang W et al (2007) A sion of an antibody fragment at high levels in focused antibody library for selecting scFvs the bacterial cytoplasm. J Mol Biol expressed at high levels in the cytoplasm. 280:117–127 BMC Biotechnol 7:81 18. Supplementary protocols plasmid DNA. 10. Robin G, Martineau P (2012) Synthetic custo- http://www.mn-net.com/tabid/12238/ mized scFv libraries. Meth Mol Biol (Clifton, default.aspx NJ) 907:109–122 19. Day LA, Wiseman RL (1978) A comparison of 11. Robin G, Sato Y, Desplancq D et al (2014) DNA packaging in the virions of fd, Xf, and Restricted diversity of antigen binding residues Pf1. Cold Spring Harbor Monogr Arch of antibodies revealed by computational ala- 08:605–625 nine scanning of 227 antibody-antigen complexes. J Mol Biol 426:3729–3743 Chapter 13

Targeting Intracellular Antigens with pMHC-Binding Antibodies: A Phage Display Approach

Zhihao Wu, Brian H. Santich, Hong Liu, Cheng Liu, and Nai-Kong V. Cheung

Abstract

Antibodies that bind peptide-MHC (pMHC) complex in a manner akin to T-cell receptor (TCR) have not only helped in understanding the mechanism of TCR-pMHC interactions in the context of T-cell biology, but also spurred considerable interest in recent years as potential cancer therapeutics. Traditional methods to generate such antibodies using hybridoma and B-cell sorting technologies are sometimes inadequate, possibly due to the small contribution of peptide to the overall B-cell epitope space on the surface of the pMHC complex (typical peptide MW ¼ 1 kDa versus MHC MW ¼ 45 kDa), and to the multiple efficiency limiting steps inherent in these methods. In this chapter, we describe a phage display approach for the rapid generation of such antibodies with high specificity and affinity.

Key words Phage display, Phage, Human leukocyte antigen, Major histocompatibility complex, Antibody, Protein expression, Fc-fusion protein, Single-chain variable fragment, scFv, T-cell receptor, Peptide MHC

1 Introduction

Since the first FDA-approved antibody drug in 1985, antibody therapeutics has become an established regimen in multiple types of diseases, especially cancers and autoimmune diseases. However, all antibody drugs approved so far have been targeting secreted or surface expressed antigens, while most neoantigens—transcription factors and signal transducers—are intracellularly expressed and are essentially “undruggable” using antibodies. The finding that degraded products of some of these intracellular neoantigens can gain access to cell surface through binding to major histocompatibility complex I (MHCI) has offered a new strategy to target abnormal cells expressing these antigens, which is to generate antibodies against these specific peptide-MHC (pMHC) complexes on the cell surface. Difficulties in this strategy lie in the generation of such antibodies through traditional hybridoma or B-cell sorting

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_13, © Springer Science+Business Media LLC 2018 255 256 Zhihao Wu et al.

methods, mainly due to the small exposed surface area of neo-epitopes in the pMHC complex and the uncontrolled specificity during animal immunization. Advances in protein engineering technologies, in particular multiple display methods, have largely overcome these limitations and stimulated new interest in the past few years in generating such antibodies. In this chapter, we describe a phage display approach to develop such antibodies. George Smith [1] 30 years ago demonstrated for the first time that bacteriophage can be engineered to display functional polypeptides on the surface and this engineered phage can be enriched by affinity. Since then, this phage display method has been extensively modified and optimized and has become a powerful tool in both antibody discovery and engineering [2–14]. In brief, the method involves (1) binding of target antigens to rare clones in a library of phages; (2) the separation of the binding phages from the non-binding ones; (3) amplification of binding phages; and (4) repetition of the process. A library typically contains tens of trillions of phages, each of which displays an antibody fragment (scFv or Fab); in total these phages make up billions of different binding specificities in a library. The whole process is independent of immune responses seen in immunization and can be carried out ex vivo with commonly available laboratory wares. With increasing size of various synthetic and immune libraries, in theory antibodies can be identified against any antigen categories (proteins, sugars, lipid, DNA), provided the library is diverse enough and the antigen is pure/stable enough. Based on these principles, pMHC-specific antibodies can be obtained by screening a phage library against target pMHC, with irrelevant pMHC (or a panel of them) as controls to eliminate nonspecific binders. This method has been used with great success to develop antibodies against an expanding list of intracellular targets like Wilms tumor 1 (WT1) [15], MAGE-A1 [16], and alpha-fetoprotein (AFP) [17]. Depending on the specific goal of the project, the identified scFv can be reformatted to obtain different functionalities (e.g., scFv-Fc fusion, T-cell recruiting bispecific full-length IgG, CAR-T cells, antibody-drug conjugates, etc.) for further development. The following protocol describes in detail how to utilize a phage library to generate TCR-like scFv’s using readily available commercial phage libraries. As an example for further development, we also detail a protocol for reformatting scFv to scFv-Fc fusion and purification of such fusion proteins. While this protocol focuses on scFv libraries, the same method can be applied to Fab libraries as well [16, 18–20], and in fact is recommended if the final desired format is full-length IgG. Targeting Intracellular Antigens with pMHC-Binding Antibodies 257

2 Materials

Prepare all buffer solutions with milliQ water or double distilled water (ddH2O) and analytical grade reagents. All reagents can be kept at room temperature unless otherwise speciﬁed. Follow all normal waste removal rules and regulations of the institution when disposing of reagents. To prevent phage cross contamination, bleach and autoclave all reusable laboratory wares. Phages are quite robust and can survive normal autoclaving procedures. Similarly, be sure to always use pipette tips with ﬁlters.

2.1 Phage Panning 1. Bovine Serum Albumin (BSA). Store at 4 C. 2. Phosphate-Buffered Saline (PBS) (1Â): Weigh 8 g of NaCl, 2 g of KCl, 17 g of Na2HPO4, and 1.63 g of KH2PO4 and dissolve in 1 l of H2O. 3. Blocking buffer: 2% (w/v) BSA in PBS. Weigh 2 g of BSA and dissolve in 100 ml of PBS. Store at 4 C. ® 4. Dynabeads M-280 Streptavidin (Thermo Fisher Scientiﬁc). Store at 4 C. 5. HuScL-3(R): Human Single Chain (scFv) Antibody Library. Store at À80 C. 6. 6-Tube Magnetic Separation Rack (New England Biolabs). 7. Biotinylated MHC monomer with target peptide (Target pMHC). Store at À80 C 8. Biotinylated MHC monomer with irrelevant peptide (Control pMHC). Store at À80 C 9. Wash Buffer: PBS containing 0.1% (v/v) Tween-20 (PBS-T). Add 1 ml Tween-20 to 1 l of PBS. 10. Elution Buffer: 0.2 M glycine-HCl. Weigh 1.5 g of glycine and 100 mg of BSA and dissolve in 100 ml of H2O. Adjust to pH 2.2 with HCl. Store at 4 C. 11. Neutralization Buffer: 2 M Tris–HCl. Weigh out 24.2 g of Tris–HCl and dissolve it in 100 ml of H2O. Adjust to pH 9.0 with HCl.

2.2 Ampliﬁcation 1. XL1-Blue E. coli. Store at À80 C. 2. Tetracycline (Tet) Stock: 5 mg/ml in ethanol. Store at À20 C.

3. Ampicillin (Amp) Stock: 100 mg/ml in H2O. Sterile ﬁlter and store at À20 C. 4. Lysogeny broth (LB): Weigh 10 g tryptone, 5 g yeast extract, and 10 g of NaCl and dissolve in 1 l H2O. Autoclave and store at 4 C. 258 Zhihao Wu et al.

5. LB with tetracycline (LB-Tet): prepare LB as described above. After autoclaving let the LB cool. Once the LB is below 55 C dilute 2 ml of tetracycline stock per 1 l of LB for a ﬁnal concentration of 10 μg/ml. Store at 4 C. 6. Shaking Incubator: Temperature set at 37 Corat30C. Rotate at 200–300 rpm. 7. Spectrophotometer: Capable of reading samples at 600 nm (OD600). Always blank with the correct growth medium. 8. LB-Agar with glucose, Amp and Tet (LB-GAT plates): Prepare LB as described above. Before autoclaving add 15 g of agar and 20 g of glucose per 1 l of LB. After autoclaving let the LB-Agar cool. Once the LB-Agar is below 55 C dilute 1 ml of ampicillin stock and 2 ml of tetracycline stock per 1 l of LB-Agar, for a ﬁnal concentration of 100 μg/ml Amp and 10 μg/ml of Tet. Pour onto 10 and 15 cm plates at desired thickness. Store at 4 C. 9. LB with glucose, Amp and Tet (LB-GAT): Prepare LB-GAT as described above (LB-GAT plates) but do not add any agar. Store at 4 C. 10. M13 K07 Helper phage. Store at 4 C.

11. Kanamycin Stock: 50 mg/ml in H2O. Sterile ﬁlter and store at À20 C. 12. PEG/NaCl (5Â): Weigh 200 g of polyethylene glycol-8000 (20% w/v) and 150 g of NaCl (2.5 M) in 1 l H2O and autoclave.

2.3 Clone Selection 1. HB2151 E. coli. 2. LB with ampicillin (LB-Amp): prepare LB as described above. After autoclaving let the LB cool. Once the LB is below 55 C dilute 1 ml of ampicillin stock per 1 l of LB for a final concentration of 100 μg/ml. Store at 4 C. 3. LB-Agar with ampicillin (LB-Amp plates): Prepare LB as described above. Before autoclaving add 15 g of agar per 1 l of LB. After autoclaving let the LB-Agar cool. Once the LB-Agar is below 55 C, dilute 1 ml of ampicillin stock for a final concentration of 100 μg/ml. Pour into 10 cm plates at desired thickness. Store at 4 C. 4. 2 ml 96-well DeepWell™ plate (Thermo Fisher Scientific). 5. Axygen™ microplate sealing film (Fisher Scientific) or any other sealing films that allow gas exchange and prevent exces- sive evaporation of liquid. 6. Isopropyl-1-thio-β-D-galactopyranoside (IPTG): weigh 6.0 g IPTG and dissolve in 50 ml of H2O. Filter sterilize and store at À20 C in 1 ml aliquots. Targeting Intracellular Antigens with pMHC-Binding Antibodies 259

2.4 Screening 1. 96-Well EIA microtiter plate. of Phage Clones 2. BSA-biotin. Reconstitute per the manufacturer’s instructions and store at À20 or À80 C. 3. Streptavidin. Reconstitute per the manufacturer’s instructions and store at À20 or À80 C. 4. Biotinylated MHC monomer with target peptide (Target pMHC). Store at À80 C. 5. Biotinylated MHC monomer with irrelevant peptide (Control pMHC). Store at À80 C. 6. Blocking buffer: 2% w/v BSA in PBS. Weigh 2 g of BSA and dissolve in 100 ml of PBS. Store at 4 C. 7. Dilution Buffer: 0.5% w/v BSA in PBS: Weigh 0.5 g of BSA and dissolve in 100 ml of PBS. Store at 4 C. 8. Mouse anti-V5 antibody. Aliquot 10-12 μl per tube at 1 mg/ ml. Store at À80 C. 9. HRP-conjugated goat anti-human antibody. 10. HRP-conjugated goat anti-mouse antibody. Aliquot 5–10 μl per tube. Store at À80 C. 11. o-Phenylenediamine dihydrochloride (OPD) tablets. 12. Development buffer: 0.05 M phosphate-citrate buffer adjusted to pH 5.0. Weigh 14.2 g of Na2HPO4, 4.8 g of C6H8O7 and dissolve in 1 l of H2O.

13. 30% v/v H2O2.

14. Stopping Solution: 5 N H2SO4. 15. Optical Plate Reader: Capable of reading samples at 490 nm (OD490).

2.5 Large-Scale 1. His GraviTrap column (GE Healthcare Life Sciences). Expression 2. 1 M Imidazole stock: Weigh 52.27 g of imidazole hydrochlo- ride and dissolve in 500 ml of H2O.

3. 10Â NaH2PO4/NaCl buffer: Weigh 12 g of NaH2PO4 and 146.1 g of NaCl and dissolve in H2O.

4. 10Â Na2HPO4/NaCl buffer: Weigh 14.2 g of Na2HPO4 and 146.1 g of NaCl and dissolve in H2O.

5. Washing buffer: 20 mM NaH2PO4,20mMNa2HPO4, 0.5 M NaCl with 40 mM imidazole. 50 ml each of 10Â NaH2PO4/ NaCl buffer and 10Â Na2HPO4/NaCl buffer mixed with 20 ml of 1 M Imidazole stock, top up with H2O, adjust pH to 7.4 and ﬁlter-sterilize. 6. Elution Buffer: 250 mM imidazole. Mix 125 ml of 1 M Imid- azole stock with 375 ml of washing buffer. 260 Zhihao Wu et al.

7. Spectrophotometer: Capable of reading samples at 280 nm (OD280). Always blank with the correct buffer to match your sample. 8. Slide-A-Lyzer Dialysis Cassettes 10 K MWCO.

9. Sodium Azide: 2% (w/v) in H2O. Weigh 2 g of NaN3 and dissolve in 100 ml of H2O.

2.6 Fc Fusion 1. pFUSE-hIgG1-Fc vector. Proteins Cloning 2. Expi293F™ Cells (Thermo Fisher Scientific). Store in liquid and Expression nitrogen. 3. ExpiFectamine™ 293 Transfection Kit (Thermo Fisher Scien- tific). Store at 4 C. 4. Expi293F™ Expression Medium (Thermo Fisher Scientific). Store at 4 C. ® 5. OptiMEM I reduced-serum medium (Thermo Fisher Scien- tific). Store at 4 C.

6. CO2 shaking incubator. 7. MabSelect protein A afﬁnity media. Store at 4 C.

3 Methods

3.1 Panning 1. Wash 500 μl of streptavidin paramagnetic Dynabeads 10 times with PBS-T, using a magnet to isolate the beads each time. Leave the beads in 500 μl of PBS-T after washing. Add 0.5 ml of 2% BSA in PBS to an Eppendorf tube and mix in 30 μl of the prewashed Dynabeads and 150 μg of biotinylated Control pMHC (see Note 1). Add about 1 Â 1012 pfu (100 μl) phage from the library to the mix and incubate for 1 hr. at 4 C with rotation (see Note 2). 2. After 1 h, apply a magnet to the sample and carefully transfer the supernatant to a new Eppendorf tube. Discard the beads. 3. Repeat step 2 with newly prepared beads. This initial “pre- screening” helps remove phages that bind to the beads or control pMHC. 4. Add 7.5 μg of the biotinylated Target pMHC to this pre-screened supernatant and incubate for 1 h at 4 C with rotation. After 1 h add 200 μl of prewashed Dynabeads to the supernatant and incubate at 4 C for another 15 min with rotation. 5. After incubation, apply a magnet to the sample, discard the supernatant from the tube, but keep the beads. Be careful not to disturb the Dynabeads while aspirating. The beads now Targeting Intracellular Antigens with pMHC-Binding Antibodies 261

contain the phages that bind the Target pMHC, but do not bind the Control pMHC. 6. Wash the beads 10 times with PBS-T and 5 times with PBS to remove nonspeciﬁc or poorly bound phages and transfer the beads to a new Eppendorf tube (see Note 3). During each wash, add 1 ml of buffer and incubate for 1–2 min before adding the magnet. After completing the ﬁnal wash, elute the bound phages by incubating the beads with 500 μl of elution buffer (Glycine-HCl, pH 2.2) for 10 min at RT. After elution, transfer the eluent to a new Eppendorf tube and neutralize with 30 μl of neutralization buffer (Tris–HCl, pH 9.0). This eluent contains the Target pMHC-binding phages, but at a relatively low titer. Therefore, it is necessary to amplify the phage for further downstream applications.

3.2 Amplification 1. Add the phages that have been affinity selected during the panning step to 20 ml of LB-Tet with XL1-blue E. coli (OD600 ¼ 1) and incubate in a shaker for 1 h at 37 C. This is when the infection of bacteria by phages occurs. Next, pellet the cells by spinning at 3000 Â g for 10 min. Discard the supernatant and resuspend cells with 0.5 ml LB, then spread the cells on a 15 cm LB-GAT plate. Incubate overnight (O/N) at 37 C. 2. Add 10 ml of LB to the LB-GAT plate and carefully scrape the cells from the plate (see Note 4). Collect the cells in a 15 ml tube. Calculate the OD600 and resuspend the cells in 17% glycerol/LB to a final density of 50.0 OD600. Freeze down 1 ml of these bacteria as a stock. Using these bacteria, inoculate 20 ml of LB-GAT to a final OD600 of 0.1 (for example use 40 μl of the 50.0 OD600 stock) and incubate at 37 Cina shaking incubator (>200 rpm) for 1 h. After this incubation, add 1 Â 1011 pfu M13 K07 helper phage (see Note 5), mix well and incubate for another 60 min at 37 C in a shaking incubator. After this incubation, pellet the cells by spinning them at 3000 Â g for 10 min and resuspend the cells in 20 ml of LB-GAT and add 20 μl of Kanamycin. Incubate the cells overnight at 30 C in a shaking incubator. 3. The following day, pellet the bacteria by spinning at 3000 Â g for 15 min, and transfer the supernatant (~20 ml) to a new tube and add in 5 ml of 5Â PEG/NaCl. Keep the tube on ice or at 4 C for 1 h to precipitate the amplified phage. During this incubation, periodically shake the tube. At this time, begin a culture of XL1-Blue (~5 ml per sample). After 1 h, pellet the precipitated phages by spinning at 6000 Â g for 30 min. Care- fully decant the supernatant, taking care not to disturb the phage pellet (see Note 6). 262 Zhihao Wu et al.

4. Carefully resuspend the phages in 1 ml of PBS (see Note 7). Be sure to wash the sides of the tube to remove the phage smear. Spin the 1 ml of phage at 12,000 Â g at 4 C for 5 min in a microcentrifuge to pellet any insoluble particles or remnants of E. coli. Titer the phages by inoculating 50 μl of OD600 0.5–1.0 À À XL1-Blue with 1 μlof108–10 12 dilutions of this phage stock. Wait 15 min after inoculation before plating the cells on 10 cm LB-GAT plates. Incubate the plates at 37 C O/N. 5. The next day, calculate the phage titer by multiplying the number of colonies on each plate by the dilution factor. This is the concentration of phages per microliter (see Note 8). Freeze down several aliquots of the ampliﬁed phages in 15% glycerol and store at À80 C, and use ~1 Â 1012 pfu phages to continue panning. 6. Repeat the above panning protocol 3–4 times, each time decreasing the amount of biotinylated target pMHC used. This will improve the speciﬁcity of the selected phages.

3.3 Clone Selection 1. After the final round of selection and amplification, titer the selected phage pool as before. After quantifying, freeze down several aliquots of the phage (as explained in Subheading 3.2, step 5) and inoculate 1 ml of an OD600 0.3–0.6 culture of HB2151 E. coli. Inoculate using a concentration of phages that gives between 100 and 200 colonies per 50 μlofculture.Plate 50 μl per plate on 3–5 LB-Amp plates and incubate O/N at 30 C. 2. The next day, pick colonies from the plates and inoculate them into a 96-well DeepWell™ plate with each well containing 400 μl LB-Amp and sealed it with Axygen™ microplate sealing film. Be sure to include at least one negative control well per experiment. This well should include the uninfected HB2151 in LB alone (see Note 9). Incubate the 96-well plate on a shaker at 37 C for 3–6 h. After the incubation, add 200 μl of 50% glycerol-LB per well. These now constitute the monoclonal glycerol stocks. Take 50 μl of each stock and inoculate a new 96-well plate with 400 μl of LB-Amp per well as before. Incu- bate until the OD600 reaches about 0.4 on over half of the plate (~3–6 h). Afterward, freeze down the first 96-well plate at À80 C for long-term storage. 3. Once the cells have reached an OD600 ~ 0.4, add 200 μlof LB-Amp þ 0.5 mM IPTG to induce soluble scFv production, and incubate O/N with shaking at 28 C. The next day, centrifuge the plates at 3000 Â g for 15 min and transfer the supernatant to a new plate for screening. This supernatant now contains soluble scFv from the monoclonal stocks. The next step will be to screen the clones for proper binding activity and specificity. Targeting Intracellular Antigens with pMHC-Binding Antibodies 263

3.4 ELISA Screening 1. For each 96-well plate, coat two ELISA plates with 50 μl/well for pMHC-Binding BSA-biotin at 10 μg/ml in PBS. Cover or seal the plate and Phage Clones incubate O/N at 4 C(see Note 10). 2. The next morning, wash the plates 3–5 times with PBS. After washing, add 50 μl/well of streptavidin at 10 μg/ml in PBS and incubate for 1 h at RT. 3. Wash the plates 5 times with PBS and coat the plates with 50 μl/well of either biotinylated Target pMHC or Control pMHC at 5 μg/ml in PBS. Incubate the plates for 1 h at RT (see Note 11). 4. Wash the plates 3–5 times with PBS and then add 150 μl/well of blocking buffer (2% BSA) to block the binding of proteins to the plates. Incubate the plates at RT for 60 min (see Note 10). 5. Wash the plate 3–5 times and add 100 μl of each monoclonal stock supernatant, or purified scFv/scFv-Fc diluted in the dilution buffer (0.5% BSA), to the plate. Incubate the plates at RT for 1 h. 6. Wash the plates five times and add 100 μl/well of mouse anti- V5 antibody at 0.5 μg/ml in dilution buffer if detecting soluble scFv. If detecting scFv-Fc (human Fc) fusion proteins, add 100 μl/well of goat anti-human HRP at 0.5 μg/ml in dilution buffer. Incubate the plates at RT for 1 h. 7. Wash the plates five times and add 100 μl/well of goat anti- mouse HRP at 0.5 μg/ml, if detecting by V5. Incubate the plates at 4 C for 1 h. If detecting by Fc skip this step and begin development of the plates. 8. During this incubation make the development buffer by adding two OPD tablets to 40 ml of OPD buffer. Keep at 4 C until ready to use. 9. Before developing the plates, wash them thoroughly (at least five times) with PBS. Right before use, add 40 μl of 30% H2O2 to the development buffer and mix well. Immediately add 150 μl/well of this development buffer and incubate at RT in the dark. Check the reaction every 5 min and stop the reaction after 30 min, when positive control wells turn dark yellow, or when negative control wells begin to turn light yellow, which- ever comes first. Stop the reaction by adding 30 μl of the stopping solution (5 N H2SO4). Add the acid quickly and carefully, being sure to not let too much time pass between the first wells and the final wells being stopped. Once all wells have been stopped, tap the sides of the plate to mix well. The acid is denser than the development solution and should mix on its own quite readily. 10. Read the plate on a spectrophotometer set to wavelength 490 nm. Wells with Target pMHC-binding scFvs should have 264 Zhihao Wu et al.

an OD490 above background (i.e., negative control wells) by at least 3 times the standard deviation of the background. Compare the binding against the Target pMHC to the Control pMHC to determine speciﬁcity (see Note 12).

3.5 Large-Scale 1. Identify the positions of the positive clones in the monoclonal Expression glycerol stocks from Subheading 3.3, step 2 and use them to inoculate 3 ml of OD600 0.6 HB2151 E. coli in LB and incubate at 30 C O/N in a shaking incubator (>200 rpm). Subculture the 1 ml culture to 0.2 l of LB-Amp and incubate at 37 C in a shaking incubator until OD600 reaches 0.6. 2. Add IPTG to a final concentration of 0.5 mM and incubate at 30 C O/N in a shaking incubator. The following day, pellet the cells at 2500 Â g for 15 min at 4 C. Resuspend pellet in 40 ml PBS and add polymyxin B to a final concentration of 1 μM. Shake at 37 C for 30 min. Polymyxin is an antibiotic that lyses the outer cell wall of bacteria and releases soluble scFv from the periplasmic space into solution. 3. Spin down and collect the supernatant and add Imidazole to a final concentration of 40 mM. 4. While incubating, prepare GraviTrap column by washing it with 10 ml of washing buffer. 5. Apply the supernatant with 40 mM imidazole onto the column. 6. Wash with 10 ml of washing buffer. 7. Elute with 3 ml of elution buffer with 250 mM imidazole. If the starting volume of culture is small or the expected yield is not good, a step-wise elution with 500 μl elution buffer can be used. 8. Using a 10K MWCO Slide-A-Lyzer cassette to dialyze the collected fractions to PBS O/N at 4 C. RT is acceptable if necessary (see Note 13). Measure the OD280 after dialysis to determine concentration and run the samples on SDS-PAGE or HPLC to determine purity. 9. Finally, aliquot the dialyzed protein and freeze it at À80 C. Minimize freeze/thaw cycles to prevent precipitation or degradation of the scFv.

3.6 Generating Fc 1. Starting from the monoclonal stocks, miniprep the clones of Fusion Proteins interest and sequence the scFv regions using sequencing primers for the vector used in the library. After determining the sequences, design primers with relevant 50 and 30 restriction sites (i.e. 50 EcoRI and 30 BglII) to PCR amplify the scFv sequences. Be sure to check the scFv sequence for these restriction sites; and if necessary use alternative restriction sites found Targeting Intracellular Antigens with pMHC-Binding Antibodies 265

on the pFUSE-hIgG1-Fc vector, or perform Gibson cloning [21]. Digest the PCR fragment and vector for 1 h at 37 C and gel purify it using a 1% agarose gel. Ligate for 30 min at RT using a 3:1 insert to vector molar ratio, and transform the plasmid into competent E. coli. Plate cells on LB-Amp plates and incubate O/N at 37 C. Pick 5–10 colonies and miniprep them. Screen by restriction digest to validate that the insert and vector bands match the approximate sizes of the pFUSE vector (4 kb) and scFv (~800 bp). Sequence the screened plasmids and select one with the correct sequence. If necessary, amplify the selected plasmid via midiprep or maxiprep (minimum 30 μg). 2. For transient transfection, begin culturing Expi293F cells according to the manufacturer’s instructions. Prepare cells one day before transfection with cell density of 2 Â 106 cells/ ml. On the day of transfection, count the number of cells and ensure viability is above 95%. ® 3. Dilute 30 μg plasmid DNA into 1.5 ml of OptiMEM I reduced-serum medium. Simultaneously mix 80 μl of ExpiFec- ® tamine™ with 1.5 ml of OptiMEM I reduced-serum. Incu- bate for no more than 5 min at RT and add the diluted DNA into the diluted transfection reagent. Incubate the mixture for another 20 min at RT. 4. While waiting, prepare 75 Â 106 Expi293F cells and top up the volume to 25.5 ml with warmed up fresh medium and place in a shaking incubator until use. 5. After 20 min incubation, slowly add the 3 ml transfection mixture to the prepared cells while swirling the flask. Incubate the cells for 4–6 days and harvest when viability drops below ~70% 6. On the day of harvest, spin down the supernatant at >3000 Â g for 1 h at 4 C. Store at 4 C until ready to begin purification. If one will be storing the supernatant for more than 2 days, it is best to add sodium azide (0.05% final) to prevent bacterial growth. Right before purification, pass the supernatant through a 0.22 μm filter to remove any precipitate or bacterial/fungal growth. 7. To purify the Fc-Fusion proteins, it is best to use protein-A beads (e.g., MabSelect). For larger volumes use an FPLC machine if possible. For smaller volumes, briefly mix protein- A resin with the filtered 293F supernatant and incubate O/N at 4 C with rotation. The next day load the resin onto a gravity column. Wash with 10 column volume (CV) of PBS before eluting with a pH gradient (i.e., from pH 7 to 2.5). Elute into 1.0 CV fractions, ten fractions per condition, and check OD280 for each fraction to identify the optimal elution conditions. Neutralize each fraction with 0.1 CV of 1 M Tris–HCl 266 Zhihao Wu et al.

(pH 9.0); combine and dialyze the fractions of interest in PBS or another suitable buffer (see Note 14). Aliquot and store samples at À80 C for long-term storage.

4 Notes

1. It is important to use the same MHC protein for control and target pMHCs to eliminate as much unwanted binding to MHC as possible. The selection of control peptides is critical in eliminating cross-reactivity to normal tissues and can be difficult in certain applications for at least two reasons. First, there is very limited information on the repertoire of “present- able” peptides in normal cells from different tissues. Second, recently it was found that around 25% of cell surface pMHC-I are derived from proteasome-catalyzed peptide splicing [22], adding further to the complexity of normal peptide repertoire. Under circumstances where a control peptide is not obvious, we generally recommend a bioinformatics approach to include a pool of homologous peptides as controls. Additionally, clones can be screened against cell lines that are known to express or not express the target peptide of interest. This is a very important step to confirm immunoreactivity. 2. Depending on the library used, it may be necessary to generate the phages from E. coli stocks. In this case, follow the manufacturer’s instructions, which should be similar to the steps from Subheading 3.2 (Amplification). 3. While PBS and PBS-T are the most common buffers for washing during the panning and screening steps, any buffer that is compatible with downstream in vitro or in vivo studies can be used, as long as it does not disturb antigen-antibody interactions. The method of washing can affect the stringency of selection. Longer incubations with wash buffer or using more wash steps is thought to select for higher affinity binders. Feel free to adjust these steps as necessary. 4. During this step, be careful not to break up the agar when scraping the plate. It is easiest to use a wide scraper to prevent this. The bacteria come off quite easily and it is not necessary to remove all traces of bacteria. If some agar does get into the mix, simply spin down the solution at a low speed (i.e., 100 g) briefly and transfer the supernatant to a fresh tube. 5. Helper phage provides the machinery necessary to package phagemid into mature virion for secretion. 6. Often the phage pellet forms only a faint smear on the side of the tube and is hard to identify. To make it easier to find, mark the bottom of the tube at the location where a pellet is expected Targeting Intracellular Antigens with pMHC-Binding Antibodies 267

to form before spinning. On an angled rotor, this is typically on the side of the tube facing away from the rotor. The smear is also apparent if the tube is held to the light and rotated. If no smear or pellet can be seen, simply wash the tube carefully and test for phage by using the tittering scheme explained in Sub- heading 3.2, step 4 but start at 101 and go up until 1010. 7. When using a library generated by animal immunization, it is important to know if the target protein was conjugated to a carrier protein or used with an adjuvant. To improve specificity, the phage pellet can be dissolved in PBS mixed with the adjuvant or protein conjugate. For example if the immunization used BSA as a carrier protein, resuspending the phage pellet in 1% BSA in PBS will negatively select the BSA-binding phages before panning, thus reducing the chance that these phages are selected and amplified. However, be sure not to use this same protein in the blocking or dilution buffers during the screening. 8. If at any stage it is clear that the selected phages are not amplifying or binding with enough specificity, simply go back to a frozen aliquot and continue panning/screening from there with less stringency. With each subsequent screening, the concentration of binding phages should increase steadily. To accurately quantify the concentration it is often necessary to dilute À the phage even further than 10 12. 9. It is recommended to pick as many colonies as possible to increase the chances of finding a good clone. Be aware, however, picking more than 95 clones per experiment means doing more than two ELISAs per experiment. Always include a negative control well but feel free to scale up screening as much as necessary. A positive control is only necessary if no binding is detectable after a primary screen. 10. The protein used during the panning steps (in this case BSA) should match the protein used in the blocking and dilution buffers in the screening ELISA; however, it does not need to be BSA. Be sure to use a protein that was not used in generating the phage library. BSA-biotin can be easily replaced with another biotinylated protein, if necessary. 11. For screening it is best to test the monoclonal stocks against both the Target pMHC and the Control pMHC to validate the specificity of the selected phage. However, for convenience one can screen a larger selection against the Target pMHC first, and then perform a secondary screen with high binders against the Control pMHC. 12. Note that at this stage it is impossible to definitively separate higher affinity binders from more stable sequences. Higher OD490 at this stage only means more scFvs were left bound 268 Zhihao Wu et al.

by the end of the ELISA, but it does not accurately distinguish between more efficient expression of the scFvs and more efficient binding to the targets. An internal control that detects the expression level of scFv can be implemented. 13. While this protocol uses a nickel purification method, this can easily be replaced with other methods such as protein-A (anti- Fc), protein-L (anti-kappa chain), or anti-V5 affinity chromatography. Each method has its benefits, but nickel purification is only used in this case for its convenience. Protein-L will sometimes have improved purity over nickel, but it does not bind all scFv sequences equally well. Protein-A can only bind to Fc-fusion or full IgG proteins and anti-V5 requires that the construct has a V5 tag. Similarly, all steps can be performed by FPLC, although small-scale purifications should be limited to gravity columns. 14. To fill the dialysis cassette, it is easiest to use a 22-G hypodermic needle fitted onto a small 1 ml syringe. Be sure not to overfill the cassette or let it sink into the dialysis buffer. When adding the sample to the cassette, keep the pointed end of the needle angled slightly downward with the cassette held parallel to the floor. This will help prevent any accidental puncturing of the membrane. The final volume after dialysis can sometimes change quite dramatically from the starting volume, so do not be alarmed if the volume appears to have dropped by up to 50%. 15. It is difficult to determine the optimal buffer formula for a given protein before enough of it can be successfully purified, but buffer optimization can substantially improve the stability of a given protein, during both short-term and long-term storage. Similarly, the buffers used during affinity chromatography can have enormous impact on the overall yield and purity of the final product. The buffers listed above should be considered a starting point but can and should be optimized for each construct produced.

References

1. Smith GP (1985) Filamentous fusion phage – 3. Griffiths AD et al (1994) Isolation of high novel expression vectors that display cloned affinity human antibodies directly from large antigens on the virion surface. Science synthetic repertoires. EMBO J 13:3245–3260 228:1315–1317. https://doi.org/10.1126/ 4. Balint RF, Larrick JW (1993) Antibody engi- Science.4001944 neering by parsimonious mutagenesis. Gene 2. Knappik A et al (2000) Fully synthetic human 137:109–118. https://doi.org/10.1016/ combinatorial antibody libraries (HuCAL) 0378-1119(93)90258-5 based on modular consensus frameworks and 5. Devlin JJ, Panganiban LC, Devlin PE (1990) CDRs randomized with trinucleotides1. J Mol Random peptide libraries: a source of specific Biol 296:57–86. https://doi.org/10.1006/ protein binding molecules. Science jmbi.1999.3444 249:404–406 Targeting Intracellular Antigens with pMHC-Binding Antibodies 269

6. Luzzago A, Felici F, Tramontano A, Pessi A, 15. Dao T et al (2013) Targeting the intracellular Cortese R (1993) Mimicking of discontinuous WT1 oncogene product with a therapeutic epitopes by phage-displayed peptides, human antibody. Sci Transl Med 5:176ra133. I. Epitope mapping of human H ferritin using https://doi.org/10.1126/scitranslmed. a phage library of constrained peptides. Gene 3005661 128:51–57. https://doi.org/10.1016/0378- 16. Chames P, Hufton SE, Coulie PG, Uchanska- 1119(93)90152-S Ziegler B, Hoogenboom HR (2000) Direct 7. McLafferty MA, Kent RB, Ladner RC, Mark- selection of a human antibody fragment land W (1993) M13 bacteriophage displaying directed against the tumor T-cell epitope disulfide-constrained microproteins. Gene HLA-A1-MAGE-A1 from a nonimmunized 128:29–36. https://doi.org/10.1016/0378- phage-Fab library. Proc Natl Acad Sci U S A 1119(93)90149-W 97:7969–7974 8. Cwirla SE, Peters EA, Barrett RW, Dower WJ 17. Liu H et al (2016) Targeting alpha-fetoprotein (1990) Peptides on phage: a vast library of (AFP)-MHC complex with CAR T cell therapy peptides for identifying ligands. Proc Natl for liver cancer. Clin Cancer Res. https://doi. Acad Sci U S A 87:6378–6382 org/10.1158/1078-0432.ccr-16-1203 9. McCafferty J, Griffiths AD, Winter G, Chiswell 18. Barbas CF III, Kang AS, Lerner RA, Benkovic DJ (1990) Phage antibodies: filamentous SJ (1991) Assembly of combinatorial antibody phage displaying antibody variable domains. libraries on phage surfaces: the gene III site. Nature 348:552–554 Proc Natl Acad Sci U S A 88:7978–7982 10. Hoogenboom HR et al (1991) Multi-subunit 19. Schoonbroodt S et al (2008) Engineering proteins on the surface of filamentous phage: Antibody Heavy Chain CDR3 to Create a methodologies for displaying antibody (Fab) Phage Display Fab Library Rich in Antibodies heavy and light chains. Nucleic Acids Res That Bind Charged Carbohydrates. J Immunol 19:4133–4137. https://doi.org/10.1093/ 181:6213–6221 nar/19.15.4133 20. Rauchenberger R et al (2003) Human combi- 11. Gram H et al (1992) In vitro selection and natorial Fab library yielding specific and func- affinity maturation of antibodies from a naive tional antibodies against the human fibroblast combinatorial immunoglobulin library. Proc growth factor receptor 3. J Biol Chem Natl Acad Sci U S A 89:3576–3580 278:38194–38205. https://doi.org/10. 12. Orum H et al (1993) Efficient method for 1074/jbc.M303164200 construction comprehensive murine Fab anti- 21. Gibson DG et al (2009) Enzymatic assembly of body libraries displayed on phage. Nucleic DNA molecules up to several hundred kilo- Acids Res 21:4491–4498. https://doi.org/ bases. Nat Methods 6:343–345. https://doi. 10.1093/nar/21.19.4491 org/10.1038/nmeth.1318 13. Hoogenboom HR, Winter G (1992) 22. Liepe J et al (2016) A large fraction of HLA By-passing immunisation. J Mol Biol class I ligands are proteasome-generated 227:381–388. https://doi.org/10.1016/ spliced peptides. Science 354:354–358. 0022-2836(92)90894-P https://doi.org/10.1126/science.aaf4384 14. Barbas CF III (1995) Synthetic human antibodies. Nat Med 1:837–839 Part II

Selection Strategies for Antibodies Chapter 14

Parallelized Antibody Selection in Microtiter Plates

Giulio Russo, Doris Meier, Saskia Helmsing, Esther Wenzel, Fabian Oberle, Andre´ Frenzel, and Michael Hust

Abstract

The most common in vitro technology to generate human antibodies is phage display. This technology is a key technology to select recombinant antibodies for the use as research tools, in diagnostic tests, and for the development of therapeutics. In this review, the high-throughput compatible selection of antibodies (scFv) in microtiter plates is described. The given detailed protocols allow the antibody selection (“panning”), screening and identiﬁca- tion of monoclonal antibodies in less than 1 week.

Key words Panning, Antibody selection, Phage display, Single-chain fragment variable (scFv), Anti- body, Monoclonal antibody screening

1 Introduction

Antibody phage display is a key technology to generate antibodies, mainly human antibodies, in vitro, independent of the restriction of the immune system. This in vitro procedure for the isolation of antibody fragments is called “panning” according to the gold washers [1]. In the panning procedure, the antigen can be immobilized to a solid surface, such as column matrixes [2], nitrocellulose [3], magnetic beads [4], or, most widely used, plastic surfaces with high protein binding capacity as polystyrene tubes, respectively microtiter wells (MTPs) [5]. A further strategy is to select antibodies in solution using biotinylated antigens followed by a “pull-down” step with streptavidin beads [6]. When generating antibodies against cell surface markers, e.g., cancer targets, the panning can be performed directly on cells [7, 8]. For the selection in MTPs, the antibody phage are incubated with the surface-bound antigen, followed by stringent washing to remove the vast excess of nonbinding antibody phage. Subsequently, the bound antibody phage will be eluted and reampliﬁed by infection of E. coli. The selection cycle will be repeated by infection of the phagemid

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_14, © Springer Science+Business Media LLC 2018 273 274 Giulio Russo et al.

bearing E. coli colonies derived from the former panning round with a helperphage to produce new antibody phage, which can be used for further panning rounds until a significant enrichment of antigen specific antibody phage is achieved. The number of antigen-specific antibody phage clones should increase with every panning round. Usually, 2–3 panning rounds are necessary to select specifically binding antibody fragments. For screening of monoclonal binders, scFvs are produced as soluble monoclonal antibody fragments, or in rare cases as monoclonal antibody phage, in microtiter plates. These monoclonal antibodies can be identified by, e.g., ELISA [9], immunoblot [5], or flow cytometer [10]. Subsequently, the gene fragments encoding the antibody fragments can be subcloned into any other antibody format, e.g., scFv-Fc or IgG [9, 11–13]. A schema of the selection procedure is given in Fig. 1. This selection procedure can be performed using patient- derived immune libraries [14, 15] or naive libraries like the McCaff- erty library [16], Pfizer library [17], Tomlinson libraries [18], or the Human/Hust antibody libraries (HAL) 7/8 and 9/10 [12, 19]. Antibody phage display libraries are valuable sources for the generation of antibodies against all kinds of target structures including therapeutic targets. Currently, six antibodies generated by phage display are FDA/EMA approved. An overview about phage display-derived therapeutic antibodies by phage display was given by Frenzel et al. [20]. The following protocols describe the panning and the screening of the selected antibody fragments completely in microtiter plates (MTPs). A “classic” protocol by plating the infected bacteria after elution during the panning can be found in an older publication [21]. The given protocol is high throughput compatible, because all steps are performed in MTPs allowing the selection of antibodies against 96 targets in parallel. The antibody selection can be performed in three days and the screening and identification of monoclonal antibodies in two further days.

2 Materials

2.1 Coating 1. Maxisorp microtiter plates or stripes (Nunc, Langenselbold, of Microtiter Wells Germany) or other polystyrole microtiter plates.

2. PBS pH 7.4: 8.0 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4∙2H2O, 0.24 g KH2PO4 in 1 L. 3. Dimethyl sulfoxide (DMSO). 4. PBST (PBS þ 0.05% (v/v) Tween 20).

2.2 Panning 1. MPBST: 2% skim milk in PBST, prepare fresh. MTP Panning 275

Fig. 1 Schema of antibody (scFv) phage display selection and screening (modiﬁed ﬁgure from former publications [20, 28])

2. Panning block solution: 1% (w/v) skim milk þ1% (w/v) BSA in PBST, prepare fresh. 3. 10 μg/mL Trypsin in PBS. 4. E. coli TG1 (supE thi-1 Δ(lac-proAB) Δ(mcrB-hsdSM)5 À À 0 q (rK mK )[F traD36 proAB lacI ZΔM15]). 5. M13K07 Helperphage (Thermo Fisher Scientiﬁc, Waltham, USA). 276 Giulio Russo et al.

6. Round-bottom polypropylene (PP) Deepwell 96 MTPs (Grei- ner, Frickenhausen, Germany). 7. Labnet VorTemp 56 benchtop shaker/incubator (Wood- bridge, NJ, USA). 8. Eppendorf 5810R, Rotor A-4-81 with MTP adapter. 9. 2ÂYT media pH 7.0: 1.6% (w/v) tryptone, 1% (w/v) yeast extract, 0.5% (w/v) NaCl. 10. 10ÂGA: 1 M glucose, 1 mg/mL ampicillin 11. 2ÂYT-GA: 2ÂYT, 100 mM glucose, 100 μg/mL ampicillin. 12. 2ÂYT-AK: 2ÂYT, containing 100 μg/mL ampicillin, 50 μg/ mL kanamycin. 13. Glycerol (99.5%).

2.3 Phage Titration 1. E. coli XL1-Blue MRF0 (recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac [F0 proAB lacIqZΔM15 Tn10 (Tetr)]). 2. 2ÂYT-GA agar plates (2ÂYT-GA þ 1.5% (w/v) agar-agar). 3. 2ÂYT-T: 2ÂYT, containing 50 μg/mL tetracycline.

2.4 Production 1. 96-well U-bottom polypropylene (PP) microtiter plates (Grei- of Soluble Monoclonal ner Bio-One, Frickenhausen, Germany). Antibody Fragments 2. AeraSeal breathable sealing ﬁlm (Excel Scientiﬁc, Victorville, in Microtiter Plates USA). 3. Potassium phosphate buffer pH 7.2–7.4: 2.31% (w/v) (0.17 M) KH2PO4 þ 12.54% (w/v) (0.72 M) K2HPO4. 4. Buffered 2ÂYT pH 7.0: 1.6% (w/v) tryptone, 1% (w/v) yeast extract, 0.5% (w/v) NaCl, 10% (v/v) potassium phosphat buffer. 5. Buffered 2ÂYT-SAI: buffered 2ÂYT containing 50 mM saccharose þ 100 μg/mL ampicillin þ 50 μM isopropyl-beta-D- thiogalactopyranoside (IPTG).

2.5 ELISA of Soluble 1. Mouse α-His-tag monoclonal antibody (α-Penta His, Qiagen, Monoclonal Antibody Hilden, Germany). Fragments 2. Mouse α-myc-tag monoclonal antibody (9E10) (Sigma, Munchen,€ Germany). 3. Mouse α-pIII monoclonal antibody PSKAN3 (Mobitec, Go¨t- tingen, Germany). 4. Goat α-Mouse IgG serum, (Fab speciﬁc) HRP conjugated (Sigma). 5. Oligonucleotide primers MHLacZ-Pro_f (50 GGCTCGTATG TTGTGTGG 30) and MHgIII:r (50 CTAAAGTTTTGTC GTCTTTCC 30). MTP Panning 277

3 Methods

The time schedule for the complete procedure from antibody selection to identiﬁcation of monoclonal antibodies is given in Table 1.

3.1 Coating 1. (a) Protein antigen: For the ﬁrst panning round, use 1–5 μg of Microtiter Plate protein/well per panning, for the following rounds use Wells 0.1–1 μg protein/well for more stringent conditions. Dissolve the antigen in 150 μL PBS (see Note 1) and incubate it in a polystyrole (PS) microtiter plate well (MTP) overnight at 4 C. (b) Oligopeptide antigen: Use 500–1000 ng oligopeptide for each panning round. Dissolve the oligopeptide in 150 μL PBS, transfer into a streptavidin-coated MTP well and incubate overnight at 4 C(see Note 2). 2. Wash the coated microtiter plate wells 3Â with PBST using an ELISA washer (see Note 3).

3.2 Panning 1. (a) Block the antigen-coated MTP wells with MPBST for 1–2 h at RT, or overnight at 4C. The wells have to be completely ﬁlled. Afterward, wash the blocked antigen-coated wells 3Â with PBST (see Note 13).

Table 1 Time schedule for antibody selection (panning) and screening of monoclonal antibodies

Day Procedure steps Preparation steps

0 – – Coating of MTP wells for ﬁrst panning round – Overnight culture of E. coli TG1 1 – First panning round – Coating of MTP wells for second panning – Infection of E. coli TG1 with eluted phage round – Infection with helperphage – Overnight culture of E. coli TG1 – Antibody phage production overnight 2 – Second panning round – Coating of MTP wells for third panning – Infection of E. coli TG1 with eluted phage round – Infection with helperphage – Overnight culture of E. coli XL1-Blue MRF0 – Antibody phage production overnight 3 – Third panning round – – Infection of E. coli XL1-Blue MRF’ with eluted phage – Titration on agar plates 4 – Picking clones for screening – – Culture overnight 5 – Production of soluble scFv overnight – Coating of MTP wells for screening ELISA 6 – Screening ELISA – 278 Giulio Russo et al.

(b) You need to perform this step only in the first panning round but we suggest performing this step also in following rounds! In parallel, block an additional MTP well (without antigen!) per panning with panning block solution for 1 h at RT, or overnight at 4C, for preincubation of the antibody gene library. The MTP wells have to be completely filled. When using biotinylated antigens, use a streptavidin MTP well (see Note 2). Wash 3Â times with PBST (see Note 3). Incubate 1011–1012 antibody phage (you should use ~50–100Â excess of phage particles compared to the library size) from the library in 50 μL panning block for 1 h at RT. This step removes unspecific binders which often occur from the antibody gene libraries due to incorrect folding of individual antibodies. 2. Carry over the preincubated antibody phage library to the blocked MTP wells or fill 1011–1012 amplified phage solved in panning block solution (final volume 150 μL) from the first or second panning round in the blocked MTP wells. Incubate at RT for 2 h for binding of the antibody phage. When using biotinylated antigens add 5 μg streptavidin for competition per MTP well. 3. Remove the unspecifically bound antibody phage by stringent washing. Therefore, wash the wells 10Â with an ELISA washer in the first panning round. In the following panning rounds increase the number of washing steps (20Â in the second panning round, 30Â in the third panning round, etc.) (see Note 3). 4. Elute bound antibody phage with 150 μL Trypsin solution (10 μg/mL) for 30 min at 37 C(see Note 4). 5. After the third panning round, use 10 μL of the eluted phage for titration (see titering). 6. Inoculate 50 mL 2ÂYT with an overnight culture of E. coli TG1 (see Note 5) in 100 mL Erlenmeyer flasks and grow at 250 rpm and 37 C to O.D.600 0.4–0.5 (see Note 6). 7. Fill 150 μL exponentially growing E. coli TG1 in a polypropylene (PP) Deepwell MTP well and mix with 150 μL of the eluted phage. Incubate the bacteria for 30 min at 37 C without shaking and 30 min at 37 C and 650 rpm (see Note 7). 8. Add 1000 μLof2ÂYT and 150 μL10Â GA (see Note 8) and incubate for 1 h at 37 C and 650 rpm. O.D. should be ~0.5 (~5 Â 108 cells/mL). 9. Infect the bacteria with 50 μL M13K07 helpherphage (2 Â 1011 phage particles/mL ¼ 1 Â 1010 phage particles, MOI 1:20). Incubate for 30 min at 37 C without shaking, followed by 30 min at 37 C at 650 rpm. 10. Centrifuge the MTP plate at 3220 Â g (e.g., use Eppendorf 5810R, Rotor A-4-81 with MTP carriers). Remove the complete supernatant with a pipette. Do not destroy the pellet (see Note 9). MTP Panning 279

11. Add 950 μL2ÂYT-AK and incubation overnight at 30 C and 850 rpm to produce new antibody phage. 12. Centrifuge the MTP plate at 3220 Â g. Transfer the supernatant (~1 Â 1012 scFv-phage/mL) into a new PP MTP. The supernatant can directly be used for the next panning round.

3.3 Phage Titration 1. Inoculate 30 mL 2ÂYT-T in a 100 mL Erlenmeyer flask with E. coli XL1-Blue MRF’ (see Note 10) and grow overnight at 37 C and 250 rpm. 2. Inoculate 50 mL 2ÂYT-T with 500 μL overnight culture and grow at 250 rpm at 37 C up to O.D.600 ~ 0.5 (see Note 6). 3. Make serial dilutions of the phage suspension in PBS. The number of eluted phage depends on several parameters (e.g., antigen, library, panning round, washing stringency, etc.). In case of a successful enrichment, the titer of eluted phage usually is 103–105 phage per well after the first panning round and increases two to three orders in magnitude per additional panning round (see Note 11). The phage titer after reamplification should be 1012–1013 phage/mL. 4. Infect 50 μL bacteria with 10 μL phage dilution and incubate 30 min at 37 C. 5. You can perform titrations in two different ways: (a) Plate the 60 μL infected bacteria on 2xYT-GA agar plates (9 cm petri dishes). (b) Pipet 10 μL (in triplicate) on 2ÂYT-GA agar plates. Here, about 20 titering spots can be placed on one 9 cm petri dish. Dry drops on work bench. 6. Incubate the plates overnight at 37 C. 7. Count the colonies and calculate the cfu or cfu/mL titer according to the dilution.

3.4 Production 1. Fill each well of a 96-well U-bottom PP MTP with 150 μL of Soluble Monoclonal 2ÂYT-GA. Antibody Fragments 2. Pick 92 clones with sterile tips from the third panning round in Microtiter Plates and inoculate each well (see Note 12). Seal the plate with a breathable sealing ﬁlm. 3. Incubate overnight in a microtiter plate shaker at 37 C and 850 rpm. 4. (a) Fill a new 96-well polypropylene microtiter plate with 180 μL2ÂYT-GA and add 10 μL of the overnight cultures. Incubate for 2 h at 37 C and 850 rpm. (b) Add 30 μL glycerol solution to the remaining 140 μL overnight cultures. Mix by pipetting and store this masterplate at À80 C. 5. Pellet the bacteria in the microtiter plates by centrifugation for 10 min at 3200 Â g. Remove 180 μL glucose containing media by carefully pipetting (see Note 9). 280 Giulio Russo et al.

6. Add 180 μL buffered 2ÂYT-SAI (containing saccharose, ampicillin and 50 μM IPTG) and incubate overnight at 30 C and 850 rpm (see Notes 13 and 14). 7. Pellet the bacteria by centrifugation for 10 min at 3200 Â g in the microtiter plates. Transfer the antibody fragment containing supernatant to a new polypropylene microtiter plate and store at 4 C for a few days or directly proceed with testing the antibody binding.

3.5 ELISA of Soluble 1. To analyze the antigen specificity of the monoclonal soluble Monoclonal Antibody antibody fragments, coat 100–200 ng antigen per well over- Fragments night at 4 C. As control coat 100–200 ng BSA or streptavidin per well (for coating see Subheading 3.2, step 1). 2. Wash the coated microtiter plate wells 3Â with PBST (washing procedure see Subheading 3.2, step 1 and Note 3). 3. Block the antigen-coated wells with MPBST for 2 h at RT. The wells have to be completely filled. 4. Fill 50 μL MPBST in each well and add 50 μL of antibody solution (see Subheading 3.2, step 4). Incubate for 1.5 h at RT (or overnight at 4 C). 5. Wash the microtiter plate wells 3Â with PBST (washing procedure see Subheading 3.2, step 1 and Note 3). 6. Incubate 100 μL/well mouse α-myc tag antibody (clone 9E10) solution for 1.5 h (appropriate dilution in MPBST). 7. Wash the microtiter plate wells 3Â with PBST (washing procedure see Subheading 3.2, step 1 and Note 3). 8. Incubate 100 μL/well goat α-mouse HRP conjugate for 1 h (appropriate dilution in MPBST). 9. Wash the microtiter plate wells 3Â with PBST (washing procedure see Subheading 3.2, step 1 and Note 13). 10. Shortly before use, mix 19 parts TMB substrate solution A and 1 part TMB substrate solution B. Add 100 μL of this TMB solution into each well and incubate for 1–30 min. 11. Stop the color reaction by adding 100 μL 1 N sulfuric acid solution per well. The color turns from blue to yellow. 12. Measure the extinction at 450 nm using an ELISA reader (reference wavelength 620 nm). 13. Identify positive candidates with a signal (on antigen) 10Â over noise (on control protein, e.g., BSA) (see Note 15). 14. Sequence the DNA of the selected scFv for identification of unique clones using the oligonucleotide primers MHLacZ- Pro_f and MHgIII_r. We suggest analyzing the antibody sequences using VBASE2 (www.vbase2.org) (Tool: Fab/ scFab/scAb/scFv Analysis). MTP Panning 281

4 Notes

1. If the protein is not binding properly to the microtiter plate surface, try bicarbonate buffer (50 mM NaHCO3, pH 9.6) (this buffer is recommended by Nunc for Maxisorp plates). 2. More hydrophobic oligopeptides may need to be dissolved in PBS containing 5–100% DMSO. If biotinylated oligopeptides are used as antigen for panning, dissolve 200 ng streptavidin in 150 μL PBS and coat overnight at 4 C. Coat two wells for each panning, one well is for the panning, the second one for the preincubation of the library to remove streptavidin or unspecific binders! Pour out the wells and wash 3Â with PBST. Dissolve 100–500 ng biotinylated oligopeptide in PBS and incubate for 1 h at RT. Alternatively, oligopeptides with a terminal cystein residue can be coupled to BSA and coated overnight at 4 C. When working with biotinylated oligopeptides, it is recommended to use 2% BSA in PBST solution instead of 2% MPBST. Soluble streptavidin (1–5 μg) should be added into the library well at least in the first panning round to further avoid streptavidin binders. 3. The washing should be performed with an ELISA washer (e.g., TECAN Columbus Plus) to increase the stringency and reproducibility. To remove antigen or blocking solutions wash 3Â with PBST (“standard washing protocol” for TECAN washer). If no ELISA washer is available, wash manually 3Â with PBST. After binding of antibody phage, wash 10Â with PBST (“stringent bottom washing protocol” in case of TECAN washer). If no ELISA washer is available, wash manually 10Â with PBST and 10Â with PBS. For stringent off-rate selection increase the number of washing steps or additionally incubate the microtiter plate in 1 L PBS for several days. 4. Phagemids like pHAL14 [12, 22] or pHAL30 [19] have coding sequences for a trypsin-specific cleavage site between the antibody fragment gene and the gIII. Trypsin also cleaves within antibody fragments but does not degrade the phage particles including the pIII that mediates the binding of the phage to the F-pili of E. coli required for the infection. We observed that proteolytic cleavage of the antibody fragments from the antibody::pIII fusion by trypsin increases not only the elution but also enhances the infection rate of eluted phage particles, especially when using Hyperphage as helperphage. 5. E. coli TG1 is growing much faster compared to XL1-Blue MRF’ and allows to perform one panning round per day.

6. If the bacteria have reached O.D.600 ~ 0.5 before they are needed, store the culture immediately on ice to maintain the F-pili on the E. coli cells for several hours. M13K07 282 Giulio Russo et al.

helperphage (kan+) or other scFv-phage (amp+) can be used as positive control to check the infectibility of the E. coli cells.

7. After 1 h of incubation an O.D.600 0.4–0.5 is reached, corresponding to ~5 Â 108 bacteria. 8. The high concentration of glucose is necessary to efﬁciently repress the Lac promoter controlling the antibody::pIII fusion gene on the phagemid. Low glucose concentrations lead to an inefﬁcient repression of the lac promoter and background expression of the antibody::pIII fusion protein. Background antibody expression is a strong selection pressure frequently causing mutations in the phagemid, especially in the promoter region and the antibody::pIII fusion gene. Bacteria with this kind of mutations in the phagemids proliferate faster than bacteria with non-mutated phagemids. Therefore, the 100 mM glucose has to be included in every step of E. coli cultivation except during the phage production! 9. To not destroy the pellet, remove the supernatant carefully by touching the pipette tip at the side of the well and aspirate slowly. An alternative is to manually shake out the supernatant (do it with a fast movement of your wrist). 10. Use E. coli XL1-Blue MRF0 for titering and production of soluble antibodies. The plasmid quality and yield using this strain is better compared to TG1. Furthermore, the XL1-Blue MRF’ slower growth rate and the more regular colony shape compared to the TG1, allow for a more accurate picking of single colonies for screening. 11. When the antibody gene library was packaged using Hyperph- age, the titer of the eluted phage after the second panning may not increase as strongly or even decreases slightly due to the change from oligovalent to monovalent display. 12. We recommend picking 92 clones when using a 96-well microtiter plate. Use the wells H3, H6, H9, and H12 for controls. H3 and H6 are negative controls (these wells will not be inoculated, but used as negative control for the following ELISA with soluble antibodies). We inoculate the wells H9 and H12 with a clone containing a phagemid encoding a known antibody fragment. In ELISA, the wells H9 and H12 are coated with the antigen corresponding to the control antibody fragment in order to check scFv production and ELISA. 13. The appropriate IPTG concentration for induction of antibody or antibody::pIII expression depends on the vector design. A concentration of 50 μM was well suited for vectors with a Lac promoter like pIT2 [23], pHENIX [24], pHAL14 [12, 22], or pHAL30 [19]. 14. Buffered culture media and the addition of saccharose enhances the production of many but not all scFvs [25]. We MTP Panning 283

observed that antibody::pIII fusion proteins and antibody phage sometimes show differences in antigen binding in comparison to soluble antibody fragments, because some antibodies can bind the corresponding antigen only as pIII fusion [26, 27]. Therefore, we recommend performing the screening procedure only by using soluble antibody fragment, to avoid false positive binders. On the other hand, some scFv binding as antibody phage, but not as soluble scFv, bind as scFv-Fc after recloning.

15. The background (noise) signals should be about O.D.450/ 620 ~ 0.02 after 1–30 min TMB incubation time.

Acknowledgments

This review is an updated and revised version of [9].

References

1. Parmley SF, Smith GP (1988) Antibody- 7. Keller T, Kalt R, Raab I, Schachner H, May- selectable filamentous fd phage vectors: affinity rhofer C, Kerjaschki D, Hantusch B (2015) purification of target genes. Gene 73:305–318 Selection of scFv antibody fragments binding 2. Breitling F, Dubel€ S, Seehaus T, to human blood versus lymphatic endothelial Klewinghaus I, Little M (1991) A surface surface antigens by direct cell phage display. expression vector for antibody screening. PLoS One 10:e0127169. https://doi.org/10. Gene 104:147–153 1371/journal.pone.0127169 3. Hawlisch H, Muller€ M, Frank R, Bautsch W, 8. Rezaei J, RajabiBazl M, Ebrahimizadeh W, Klos A, Ko¨hl J (2001) Site-specific anti-C3a Dehbidi GR, Hosseini H (2016) Selection of receptor single-chain antibodies selected by single chain antibody fragments for targeting differential panning on cellulose sheets. Anal prostate specific membrane antigen: a compar- Biochem 293:142–145 ison between cell-based and antigen-based 4. Moghaddam A, Borgen T, Stacy J, approach. Protein Pept Lett 23:336–342 Kausmally L, Simonsen B, Marvik OJ, Brekke 9. Frenzel A, Kugler€ J, Wilke S, Schirrmann T, OH, Braunagel M (2003) Identification of Hust M (2014) Construction of human anti- scFv antibody fragments that specifically recog- body gene libraries and selection of antibodies nise the heroin metabolite by phage display. Methods Mol Biol 6-monoacetylmorphine but not morphine. J 1060:215–243. https://doi.org/10.1007/ Immunol Methods 280:139–155 978-1-62703-586-6_12 5. Hust M, Maiss E, Jacobsen H-J, Reinard T 10. Ayriss J, Woods T, Bradbury A, Pavlik P (2007) (2002) The production of a genus-specific High-throughput screening of single-chain recombinant antibody (scFv) using a recombi- antibodies using multiplexed flow cytometry. J nant potyvirus protease. J Virol Methods Proteome Res 6:1072–1082. https://doi.org/ 106:225–233 10.1021/pr0604108 6. Schutte€ M, Thullier P, Pelat T, Wezler X, 11. Hoet RM, Cohen EH, Kent RB, Rookey K, Rosenstock P, Hinz D, Kirsch MI, Schoonbroodt S, Hogan S, Rem L, Frans N, Hasenberg M, Frank R, Schirrmann T, Daukandt M, Pieters H, van Hegelsom R, Neer Gunzer M, Hust M, Dubel€ S (2009) Identifi- NC, Nastri HG, Rondon IJ, Leeds JA, Hufton cation of a putative Crf splice variant and gen- SE, Huang L, Kashin I, Devlin M, Kuang G, eration of recombinant antibodies for the Steukers M, Viswanathan M, Nixon AE, Sex- specific detection of Aspergillus fumigatus. ton DJ, Hoogenboom HR, Ladner RC (2005) PLoS One 4:e6625. https://doi.org/10. Generation of high-affinity human antibodies 1371/journal.pone.0006625 by combining donor-derived and synthetic 284 Giulio Russo et al.

complementarity-determining-region diversity. Generation and analysis of the improved human Nat Biotechnol 23:344–348 HAL9/10 antibody phage display libraries. BMC 12. Hust M, Meyer T, Voedisch B, Rulker€ T, Biotechnol 15:10. https://doi.org/10.1186/ Thie H, El-Ghezal A, Kirsch MI, Schutte€ M, s12896-015-0125-0 Helmsing S, Meier D, Schirrmann T, Dubel€ S 20. Frenzel A, Schirrmann T, Hust M (2016) (2011) A human scFv antibody generation Phage display-derived human antibodies in pipeline for proteome research. J Biotechnol clinical development and therapy. MAbs 152:159–170. https://doi.org/10.1016/j. 8:1177–1194. https://doi.org/10.1080/ jbiotec.2010.09.945 19420862.2016.1212149 13. J€ager V, Bussow€ K, Wagner A, Weber S, 21. Schirrmann T, Hust M (2010) Construction of Hust M, Frenzel A, Schirrmann T (2013) human antibody gene libraries and selection of High level transient production of recombi- antibodies by phage display. Methods Mol Biol nant antibodies and antibody fusion proteins 651:177–209. https://doi.org/10.1007/ in HEK293 cells. BMC Biotechnol 13:52. 978-1-60761-786-0_11 https://doi.org/10.1186/1472-6750-13-52 22. Kirsch M, Hulseweh€ B, Nacke C, Rulker€ T, 14. Trott M, Weiβ S, Antoni S, Koch J, von Schirrmann T, Marschall H-J, Hust M, Dubel€ Briesen H, Hust M, Dietrich U (2014) Func- S (2008) Development of human antibody tional characterization of two scFv-Fc antibo- fragments using antibody phage display for dies from an HIV controller selected on soluble the detection and diagnosis of Venezuelan HIV-1 Env complexes: a neutralizing V3- and a equine encephalitis virus (VEEV). BMC Bio- trimer-specific gp41 antibody. PLoS One 9: technol 8:66 e97478. https://doi.org/10.1371/journal. 23. Goletz S, Christensen PA, Kristensen P, pone.0097478 Blohm D, Tomlinson I, Winter G, Karsten U 15. Chan SW, Bye JM, Jackson P, Allain JP (1996) (2002) Selection of large diversities of antiidio- Human recombinant antibodies specific for typic antibody fragments by phage display. J hepatitis C virus core and envelope E2 peptides Mol Biol 315:1087–1097 from an immune phage display library. J Gen 24. Finnern R, Pedrollo E, Fisch I, Wieslander J, Virol 77(Pt 10):2531–2539 Marks JD, Lockwood CM, Ouwehand WH 16. Schofield DJ, Pope AR, Clementel V, Buckell J, (1997) Human autoimmune anti-proteinase Chapple SD, Clarke KF, Conquer JS, Crofts 3 scFv from a phage display library. Clin Exp AM, Crowther SRE, Dyson MR, Flack G, Grif- Immunol 107:269–281 fin GJ, Hooks Y, Howat WJ, Kolb-Kokocinski- 25. Hust M, Steinwand M, Al-Halabi L, A, Kunze S, Martin CD, Maslen GL, Mitchell Helmsing S, Schirrmann T, Dubel€ S (2009) JN, O’Sullivan M, Perera RL, Roake W, Shad- Improved microtitre plate production of single bolt SP, Vincent KJ, Warford A, Wilson WE, chain Fv fragments in Escherichia coli. New Xie J, Young JL, McCafferty J (2007) Applica- Biotechnol 25:424–428. https://doi.org/10. tion of phage display to high throughput anti- 1016/j.nbt.2009.03.004 body generation and characterization. Genome 26. Goffinet M, Chinestra P, Lajoie-Mazenc I, Biol 8:R254. https://doi.org/10.1186/gb- Medale-Giamarchi C, Favre G, Faye J-C 2007-8-11-r254 (2008) Identification of a GTP-bound Rho 17. Glanville J, Zhai W, Berka J, Telman D, specific scFv molecular sensor by phage display Huerta G, Mehta GR, Ni I, Mei L, Sundar selection. BMC Biotechnol 8:34 PD, Day GMR, Cox D, Rajpal A, Pons J 27. Lillo AM, Ayriss JE, Shou Y, Graves SW, Brad- (2009) Precise determination of the diversity bury ARM (2011) Development of phage- of a combinatorial antibody library gives based single chain Fv antibody reagents for insight into the human immunoglobulin reper- detection of Yersinia pestis. PLoS One 6: toire. Proc Natl Acad Sci U S A e27756. https://doi.org/10.1371/journal. 106:20216–20221. https://doi.org/10. pone.0027756 1073/pnas.0909775106 28. Kuhn P, Fuhner€ V, Unkauf T, Moreira GMSG, 18. de Wildt RM, Mundy CR, Gorick BD, Tom- Frenzel A, Miethe S, Hust M (2016) Recom- linson IM (2000) Antibody arrays for high- binant antibodies for diagnostics and therapy throughput screening of antibody-antigen against pathogens and toxins generated by interactions. Nat Biotechnol 18:989–994. phage display. Proteomics Clin Appl https://doi.org/10.1038/79494 10:922–948. https://doi.org/10.1002/prca. 19. KuglerJ,WilkeS,MeierD,TomszakF,€ 201600002 Frenzel A, Schirrmann T, Dubel€ S, Garritsen H, Hock B, Toleikis L, Schutte€ M, Hust M (2015) Chapter 15

Mass Spectrometry Immuno Assay (MSIA™) Streptavidin Disposable Automation Research Tips (D.A.R.T’s®) Antibody Phage Display Biopanning

Chai Fung Chin, Yee Siew Choong, and Theam Soon Lim

Abstract

Antibody phage display has been widely established as the method of choice to generate monoclonal antibodies with various efﬁcacies post hybridoma technology. This technique is a popular method which takes precedence over ease of methodology, time- and cost-savings with comparable outcomes to conventional methods. Phage display technology manipulates the genome of M13 bacteriophage to display large diverse collection of antibodies that is capable of binding to various targets (nucleic acids, peptides, proteins, and carbohydrates). This subsequently leads to the discovery of target-related antibody binders. There have been several different approaches adapted for antibody phage display over the years. This chapter focuses on the semi-automated phage display antibody biopanning method utilizing the MSIA™ ® streptavidin D.A.R.T’s system. The system employs the use of electronic multichannel pipettes with predeﬁned programs to carry out the panning process. The method should also be adaptable to larger liquid handling instrumentations for higher throughput.

® Key words Biopanning, Disposable automation research tips (D.A.R.T’s ), Mass spectrometry immunoassay (MSIA™), Monoclonal antibodies, Phage display

1 Introduction

George P. Smith first introduced the concept of phage display in 1985 and applied it for the display of peptides [1, 2]. Since then, the application of phage display technology has extended to the production of antibodies, enzyme evolution, and even nanotechnology. Antibody phage display is a powerful tool used to select for monoclonal antibodies with specific binding properties against a specific target antigen [3]. Its fundamental concept relies on the physical linkage of the phenotype to the genotype, established by the connection between the antibody fragment displayed on the phage surface (phenotype) and the genetic information encoding the displayed protein (genotype) encapsulated in the phage [4, 5]. The starting point of antibody phage display is the

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_15, © Springer Science+Business Media LLC 2018 285 286 Chai Fung Chin et al.

preparation of the antibody library, followed by repeated rounds of biopanning for the selection and enrichment of the specialized antibody population that is specific to the target antigen [5]. Numerous biopanning methods based on phage display technology were introduced by employing different solid phase surfaces such as microtiter plates, magnetic beads, or immunotubes [6] for target immobilization. The immobilization of the target antigen allows the physical separation of binding phages from non-binders, thus playing an important role in biopanning [7]. Traditionally, antigen immobilization is carried out by passive adsorption to polystyrene plate surfaces or by affinity capture of biotinylated antigens on streptavidin-coated plates [7, 8]. The use of magnetic beads is another option for target immobilization with the added advantage of an increased surface area for binding and an efficient washing step by physical isolation during biopanning [7]. The MSIA™ system was developed as an approach for protein analysis, involving immunoaffinity capture of the target antigens in microscale for mass spectrometry analysis. The MSIA™ Streptavi- ® din Disposable Automation Research Tips (D.A.R.T’s ) was developed to isolate proteins for mass spectrometry analysis through affinity selection of biotinylated antigens captured by the streptavidin on the tip matrix [9]. In order to capture the target protein, a small porous matrix is fitted in a pipette tip for antibody immobili- ® zation. In the context of the MSIA™ Streptavidin D.A.R.T’s , repeated aspiration and dispensing of biotinylated antigens through the tip allows for the immobilization of biotinylated target proteins to the tip matrix by streptavidin-biotin interaction [10]. ® The MSIA™ Streptavidin D.A.R.T’s are pipette tips fitted with a small piece of porous monolithic packing material that contains covalently bound streptavidin [11]. Thus, the porous medium in the tips will help in immobilizing the biotinylated antigen effectively due to the high affinity between streptavidin and biotin [11, 12]. The immobilized antigens will be used to capture antibodies that are highly specific to the targets. In addition, an electronic multi-channel pipette with various aspirations and dispensing speed settings, attached to an adjustable pipette stand ® is used complementarily with the streptavidin D.A.R.T’s in this biopanning method. Consequently, it helps to create a semi- automated biopanning system that has a small platform, convenient to use and time-saving. Similar to the conventional panning methods, this newly developed MSIA™ streptavidin tips-based biopanning consists of the common steps in phage display biopanning with slight modifications. It initially starts with immobilizing the antigen on the porous ® matrix in the D.A.R.T’s as illustrated in Fig. 1, followed by the washing and blocking steps. These steps are essential in minimizing nonspecific binding [13]. By repeatedly aspirating and dispensing ® the antibody phage library in the antigen-coupled D.A.R.T’s , MSIA Streptavidin DARTs Panning 287

Fig. 1 Immobilization of target antigen in D.A.R.T’s® with (a) aspiration and (b) dispensing of the antigen- containing buffer with the streptavidin matrix

specific antibodies will bind to the target antigen. Once the antibodies are bound, another round of wash step is carried out to remove nonspecific antibodies or low affinity-antibodies that bind loosely to the antigen [10]. Finally, an elution buffer is used to elute the bound phages, followed by amplification of the enriched phage pool which will be used in the subsequent rounds of biopanning. This biopanning approach is illustrated in Fig. 2. A similar principle ® is applied to the D.A.R.T’s for a modified use in antibody phage display biopanning [11]. The use of mass spectrometry immunoassay (MSIA™) system in biopanning is an attractive alternative to the conventional antibody phage display biopanning to produce antibodies that bind specifically to the target antigens. The equipment used in this biopanning method is inexpensive and practical compared with other semi-automated methods that are currently available [11]. In the following section, we will introduce the materials and several protocols to perform or validate results of the MSIA™ ® Streptavidin D.A.R.T’s antibody biopanning. In addition, methods in producing and validating the soluble monoclonal antibodies will be provided as these are important steps in phage display technology to generate soluble antibodies for downstream applications. The biopanning approach discussed in this chapter is used by the authors to generate monoclonal antibodies against other targets, even for epitope peptides. 288 Chai Fung Chin et al.

Fig. 2 Process of MSIA™ Streptavidin D.A.R.T’s® antibody biopanning approach

2 Materials

2.1 Preparation 1. An in-house synthetic domain antibody phagemid library [14] of Phage Display was used (see Note 1). Antibody Library 2. The vector used for the antibody library is the pLABEL phagemid which employs pIII minor coat protein for antibody 2.1.1 Phage Display display. Antibody Libraries 3. The library size for the synthetic domain antibody library is and E. coli Host Strains 6.6 Â 109 CFU/mL. 4. E. coli XL1Blue: tetracycline resistant; endonuclease (endA) deﬁcient which greatly improves the quality of miniprep DNA; recombination (recA) deﬁcient which improves insert stability; hsdR mutation prevents the cleavage of cloned DNA by the EcoK endonuclease system; lacIqZΔM15 gene on the F0 episome which allows for blue-white color screening.

2.1.2 Preparation 1. Sterile conical centrifuge tubes: 15 and 50 mL (Nunc). of Antibody Library 2. 1.5 mL sterile microcentrifuge tubes (Eppendorf). MSIA Streptavidin DARTs Panning 289

3. Erlenmeyer flasks (1 and 2 L) (Schott Duran). 4. Sterile petri dishes 94 mm Â 16 mm. 5. 2 YT: Prepare 16 g tryptone, 10 g yeast extract, and 5 g NaCl in 1LdH2O, autoclave and store at room temperature. 6. 50 mg/mL ampicillin stock solution: Prepare 0.5 g ampicillin in 10 mL of 50% (v/v) ethanol, filter-sterilize and store at À20 C. 7. 30 mg/mL kanamycin stock solution: Prepare 0.3 g kanamycin in 10 mL of dH2O, filter-sterilize, aliquot and store at À20 C. 8. 40% glucose stock solution: Prepare 40 g glucose in 100 mL of dH2O, autoclave and store at room temperature.

9. 80% Glycerol: Prepare 80 mL glycerol in 20 mL of dH2O, autoclave and store at room temperature. 10. 2 YT agar: Prepare 31 g premixed 2 YT and 15 g agar in 1 L dH2O, autoclave, cool to 55 C, add 2% glucose and appropriate antibiotics. 11. 20% Polyethylene glycol 6000/2.5 M NaCl (PEG/NaCl) solution: Prepare 200 g PEG and 146 g NaCl in 1 L of dH2O, autoclave and store at room temperature.

12. PBS buffer: Add 8 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4 and 0.24 g KH2PO4 in 1 L dH2O, adjust to pH 7.4, autoclave and store at room temperature. 13. M13KO7 helper phage (NEB). 14. CryoTube™ vials (Thermo Scientiﬁc).

2.2 Phage Display 1. Target antigen: Biotinylated recombinant Hemolysin E protein MSIA™ Streptavidin (see Note 2) in PBS/bicarbonate buffer (0.1 M NaHCO3, D.A.R.T’s® Biopanning pH 8.6 (see Note 3). 2. TG1 E. coli cell: Contains the lacIqZΔM15 gene on the F0 episome which allows blue-white screening for recombinant plasmids.

3. 0.1 M NaHCO3 (Bicarbonate Buffer) : Add 0.84 g NaHCO3 in 100 mL of dH2O, adjust to pH 8.6. 4. 0.5% PBS-T: Add 5 mL Tween 20 into 1 L PBS. 5. 3% PTM: Add 3 g skimmed milk in 100 mL 0.1% PBS-T.

6. 0.2 M glycine, pH 2.2: Prepare 1.5 g glycine in 100 mL dH2O, adjust to pH 2.2, autoclave and store at room temperature.

7. 1 M Tris–HCl: Add 3.0275 g Tris in 25 mL dH2O, adjust to pH 9.1. ® 8. MSIA™ Streptavidin D.A.R.T’s (Thermo Scientiﬁc) (see Note 4). 290 Chai Fung Chin et al.

2.3 Phage ELISA 1. 96-well microtiter plate (Greiner) and 96-well strip microtiter ® plate Costar (Corning). 2. Anti-M13 horseradish peroxidise (HRP)-conjugated monoclonal antibody. 3. 2% BSA: Prepare 2 g BSA in 100 mL of 0.1% PBS-T. 4. ABTS developing solution: add one 10 mg tablet ABTS in 5 mL of 50 mM citric acid, 5 mL of 50 mM trisodium citrate, and 10 μLH2O2. Store in the dark.

2.4 DNA Sequencing 1. Minipreps: QIAprep spin Plasmid Miniprep kit (Qiagen). 2. Primers: LMB3_Fw—50 CAGGAAACAGCTATGAC 30 and PIII_Rv—50 GTTAGCGTAACGATCTAA 30.

2.5 Soluble Antibody 1. 1 M Isopropyl β-D-1-thiogalactopyranoside (IPTG) stock solu- Fragments Detection tion: Prepare 2.38 g IPTG in 10 mL of dH2O, aliquot and store at À20 C. 2. 1Â TES buffer [30 mM Tris (pH 8), 1 mM EDTA and 20% sucrose]: Add 1.5 mL 1 M Tris (pH 8.0), 0.05 mL 1 M EDTA and 10 g sucrose into dH2O in a total volume of 50 mL. Store at 4 C. 3. 1:5 dilution 1Â TES buffer: Add 10 mL of 1Â TES buffer into 40 mL dH2O. Store at 4 C. 4. 3% PTM. 5. Horseradish peroxidase-anti-c-Myc antibody. 6. PBS.

3 Methods

3.1 Preparation In order to prepare sufficient starting material for the biopanning of Antibody Library process to be carried out, a synthetic domain antibody library was Phage amplified from the library stock. Since the phagemid system is employed, an additional co-infection with M13KO7 helper phage is required during the packaging process of the antibody library phage. 1. Thaw the glycerol stock of the antibody library and start culturing in 500 mL of 2 YT containing 2% glucose and ampicillin (100 μg/mL) whereby the starting inoculation is at OD600 ~ 0.1. 2. Grow the culture at 37 C with 200 rpm shaking until OD600 ~ 0.5. 3. Divide the culture equally into two flasks; one is for phage packaging whereas the other is stored as the first generation stock. MSIA Streptavidin DARTs Panning 291

3.1.1 Library Phage 1. For phage packaging purposes, co-infect the culture with Packaging M13KO7 helper phage (1011 CFU/mL) by incubation at 37 C static for 30 min (see Note 5). 2. Centrifuge the culture at 1726 Â g for 30 min and discard the supernatant. 3. Reconstitute the pellet with 250 mL of 2 YT medium containing 0.1% glucose, ampicillin (100 μg/mL) and kanamycin (60 μg/mL) with gentle mixing (see Note 6). 4. Grow the bacteria in 2 YT medium o/n at 30 C with 180 rpm agitation. This step is to amplify/package phagemid bearing phage particles. 5. The next day, centrifuge the culture at 1726 Â g for 30 min to collect phage-containing supernatant. 6. To the supernatant, add an additional 1/6 of the total supernatant volume with PEG/NaCl and chill on ice for 1 h to precipitate the phage. 7. Centrifuge the mixture at 1726 Â g for 30 min. 8. Discard the supernatant and air-dry the white color phage pellet. 9. Resuspend the pellet with 1 mL of PBS buffer. 10. Centrifuge the mixture at maximum speed (21,130 Â g) for 20 min. Additional centrifugation may be required to ensure total removal of bacterial culture in the supernatant. 11. Store the supernatant containing the library phage at 4 C until ready for use. 12. Perform phage titration to estimate the amount of phage particles present. Prepare a series of 1:10 phage dilution by adding 10 μL of phage with 90 μL PBS. Then, add 200 μL of the TG1 culture (OD600 ~ 0.5) to the 100 μL ﬁnal volume of phage dilution prepared earlier and incubate static at 37 C for 30 min. Spot 10 μL of each dilution, TG1 and PBS (as negative controls) on 2 YT agar plates containing ampicillin (100 μg/ mL) and kanamycin (60 μg/mL) respectively. The phage particles with the desired antibody phagemid genome would survive on ampicillin agar plate after infection with TG1 pilus- bearing bacteria, whereas phage with M13KO7 genome would only survive on kanamycin plates. Incubate the agar plates o/n at 37 C and calculate the number of colonies on the ampicillin agar plate.

3.1.2 Preparation of First 1. For the ﬁrst generation library stock, leave the culture to grow Generation Stock o/n at 37 C with 200 rpm agitation. 292 Chai Fung Chin et al.

2. The next day, centrifuge the culture at 1726 Â g for 30 min to collect the cell pellet that is later resuspended with fresh 5 mL of 2 YT containing 2% glucose and ampicillin (100 μg/mL). 3. Add 20% glycerol to the mixture, aliquot into CryoTube™ vials for storage at À80 C.

® 3.2 Phage Display MSIA™ streptavidin D.A.R.T’s antibody phage display biopan- Biopanning ning is according to the protocol by Chin et al. [11]. Nonetheless, in order to choose the best biopanning approach that is feasible for your laboratory, there are several aspects that you may need to take into consideration: (1) Choose the biopanning approach based on equipment/materials available in your laboratory. For instance, in ® order to perform MSIA™ streptavidin D.A.R.T’s antibody phage display biopanning, Finnpipette™ Novus i Electronic 12-channel Pipette (Thermo Scientific) and pipette stand are needed in addi- ® tion to D.A.R.T’s . (2) Target antigen used. Biotinylated antigens ® are required for MSIA™ streptavidin D.A.R.T’s antibody phage display biopanning. The optimized condition whereby each biopanning is carried out may vary. Therefore, whenever there is a high background generated during the biopanning, repetition of current or earlier step is required until the enrichment of target-specific antibodies is achieved. Modifications may need to be carried out to optimize the panning process for different targets.

® 3.2.1 MSIA™ 1. Mount the MSIA™ Streptavidin D.A.R.T’s to a multi- Streptavidin D.A.R.T’s® channel electronic Finnpipette™ for loading of biotinylated Loading of Biotinylated target antigens to take place. Antigen 2. Load biotinylated recombinant antigen at 100 μg(see Note 7) ® in bicarbonate buffer to MSIA™ Streptavidin D.A.R.T’s by continuous aspiration and dispensing. Set the electronic pipette program for 999 cycles at a Speed Setting of 5 with a ﬁxed volume of 150 μL. Continuous aspiration and dispensing at a moderate speed could help in binding the biotinylated targets ® to the streptavidin in the D.A.R.T’s . ® 3. Wash the MSIA™ Streptavidin D.A.R.T’s two times (20 cycles, Speed Setting 8 and volume 200 μL) with 0.5% PBS-T followed by one time (20 cycles, Speed Setting 8 and ® volume 200 μL) with PBS. The antigen-captured D.A.R.T’s is now ready for use in biopanning.

3.2.2 MSIA™ 1. Block the antigen-coupled tip with 3% PTM with continuous Streptavidin D.A.R.T’s® aspiration and dispensing at 500 cycles with a Speed Setting of Antibody Biopanning 5 and volume of 200 μL. MSIA Streptavidin DARTs Panning 293

2. At the same time, preincubate ~1012 phage particles (see Note 8) of the antibody library with 3% PTM to minimize background from the system. ® 3. Subsequently, wash the D.A.R.T’s two times (20 cycles, Speed Setting 8 and volume 200 μL) with 0.5% PBS-T followed by one time (20 cycles, Speed Setting 8 and volume 200 μL) with PBS. 4. Capture the antibody phage in PTM by performing repetitive pipetting with a fixed volume of 150 μL, 999 cycles repeat and a Speed Setting of 5. ® 5. Rinse the D.A.R.T’s for five rounds with 0.5% PBS-T and another five rounds with PBS (see Note 9). Each wash cycle constitutes 20 cycles of aspirating and dispensing with a speed setting of 8 and volume of 200 μL. 6. Elute the bound phages by using 100 μL of 0.2 M glycine, pH 2.2 (see Note 10) with 300 cycles of repetitive pipetting (Speed Setting 3) (see Note 11). 7. Immediately neutralize the eluted fraction with 1 M Tris–HCl, pH 9.1 to achieve pH 7. This step has to be done immediately to prevent the decrease of phage infectivity. 8. Infect the eluted phages with an exponentially growing 2–4 mL TG1 culture (OD600 ~ 0.5) with incubation at 37 C for 30 min static for phage rescue. 9. At the same time, perform phage titration as described in step 12, Subheading 3.1.2 (see Note 12). 10. Centrifuge the infected culture at 9000 Â g for 10 min. 11. Discard the supernatant and mix the cell pellet with 20 mL 2 YT medium containing 2% glucose and ampicillin (100 μg/ mL). 12. Grow the culture at 37 C with 200 rpm agitation for approximately 3–4 h (see Note 13). 13. Equally divide the culture into two where one is kept as glycerol stock as described earlier (Subheading 3.1, step 3)in1mL 2 YT medium instead of 5 mL and the other half is co-infected with helper phage (~1010 CFU/mL) and incubated at 37 C static for 30 min. 14. Centrifuge the co-infected culture at 9000 Â g for 10 min. 15. Reconstitute the pellet with the same volume of 2 YT medium containing 0.1% glucose, ampicillin (100 μg/mL), and kanamycin (60 μg/mL). Grow the culture o/n at 30 C with 180 rpm agitation. 16. The next day, centrifuge the culture at 9000 Â g for 30 min to collect phage-containing supernatant. 294 Chai Fung Chin et al.

17. Perform phage precipitation (steps 6–11, Subheading 3.1.2) and titration (step 12, Subheading 3.1.2) as previously described. The ﬁnal volume for the antibody phage in PBS is 300 μL. Repeat this process for 2–5 rounds in order to obtain target-enriched phages (see Note 14).

3.3 Phage ELISA After approximate three rounds of biopanning, polyclonal phage ELISA is performed in order to observe the enrichment patterns of the biopanning. After the successive rounds of biopanning, the clones from the panning round with the highest enrichment will be plated out to screen for target-speciﬁc monoclonal antibodies. Propagate the monoclonal antibody phage and screen with monoclonal ELISA.

3.3.1 Polyclonal Phage 1. For three rounds of biopanning, coat three wells of Costar ELISA EIA/RIA microtiter plate with 100 μL of target antigen (10 μg) in bicarbonate buffer/PBS buffer o/n at 4 C with 700 rpm agitation. Coat another three wells with 300 μLof2% BSA concurrently to be used as background control (see Note 15). 2. The next day, wash the plate three times with 0.5% PBS-T. 3. Block the wells with 300 μL of 2% BSA for 1–2 h with 700 rpm agitation to reduce nonspeciﬁc binding. 4. Add 100 μL of (109) enriched phage particles in 2% BSA from the biopanning process to the wells coated with target as well as preblocked coated wells. Incubate the plate for 1–2 h with 700 rpm agitation. 5. Wash the plate thrice with 0.5% PBS-T. 6. Add 150 μL of anti-M13 HRP (1:5000) in 2% BSA to the wells and incubate for 1–2 h with 700 rpm agitation. 7. Wash the plate three times with 0.5% PBS-T. 8. Add 150 μL of the ABTS developing solution to detect bound phages. After 30 min incubation in the dark, the absorbance reading at 405 nm (OD405) is recorded with a microplate reader. The incubation is done in the dark as ABTS solution is light-sensitive.

3.3.2 Monoclonal Phage 1. Plate out the phages infected with bacteria from the biopan- Propagation ning with the highest enrichment. Dilute the polyclonal phage À in 1:10 serial dilution until 10 10 and further infect with 200 μL of TG1 culture (OD600 ~ 0.5) at 37 C, static for 30 min. Then, 100 μL of the infected culture is plated on 2 YT agar plate containing ampicillin (100 μg/mL). The plates are incubated at 37 C for o/n. MSIA Streptavidin DARTs Panning 295

2. Pick a total of 93 single colonies and grow in 2 YT containing 2% glucose and ampicillin (100 μg/mL) at 37 C, 900 rpm o/n in a round-bottom microtiter plate. On the plate, wells in position A2 were left empty as negative control while position A1 was cultured with a known clone as positive control (see Note 15). 3. The next day, inoculate 10 μL of the o/n culture in 200 μLof 2 YT containing 2% glucose and ampicillin (100 μg/mL) and further grow at 37 C, 900 rpm for 2.5 h. The o/n culture is added with glycerol to a ﬁnal concentration of 20% and kept at À80 C as stock. 4. After the incubation, add 20 μL of M13KO7 helper phage (~109) for co-infection by static incubation at 37 C for 30 min. 5. Centrifuge the culture at 563 Â g for 10 min. 6. After discarding the supernatant, resuspend the cell pellet with 220 μL of 2 YT containing 0.1% glucose, ampicillin (100 μg/ mL), and kanamycin (60 μg/mL). 7. Incubate the culture at 30 C with 900 rpm agitation for o/n. 8. The next day, centrifuge the culture at 563 Â g for 10 min and phage containing supernatant is collected as well as kept at 4 C until ready for use.

3.3.3 Monoclonal ELISA Perform the monoclonal phage ELISA in a similar way to polyclonal phage ELISA (Subheading 3.3.1) described earlier. A total of 50 μL sample monoclonal antibody phage is used to perform monoclonal ELISA. In addition, positive and negative controls are included in the ELISA to ensure the validity of the ELISA (see Note 15).

3.4 DNA Sequencing 1. Mimiprep of clones that showed positive binding activities using QIAprep Spin Miniprep Kit. 2. The puriﬁed plasmid-DNA is sent for sequencing with LMB3 forward and pIII reverse primers.

3.5 Generation After verifying the positive binders, expression and validation of the of Soluble Antibody antibody fragments in soluble form will be of priority. Nonetheless, Fragments the soluble antibody fragments that are expressed will still have pIII minor coat protein attached, as the amber stop codon will not be read as stop in amber suppressor strains, i.e., TG1 cell. Alterna- tively, the target clones can be infected into non-amber suppressor E. coli strains such as Top10 F0 in order to express the soluble antibody fragment independent of the pIII. 296 Chai Fung Chin et al.

3.5.1 Expression 1. Pick single colonies from the target-speciﬁc monoclonal anti- and Extraction of Soluble body bacteria colonies growing on the agar plate. Culture the Antibody colonies o/n at 37 C in 5 mL 2 YT medium containing 2% glucose and 100 μg/mL ampicillin. 2. The next day, inoculate the o/n culture in 100 mL 2 YT medium containing ampicillin (100 μg/mL) and 0.1% glucose at 1:100 ratio and further grow it at 37 CtoOD600nm ¼ 0.6. 3. Induce the lac promoter from pLABEL phagemid with 1 mM IPTG and further express the clones o/n at 25 C with 160 rpm agitation for 16 h. 4. At the following day, centrifuge the culture at 1726 Â g for 30 min to collect the cell pellet in 100 mL fraction. 5. Resuspend the cell pellet fraction of 100 mL expressed antibody in 1 mL of cold 1Â TES buffer. 6. Add 1.5 mL of 1:5 dilution cold 1Â TES buffer and mix gently. Incubate the mixture on ice for 1 h. The protein extraction method used is by hypotonic shock to release soluble antibodies especially in the periplasmic region of the bacteria. 7. Centrifuge at 9000 Â g for 10 min and collect the antibody containing supernatant. Keep the soluble antibody at À20 C until ready for use.

3.5.2 Soluble ELISA 1. Coat 10 μg/well of target antigen on the microtiter plate o/n at 4 C in PBS buffer and 3% PTM as background control (see Note 15). 2. Wash the wells three times with 0.5% PBS-T. 3. Block the wells with 3% PTM for 1 h with 700 rpm agitation to reduce nonspeciﬁc binding. 4. Wash the wells thrice with 0.5% PBS-T. 5. Incubate the monoclonal soluble antibody with target-coated wells and preblocked wells for 1 h with 700 rpm agitation. 6. Add anti-c-Myc-HRP antibody (1:2500 in PTM) into the wells and incubate for 1 h with 700 rpm agitation. 7. Add ABTS developing solution in the dark to detect bound antibody carrying the peroxidase enzyme that covert the substrate into color product. Record the absorbance reading (OD405nm) using a microplate reader.

® 3.6 Analysis After performing MSIA™ streptavidin D.A.R.T’s biopanning, typical enrichment pattern for polyclonal phage ELISA will be observed for successful biopanning. After the enrichment pattern is observed, the monoclonal domain antibodies selection is performed to isolate target-speciﬁc antibodies from the antibody library. Phage enrichment ratios may be used to gauge the MSIA Streptavidin DARTs Panning 297

enrichment process of the panning experiment. A similar polyclonal antibody ELISA pattern can be expected using the MSIA™ strep- ® tavidin D.A.R.T’s biopanning protocol. The positive clones obtained will also need to be sequenced to determine the identity and diversity of the clones enriched. In addition, soluble monoclonal antibody ELISA must be performed to validate the functionality of isolated monoclonal antibodies in soluble form. This will allow the identiﬁcation of functional and soluble antibody clones that can be applied for multiple downstream applications.

4 Notes

1. Other phage display antibody library with different antibody ® formats could also be used in MSIA™ Streptavidin D.A.R.T’s biopanning. The method is not limited to only domain antibodies but can also be applied for scFv and Fab libraries. 2. Biotinylated targets are needed for coupling to the streptavidin ® D.A.R.T’s in order to perform the biopanning. Target antigens can be prepared either by chemical conjugation of biotin or by in vivo biotinylation methods. The strong biotin- streptavidin interaction enables target to be coupled and presented on the surface of the tip with higher efficiency and efficacy. ® 3. Buffers used to load the antigen to streptavidin D.A.R.T’s can be varied depending on the suitability of the buffer to maintain the stability of the antigens. 4. An electronic multichannel pipette with programming functions for continuous aspiration is required. We applied the Finnpipette™ Novus i Electronic 12-channel with an adjustable pipette stand that was designed for use with the MSIA™ ® Streptavidin D.A.R.T’s system. However, other electronic multichannel pipette systems with similar functions can be applied. 5. The static condition is to prevent the destruction of the pilus that may affect the bacteria infectivity. 6. Antibiotics selection enables differentiation between bacteria cells with antibody phagemid (ampicillin resistant) and M13KO7 genome (kanamycin resistant). 7. The starting materials or target antigens can be scaled up for ® optimum coupling to the D.A.R.T’s if the coupling is not efficient. 8. Try to ascertain the similar amount of phage as input for each biopanning round to ensure a successful biopanning. 298 Chai Fung Chin et al.

9. A more stringent/increment of wash in washing step for each ® increased round of MSIA™ Streptavidin D.A.R.T’s biopanning, especially before acid elution may help in enriching target-specific antibody/reduce background interference. 10. Acid elution within acidic pH range can be used in MSIA™ ® Streptavidin D.A.R.T’s system for the elution of target- specific phage antibodies and does not interrupt the biotin- streptavidin interaction between the biotinylated targets with the streptavidin in the system. 11. The elution step that involves acid elution of the bound phages can be modified with more cycles of repetitive pipetting (300–- 500 cycles) in order to elute the bound phages more effectively. With that, target enriched phages can be rescued and retrieved more efficiently. Be careful not to apply too many cycles as the extended exposure to acidic buffers may not be favorable for certain antigens and antibody clones. 12. The concentration of the eluted phage may vary from each round of biopanning. However, it is critical to ensure the amount of blank phage is less than target phage to promote a higher success rate for biopanning. ® 13. MSIA™ Streptavidin D.A.R.T’s biopanning process usually takes one day to accomplish one round of biopanning. None- theless, if the enriched phage that is infected with bacterial cell does not grow well, o/n growth may be required and followed by co-infection with helper phage in the next day. This could result from a low level of phage recovery from that particular biopanning round. 14. Usually three rounds of biopanning will suffice to enrich target-specific phage; however, sometimes, up to five rounds of biopanning may be required to enrich target-specific phages. 15. Make sure there are appropriate controls, especially for ELISA to ensure validity of the data.

Acknowledgments

The authors would like to acknowledge the support of the Malay- sian Ministry of Higher Education Fundamental Research Grant Scheme (203/CIPPM/6711381) and Malaysian Ministry of Edu- cation through the Higher Institution Centre of Excellence (HICoE) Grant (Grant No.311/CIPPM/4401005). MSIA Streptavidin DARTs Panning 299

References

1. Zwick MB, Shen J, Scott JK (1998) Phage- Kontermann R, Dubel€ S (eds) Antibody engi- displayed peptide libraries. Curr Opin Biotech- neering. Springer, Berlin, pp 267–287 nol 9:427–436 9. Nelson RW, Krone JR, Bieber AL, Williams P 2. Smith GP (1985) Filamentous fusion phage: (1995) Mass spectrometric immunoassay. Anal novel expression vectors that display cloned Chem 67:1153–1158 antigens on the virion surface. Science 10. Trenchevska O, Nelson R, Nedelkov D (2016) 228:1315–1317 Mass spectrometric immunoassays in character- 3. Brichta J, Hnilova M, Viskovic T (2005) Gen- ization of clinically significant proteoforms. eration of hapten-specific recombinant antibo- Proteomes 4:13. https://doi.org/10.3390/ dies: antibody phage display technology: a proteomes4010013 review. Vet Med 50:231–252 11. Chin CF, Ler LW, Choong YS, Ong EB, 4. Hammers CM, Stanley JR (2014) Antibody Ismail A, Tye GJ, Lim TS (2016) Application phage display: technique and applications. J of streptavidin mass spectrometric immunoas- Invest Dermatol 134:e17. https://doi.org/ say tips for immunoaffinity based antibody 10.1038/jid.2013.521 phage display panning. J Microbiol Methods 5. Carmen S, Jermutus L (2002) Concepts in 120:6–14 antibody phage display. Brief Funct Genomic 12. Wilchek M, Bayer EA (1988) The avidin-biotin Proteomic 1:189–203 complex in bioanalytical applications. Anal Bio- 6. Moulard M, Zhang MY, Dimitrov DS (2004) chem 171:1–32 Novel HIV neutralizing antibodies selected 13. Williams S, Van der Logt P, Germaschewski V from phage display libraries. In: Subramanian (2001) Phage display libraries. In: Westwood G (ed) Antibodies, Novel technologies and OMRH, F. C. (eds) Epitope mapping: a prac- therapeutic use, vol 2. Kluwer Academic, tical approach. Oxford University Press, New York, NY, pp 105–118 Oxford, pp 225–254 7. McConnell SJ, Dinh T, Le MH, Spinella DG 14. Hairul Bahara NH, Chin ST, Choong YS, Lim (1999) Biopanning phage display libraries TS (2016) Construction of a semisynthetic using magnetic beads vs. polystyrene plates. human VH single-domain antibody library BioTechniques 26(208–210):214 and selection of domain antibodies against 8. Konthur Z, Wilde J, Lim TS (2010) Semi- alpha-crystalline of Mycobacterium tuberculo- automated magnetic bead-based antibody sis. J Biomol Screen 21:35–43 selection from phage display libraries. In: Chapter 16

Magnetic Nanoparticle-Based Semi-Automated Panning for High-Throughput Antibody Selection

Angela Chiew Wen Ch’ng, Nurul Hamizah Binti Hamidon, Zolta´n Konthur, and Theam Soon Lim

Abstract

The application of recombinant human antibodies is growing rapidly mainly in the field of diagnostics and therapeutics. To identify antibodies against a specific antigen, panning selection is carried out using different display technologies. Phage display technology remains the preferred platform due to its robustness and efficiency in biopanning experiments. There are both manual and semi-automated panning selections using polystyrene plastic, magnetic beads, and nitrocellulose as the immobilizing solid surface. Magnetic nanoparticles allow for improved antigen binding due to their large surface area. The Kingfisher Flex magnetic particle processing system was originally designed to aid in RNA, DNA, and protein extraction using magnetic beads. However, the system can be programmed for antibody phage display panning. The automation allows for a reduction in human error and improves reproducibility in between selections with the preprogrammed movements. The system requires minimum human intervention to operate; however, human intervention is needed for post-panning steps like phage rescue. In addition, polyclonal and monoclonal ELISA can be performed using the semi-automated platform to evaluate the selected antibody clones. This chapter will summarize the suggested protocol from the panning stage till the monoclonal ELISA evaluation. Other than this, important notes on the possible optimization and troubleshooting are also included at the end of this chapter.

Key words Panning, Antibody library, Monoclonal antibodies, Phage display, Semi-automated, Mag- netic nanoparticle

1 Introduction

The rapid growth of recombinant human antibodies can mainly be attributed to its role in biomedical applications ranging from laboratory scale research, medical diagnostics and more importantly as therapeutics [1–4]. To date, the recent developments in molecular biology and recombinant DNA technology have allowed the establishment of various novel display technologies [5]. Many of these methods have been used for human antibody generation, namely yeast display, ribosome display, mammalian cell display, bacterial display, covalent DNA display, and mRNA display

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_16, © Springer Science+Business Media LLC 2018 301 302 Angela Chiew Wen Ch’ng et al.

[1, 5–8]. However, phage display technology remains the most favored platform for monoclonal antibody generation due to its robustness and efficiency in antibody development projects [6, 7]. This is evident with the number of phage-derived antibodies that have found their way into clinical trials and the market. First and foremost, the antibody generation process with phage display involves the availability of phage-derived antibody libraries for screening. The typical libraries used for phage display can be classified as naive, immune, or synthetic in nature [9–11]. The characteristics of the different libraries have been extensively reviewed elsewhere [12–17]. In essence, these libraries differ mainly in the cDNA source for library generation. The quality of a library is mainly attributed to its size and diversity. Large library sizes are preferred as the higher number of clones available allows an increased probability to enrich binders against a specific antigen. The diverse repertoire is also vital to allow enrichment of higher affinity antibodies. The choice of antibody library to be used is massively dependent on the antigen specificity. However, generally naive and synthetic libraries are virtually universal but recent reports highlight the ability of immune libraries to be applied for antibody generation against non-disease-specific targets [13, 18]. The eventual yardstick would be dependent on the quality of the clones rather than the ability to enrich a clone. In general, antibody generation using phage display is an iterative process that allows continuous concentration of bound clones by constant isolation and multiplication of an explicit pool of clones [19]. The selection procedure commonly known as panning involves several key stages. This includes affinity-induced target capture, isolation, retrieval, and rescue prior to analysis usually by ELISA [20, 21]. In order to facilitate target capture, antigens are anchored to solid supports, such as column packing materials, magnetic particles, nitrocellulose membranes with plastic surfaces being the most preferred surface [22, 23]. The common plastic surfaces used include polystyrene immune tubes or microtiter wells. These surfaces function as physical materials to capture and hold the antigen in place for antibody capture by affinity. A collection of antibody presenting phage particles are incubated with the antigen bound solid surfaces to allow target-specific antibodies to bind to the target antigen. Then a wash step is introduced to isolate nonspecific, unbound, or even weak binders from the solid surface. The remaining phage particles are then retrieved in most cases by elution using salts, pH dissociation [24], or even enzymatic cleavage [25]. In some cases, a direct rescue can also be carried out without the need of phage elution. A rescue process whereby retrieved phage particles are re-infected with Escherichia coli is carried out to propagate a concentrated population of phage. The re-amplified phage can then be used for subsequent rounds of panning for further clone enrichment or analysis. The number of panning Magnetic Nanoparticle Based Panning 303 rounds to be carried out normally falls in between two to four rounds. However, this is normally carried out until a satisfactory enrichment pattern is obtained. It is important to note that phage infection and propagation steps although robust are labor intensive as it involves several biological processes that require care, making it difficult to automate. The introduction of automation in laboratory routines has allowed for more reproducible, efficient, and faster processes to be carried out [26]. An automated system refers to the ability to carry out a set of predefined processes in a pipeline without the need of human intervention. Semi-automated processes however require the involvement of humans for particular steps in certain stages throughout the process. This allows only particular stages in a pipeline to be automated independently of the remaining stages. In most instances, the need to automate a workflow requires initial investment in terms of cost and also the added concern of multiple instruments that may not be compatible [20, 23]. Therefore, in order to automate a workflow in smaller scale laboratories, the ability to have one platform or instrument that can automate majority of the stages in the workflow would be ideal. In terms of phage display panning, the adaptation of magnetic nanoparticles with a magnetic particle processor would ideally suit the need for semi-automation of the panning process. Looking back at the stages in the phage display panning process, it is clear that panning process is a multi-stage process where only parts of the process can be automated [20, 22]. The main stages that are compatible for automation are panning, infection, propagation, and ELISA evaluation using just one similar instrument. Colony picking processes can be automated provided access to a costly robotic picking instrument is available. There are several magnetic particle processors (MPP) available in the market from different manufacturers, which includes the Kingfisher line, MagMax series, and Maxwell systems. All these systems are similar in function that the sole function of the instrument is to capture, release, and move magnetic particles in the solution. Most of these systems were originally designed for automating RNA, DNA, and protein extraction by using magnetic beads. The open platform of most of these equipments means that antibody selection processes could be carried out using these systems too. The main advantage of automating several stages in the panning process is the reduction of human errors and increase in process consistency. Multi-sample processing is also possible with the use of larger versions of these units that can handle 96-well microtiter plates. This chapter is an updated version of the previous chapter published by Konthur (2010). We applied a pin-based magnetic particle processor (MPP) (KingFisher Flex, Thermo) for semi- automation of the panning process. The MPP can accommodate a 304 Angela Chiew Wen Ch’ng et al.

96-magnet pins handle on the moving arm that is aligned according to the positions of a standard 96-well microtiter plate [27]. The MPP magnet pins are equipped with a plastic cover to function as a border between magnet particles from the magnetic pins. The MPP magnet pins function to physically relocate magnetic particles from one well to another. The intuitive software that controls the movements of the magnetic pins allows for personalized incubation times, capture release frequency, position movement, frequency of pin movement, temperature, and pin movement speeds. We find that adaptation of the MPP system has several advantages over the conventional microtiter plate panning. Applying the MPP system allows for higher reproducible sample handling with reduced exper- imentation errors even when dealing with multiple samples. The physical transfer of magnetic particles from one well to another reduces background selection of nonspecific binders that may be trapped on the surfaces of solid supports and ensures minimal volume transfer [23]. The panning protocol is generally conducted using a MPP over four rounds of selection, as shown in Figs. 1 and 2, Tables 1 and 2. Applying the MPP protocol allows for standardization of panning parameters as well as personalization of parameters when dealing with different targets or for the enrichment of antibodies with specific characteristics. The controlled condition and automated process allow for reproducible multi-target panning and multi- library panning using different buffers. As highlighted earlier, the entire panning process involves multiple biological processes that make full automation of the panning process expensive. The semi- automated protocol would still require an element of human intervention at particular steps in the process. In order to produce a convenient standard operating procedure, human intervention is kept minimal and simple. Another obvious convenience in using the 96-well format MPP is that any manipulation that requires human participation can be carried out using multichannel pipet- tors. The entire protocol has been set up in a streamlined manner to conduct all major stages of panning using the MPP with minimal human intervention. This includes antigen loading to magnetic beads, phage selection process by affinity selection, phage rescue/ amplification in between rounds, and the sample confirmation by ELISA. The panning round evaluation consists of two stages of ELISA, with the first focusing on polyclonal level analysis to provide an idea of the enrichment pattern as shown in Fig. 3. The second stage involves the selection of individual clones in a 96-well format from specific selection rounds with good enrichment ratios based on the polyclonal ELISA and phage titer results. The individual clones would then be packaged independently as antibody presenting phage particles for analysis using ELISA. In summary, semi-automation of the phage panning process allows for an increase in the number of target antigens using Magnetic Nanoparticle Based Panning 305

Fig. 1 Rotating table of Kingﬁsher 96 magnetic particle processor of Thermo Scientiﬁc

different libraries or buffer conditions in parallel. The protocol would be practical for research laboratories or even for larger scale antibody discovery centers. The minimum human intervention throughout the process would provide higher confidence in results, thus minimizing the need for repeats. This allows for the elevation of the bottleneck associated with conventional panning strategies from the early stages down the pipeline to the monoclonal binder analysis and identification (Fig. 4). Even so, several other strategies have been introduced to allow rapid and automated monoclonal binder isolation, screening, and identification [1]. The protocol allows for a reasonably cost semi-automated process as an alternative to the labor-intensive conventional methods and expensive fully robotic processes for antibody generation.

2 Materials

2.1 Loading 1. Dynabeads™ M-280 Streptavidin (Invitrogen Dynal AS, Oslo, of Magnetic Beads Norway). 306 Angela Chiew Wen Ch’ng et al.

Fig. 2 The operating mode of magnetic particle processor in magnetic beads capturing and releasing by the magnetic head covered with plastic comb, then transfer from 1 plate to another plate by rotating the plate holder table clockwise. (I) The rod shape magnet is covered by plastic comb and move into the solution containing preloaded magnetic beads. (II) Mixing is done by moving slowly up and down. (III) Moving the covered magnet to plate 2 to transfer the beads to a new solution. (IV) The magnet is removed from the plastic cover, the beads slowly suspended into the solution again. (V) The magnet head and plastic cover were removed up to the starting position and continued the next stage process

Table 1 Overview of the automated magnetic particles panning procedure with Kingfisher 96

Plate no. Panning round 1 Panning round 2 Panning round 3 Panning round 4

1 Bead plate Bead plate Bead plate Bead plate 2 Phage plate Phage plate Phage plate Phage plate 3 Wash plate 1 Wash plate 1 Wash plate 1 Wash plate 1 4 Release plate Wash plate 2 Wash plate 2 Wash plate 2 5 E. coli culture plate Release plate Wash plate 3 Wash plate 3 6– E. coli culture plate Release plate Wash plate 4 7– – E. coli culture plate Release plate 8– – – E. coli culture plate Total time ~135 min ~145 min ~155 min ~165 min

2. Phosphate-buffered saline (PBS): 8 g/L NaCl, 0.2 g/L KCL, 1.44 g/L Na2HPO4∙2H2O, and 0.24 g/L KH2PO4, pH 7.4. 3. Phosphate-buffered saline Tween (PBST): PBS and 0.1% Tween-20. Magnetic Nanoparticle Based Panning 307

Table 2 The actual process during automated magnetic particles in operation mode

Plate Plate Volume Time no. name Process (μL) (min)

1 Bead plate Blocking the antigen loaded on magnetic beads with PTM 200 60 2 Phage Selection of antibody phage with magnetic beads 200 60 plate 3 Wash plate Wash 1 of the magnetic beads in PBST 200 10 1 4 Wash plate Wash 2 of the magnetic beads in PBST 200 10 2 5 Wash plate Wash 3 of the magnetic beads in PBST 200 10 3 6 Wash plate Wash 4 of the magnetic beads in PBST 200 10 4 7 Release Released the magnetic beads with speciﬁc binders after wash 200 5–10 plate with PBST 8– Total time 170

2.2 Semi-automated 1. TG1 genotype: supE thi-1 Δ(lac-proAB) Δ(mcrB-hsdSM)5 (rK– 0 Panning Using mK–) [F traD36 proAB lacIq ZΔM15]. a Magnetic Particle 2. 96-well V-bottom polypropylene (PP) microtiter plates (Nunc, Processor Wiesbaden, Germany). 3. 96-well U-bottom polypropylene (PP) microtiter plates (Nunc, Wiesbaden, Germany). 4. Aera Seal breathable sealing ﬁlm (Sigma-Aldrich, Taufkirchen, Germany). 5. Phosphate-buffered saline Tween Milk powder (PTM): PBS, 1% Tween-20, 2% non-fat dry milk powder, prepare fresh. 6. 2YT medium: 1.6% (w/v) tryptone, 1% (w/v) yeast extract, and 0.5% NaCl, pH 7.0. 7. 10Â Amp/Glu solution: 1 mg/mL ampicillin, and 20% (w/v) glucose in 2YT medium.

2.3 Packaging 1. M13 K07 Helperphage (New England BioLabs, Frankfurt, of Phagemids Germany). 2. 96-well ﬁltration plate: MultiScreenHTS Plates with hydrophilic Durapore PVDF membrane with 0.65 mm pore size (Millipore, Schwalbach/Ts, Germany). 308 Angela Chiew Wen Ch’ng et al.

Fig. 3 The typical ELISA result highlighting the enrichment patterns of antibody phage selection for four rounds. Pooled phage obtained from four rounds of panning against scFV library was added to ELISA well coated with the parental antigen

Fig. 4 The typical selected positive clones from the monoclonal ELISA. Individual phage clones were chosen from three and four round pooled of phage. Several clones from each pooled phage indicate the binding of the phage to the respective antigen Magnetic Nanoparticle Based Panning 309

3. 2YT-AG-2: 2YT medium containing 100 mg/mL ampicillin, 2% (w/v) glucose. 4. 2YT-AKG: 2YT medium containing 100 mg/mL ampicillin, 60 mg/mL kanamycin, 0.1 (w/v) glucose. 5. Glycerol solution: 80% (w/v) glycerol in distilled water, then autoclave.

2.4 Titration 1. 2YT-AG agar plates: 2YT medium containing 100 mg/mL of Phage Particles ampicillin, 2% (w/v) glucose, and 1.5% (w/v) agar-agar. 2. 2YT-KG agar plates: 2YT medium containing 60 mg/mL kanamycin, 2% (w/v) glucose, and 1.5% (w/v) agar-agar.

2.5 Magnetic Particle 1. Matrix 96-well polystyrene microtiter plates (Thermo Scien- ELISA of Polyclonal tiﬁc, Dreieich, Germany). Antibody Phage 2. Anti-M13 Horseradish Peroxidase (HRP)-conjugated monoclonal antibody. 3. ABTS developing solution: 10 mg tablet ABTS in 5 mL of 50 mM citric acid, 5 mL of 50 mM trisodium citrate, and 10 μLH2O2. Store in the dark.

2.6 Production 1. E. coli HB2151 genotype: K12 ara D(lac-proAB) thi/F0 of Soluble Monoclonal proA þ B lacIq lacZDM15. Antibody Fragments 2. 2YT-AG-0.1: 2YT medium containing 100 mg/mL ampicillin, in Microtiter Plates 0.1% (w/v) glucose. 3. 20 mM isopropyl-b-D-thiogalactopyranoside (IPTG).

2.7 ELISA of Soluble 1. Bovine Serum Albumin (BSA): 10 mg/mL stock solution Monoclonal Antibody in PBS. Fragments 2. Recombinant Protein L horseradish peroxidise (HRP)- in Microtiter Plates conjugated antibody.

3 Methods

All the designed protocols in this section are designed to allow the selection of more than one target antigen in parallel. The semi- automated phage display panning procedure requires minimal human intervention when compared with the conventional panning protocol (see Note 2). The antibody library used can be either naive, immune, or synthetic regardless of the format. The protocol can be applied for scFv, Fab, or domain antibodies (see Note 1). With the standard protocol outlined here, the sample handling via this automated selection will become more straightforward to produce high-throughput results with minimal handling by humans (see Notes 3 and 5). 310 Angela Chiew Wen Ch’ng et al.

3.1 Loading 1. Take 1 mg of Dynabeads™ M-280 Streptavidin magnetic of Magnetic Beads beads and wash 5 min for three times with 1.5 mL PBST at room temperature (RT). At the same time, dissolve 100–200 mg of biotinylated protein antigen or 1–2 mg of biotinylated peptide antigen in 1 mL PBS (see Note 4). Then, resuspend the 1 mL antigen solution with magnetic beads and incubate the mixture overnight (o/n) at 4 Cor1hatRTona rotator. 2. Remove the antigen solution and wash the magnetic beads three times with 1.5 mL PBST. 3. Discard the wash solution and resuspend the magnetic beads with 200 μL PBS and store antigen-loaded bead stock at 4 C until use.

3.2 Semi-automated 1. Culture single clone of TG1 in 5 mL of 2YT at 37 C, 200 rpm Panning on Magnetic overnight shaking. Particle Processor 2. Inoculate 20 mL 2YT in a 250 mL Erlenmeyer flask with 0.5 mL of a fresh overnight TG1 at 37 C and 200 rpm until OD600 ~0.5 3. Arranging bead plate (Plate no. 1). Fill the positions A1–A12 of a 96-well V-bottom PP (PP) microtiter plate with 180 μL PTM. Add 20 μL from corresponding antigen-loaded bead stocks for each antigen to the specified position, namely beads antigen 1 to positions A1, beads of antigen 2 to positions A2, and so on (Total of 12 antigens in this case) (see Notes 6 and 7). 4. Preincubate the unselected antibody phage library with unloaded magnetic beads with PTM to deplete selection matrix binders. In a 15 mL PP tube, add 2 mg Dynabead M-280 Streptavidin to 1 Â 1013–3 Â 1013 phage particles in 10 μL PTM and incubate for 1–2 h at RT. 5. Then, transfer the antibody phage library to a new 15 mL PP tube after centrifugation at 894 Â g. The magnetic beads are discarded. 6. Arrange the phage-plate (Plate no. 2 as shown in Fig. 1 and Table 1) for the first round. Fill positions A1–A12 of a 96-well V-bottom PP microtiter plate with 200 μL of the antibody phage library solution. The following rounds of panning will continue with step 9. 7. Arrange phage-plate for subsequent rounds and fill positions A1–A12 of a 96-well V-bottom PP microtiter plate with 100 μL PTM. Add 100 μL of the amplified phage solutions of the previous round according to the same antigen order in positions A1–A12. Magnetic Nanoparticle Based Panning 311

8. Prepare wash plate (plate nos. 3, 4, 5, 6 as shown in Fig. 1 and Table 1) and fill positions A1–A12 of a 96-well V-bottom PP microtiter plate with 200 μL PBST. 9. Prepare release plate (Plate nos. 4, 5, 6, 7 as shown in Fig. 1 and Table 1) and fill positions A1–A12 of a 96-well V-bottom PP microtiter plate with 200 μL PBS. 10. Place the plates in the Kingfisher 96 plate holder table according to the plate numbering in Table 1 and start the magnetic bead-based panning program. The magnetic beads should then move from plate to plate according to the program. 11. Incubate the beads in each plate. The beads should be kept in suspension by moving plastic tips up and down in the wells at medium speed (30–50 mm/s) during incubation (Fig. 2). Once the panning program has finished, prepare E. coli culture plate and fill positions A1–A12 of a 96-well U-bottom PP microtiter plate with 200 μLofE. coli TG1. Place E. coli culture plate in Kingfisher 96 instrument and start Transfer Program. This program simply transfers the beads from the release plate to the E. coli culture plate (see Note 17). 12. Take out the selection stock plate from the Kingfisher 96 instrument, cover with plastic lid, and incubate for 30 min at 37 C. 13. Remove the beads and then add into 20 μL 10 Amp/Glu solution, seal with breathable sealing film, and incubate in a microplate shaker for 2 h at 37 C and 1400 rpm. 14. Then, proceed to Packaging of Phage Particles protocol (Sub- heading 3.4). 15. Refer to Figs. 1 and 2 for the actual positions and operating mode.

3.3 Packaging 1. Take selection stock plate and add 200 μL of pre-warmed of Phage Particles 2YT-AG medium to culture, mix thoroughly, and transfer 200 μL into a 96-well filtration plate. 2. Seal the selection stock plate again with breathable sealing film and continue incubation in a microplate shaker overnight at 37 C at 1200 rpm. 3. Add 20 μL M13K07 helperphage ~109 phage particles covered with plastic lid to the filtration and incubate stationary for 30 min at 37 C. 4. Filter the bacterial culture by the centrifuge microtiter plate for 5 min at 894 Â g. 5. Discard the supernatant with remaining M13K07 helperphage. 6. Resuspend bacteria in 220 μL pre-warmed 2YT-AKG and transfer to a fresh 96-well U-bottom PP microtiter plate. Seal the phage production plate with breathable sealing film and 312 Angela Chiew Wen Ch’ng et al.

incubate in a microplate shaker overnight at 30 C shaking at 1400 rpm. 7. The next day, add 160 μL glycerol solution to the selection stock plate. Then, mix and store as glycerol stock at À80 C(see Note 11). 8. Pellet down the bacteria in phage production plate by centrifugation 10 min at 894 Â g. Transfer the supernatant carefully without disturbing the pellet to a 96-well filtration plate. 9. Place the filtration plate on the top of a new 96-well U-bottom PP microtiter plate and fix with sticky tape. 10. Filter antibody presenting phage particles to remove possible contaminating E. coli cells by centrifugation for 2–5 min at 894 Â g. 11. Store the filtrate. Discard bacteria pellets. 12. Add 50 μL PBS to each well of the phage stock plate and mix thoroughly. Use 100 μL for the next round of selection, use 10 μL for phage titration (see Notes 9, 10, and 14).

3.4 Titration 1. Inoculate 5 mL of 2YT in a 50 mL falcon tube with a single of Phage Particles clone of TG1 from an agar plate and grow overnight at 37 C and 250 rpm. 2. Inoculate 50 mL 2YT in a 250 mL Erlenmeyer ﬂask with 500 μL of overnight culture and grow at 37 C and 250 rpm until OD600~0.5. 3. Make 1:10 serial dilutions of phage suspension in PBS (see Notes 9, 10, 12, and 14). 4. Infect 200 μLofE. coli TG1 to phage dilutions and incubate for 30 min at 37 C without shaking. 5. Mix infected E. coli cultures and plate out 10 μL droplets of each dilution series on a single 2YT-AG and 2YT-K agar plates per enriched library. Incubate plates overnight at 37 C after the droplets are dried (see Notes 9, 10, 13, and 14).

3.5 ELISA 1. Fill each position of a 96-well V-bottom PP microtiter plate of Polyclonal Antibody with 180 mL 2% PTM and add 20 μL of antigen-loaded bead Phage stock according to the plate layout. Then, add magnetic beads of antigen 1 to positions A1–D1, beads of antigen 2 to positions A2–D2, and so on. 2. Use empty beads as a negative control. Take 5 mg (500 μL) Dynabeads™ M-280 Streptavidin magnetic beads and wash three times with 1.5 mL PBST and 1Â with 1.5 mL PBS at RT. Discard the last wash solution and resuspend in 1 mL. Add 20 μL to positions E1–H12 (Note: the wash can be done in rotary). Magnetic Nanoparticle Based Panning 313

3. Fill each position of a 96-well V-bottom PP microtiter plate with 150 mL PTM. Add 50 μL of phage solution from the phage stock plates of the individual rounds to the plate according to the layout. Add phage stocks of selection rounds 1–4 on antigen 1 to position A1–D1 and E1–H1 respectively. Add phage stocks of selection rounds 1–4 on antigen 2 to position A2–D2 and E2–H2 respectively and so on. 4. Prepare wash plates 1–3 and fill 96-well V-bottom PP microtiter plates with 200 μL PBST (see Note 15). 5. Prepare wash plate 4 by filling 96-well V-bottom PP microtiter plates with 200 μL PBS (see Note 15). 6. Add 4 μL mouse monoclonal anti-M13 HRP-conjugated to 20 mL 2% PTM (1:5000). Fill 96-well V-bottom PP microtiter plates with 200 μL antibody solution (see Note 15). 7. Place the plates in the Kingfisher 96 plate holder table and start magnetic bead-based ELISA program. The program should be set to move magnetic beads from plate to plate and incubate the beads in each plate (Fig. 2). During all incubations, the beads should be kept in suspension by moving plastic tips up and down in the wells at medium speed (30–50 mm/s). 8. While ELISA program is running, prepare the substrate plate. Dissolve one ABTS tablet (10 mg) in 20 mL substrate buffer. Shortly after the antibody plate incubation step in the ELISA process is finished, add 10 μL hydrogen peroxide to substrate solution and pipette 200 μL to each well of a Matrix 96-well polystyrene microtiter plate and place the plate in Kingfisher 96 (see Note 15). 9. Once beads are incubated in the substrate and color developed for 20 min, remove beads from the substrate by transferring them back to wash plate 4. 10. Take out Substrate plate from the Kingfisher 96 plate holder table and measure substrate specific extinction at 405 nm in an ELISA reader. 11. For each individual selection target, evaluate enrichment by plotting the obtained values for antigen-loaded and control beads of each phage selection round next to each other (Fig. 3)(see Note 9).

3.6 Production 1. Inoculate 5 mL of 2YT in a 15 mL PP tube with a single clone of Soluble Monoclonal of HB2151 from an agar plate and grow shaking overnight at Antibody Fragments 37 C and 250 rpm. in Microtiter Plates 2. Inoculate 50 mL 2YT in a 250 mL Erlenmeyer ﬂask with 0.5 mL of overnight HB2151 culture and incubate shaking at 37 C and 250 rpm until OD600 ~ 0.4–0.5. 3. Meanwhile, prepare a 1:10 dilution series of the desired panning round from the corresponding phage stock plate by adding 10 μL phage to 90 μL PBST. 314 Angela Chiew Wen Ch’ng et al.

4. Add 100 μL of TG1 E. coli cell (OD600 ~ 0.4–0.5) to phage dilutions and incubate for 30 min at 37 C. 5. Mix infected E. coli cultures and plate out 10 μL of each dilution series on a 2YT-AG agar plate. Once dried, incubate plates top-down at 37 C overnight. 6. Pick 92 clones into a 96-well U-bottom PP microtiter plate ﬁlled with 200 μL 2YT-AG (see Note 8). 7. Leave positions H3, H6, H9, and H12 empty for controls. Seal the mother plate with breathable sealing ﬁlm and incubate in a microplate shaker overnight at 37 C and 1400 rpm. 8. Next day, inoculate the fresh 96-well U-bottom PP microtiter plate containing 180 μL 2YT-AG with 20 μL of the overnight culture and incubate the daughter plate for 2 h at 37 C and 1400 rpm (see Note 15). 9. Add 150 μL glycerol solution to each well of the mother plate and store as glycerol stock À80 C(see Notes 11 and 15). 10. Induce soluble antibody fragment production in the daughter plate by adding 11 μL of 20 mM IPTG to each well and continue incubating overnight at 30 C, 1400 rpm (see Note 15). 11. Pellet down the bacteria by centrifuging the microtiter plates for 10 min at 2012 Â g. 12. Transfer soluble monoclonal antibody fragment containing culture supernatant into a fresh 96-well U-bottom PP microtiter plate and store until further use at 4 C. Discard the pellet- containing plate (see Note 15).

3.7 ELISA of Soluble 1. To analyze the antigen specificity of the soluble antibody frag- Monoclonal Antibody ment, coat half of a Matrix 96-well microtiter plate (positions Fragments A1–H6) by transferring (a) 1–2 mg protein antigen in 100 μL in Microtiter Plates PBS or (b) 10–20 μg peptide antigen in 100 μL PBS to each well. At the same time, coat the other half of the plate (positions A7–H12) with 100 μL/well of an appropriate negative control, such as Bovine Serum Albumin (10 mg/mL in PBS) or PTM and incubate the microtiter plate overnight at 4 C. 2. Discard coating solution and wash all wells two times for 5 min by completely filling them with PBST. 3. Block all wells by completely filling them with PTM and incubate for 1 h at RT. 4. Discard blocking solution and wash all wells three times for 5 min by completely filling them with PBST. 5. Fill each well with 50 μL PTM and 50 μL soluble antibody fragment solution of the respective 46 clones to each half of the Magnetic Nanoparticle Based Panning 315

plate (containing target antigen and a negative control, respectively) and incubate for 1 h at RT (see Note 15). 6. Discard soluble antibody fragment solution and wash wells three times for 5 min by completely filling them with PBST. 7. Add 100 μL of recombinant Protein L-HRP (1:5000 in PTM) to each well and incubate for 1 h at RT (see Note 15). 8. Discard recombinant Protein L-HRP solution. Then, wash the wells three times with PBST and two times with PBS (see Note 16). 9. Meanwhile, prepare substrate by dissolving one ABTS tablet (10 mg) in 20 mL substrate buffer. Immediately prior to use, add 10 μL hydrogen peroxide to the substrate solution. 10. Finally, add 100 μL of the substrate to each well and allow developing for 5–30 min at RT in the dark (see Notes 15 and 16). 11. Read substrate-specific extinction at 405 nm in an ELISA reader. 12. Plot the obtained values for antigen and negative control protein for each soluble monoclonal antibody fragment next to each other and identify positive candidates with an acceptable signal-to-background ratio (Fig. 4). The semi-automated magnetic bead-based panning allows for a physical interaction between the antibody presenting phage particles and the antigen bound nanoparticles for selection. It allows for easy in-vitro selection of specific antibodies against immobilized target antigen. This selection protocol can be optimized with different parameters such as incubation time, speed of motion, number and volume of washing step. With the reference of Fig. 3,it shows a typical ELISA result highlighting the enrichment patterns of antibody phage selection for four rounds. Pooled phage obtained from four rounds of panning against scFV library was added to ELISA well coated with the parental antigen, whereas Fig. 4 shows the typical selected positive clones from the monoclonal ELISA. Individual phage clones were chosen from three and four round pooled of phage. Several clones from each pooled phage indicate the binding of the phage to the respective antigen. Thus, the semi-automated selection protocol is more efficient panning compared to the manual method because it allows up to 96 phage display selection in one round and it can increase surface area on beads compared to plates. This method can be performed in solution using streptavidin magnetic beads coupled with an automated bead processor. Furthermore, bead-based ELISA screening can allow for the detection of antigens normally difficult to assess using conventional ELISA. 316 Angela Chiew Wen Ch’ng et al.

4 Notes

1. The target molecule for automated panning can be natural source recombinant proteins or synthetic compounds, especially different types of antibody library [12, 20, 26]. 2. There are four different techniques for the selection of antibodies: (a) Attached to the bottom of 96-well microtiter plates; (b) Attached to immunopins in a 96-well format; (c) Transferred electrophoretically to membranes from 2D-gels [28]; or (d) Attached to magnetic particles using a magnetic particle processor containing 96 magnetic pins [26, 27]. 3. Semi-automated can help to elevate the restrictions of conventional panning methods. The use of magnetic nanoparticles allows the conjugation of target proteins on the surface and separation of complexes by magnetic force [29]. The main advantage of using magnetic nanoparticles for panning is the effective binding of proteins on the surface of nanoparticles compared to a larger solid phase like microtiter plates. 4. The availability of a larger surface area increases accessibility of proteins to the phage particles for binding. Incorporating automation into the panning process ensures it to be more robust, efﬁcient, and reproducible. 5. The semi-automated protocol utilizing a magnetic particle processor has allowed successful generation of mAbs against various antigens simultaneously which is considered effective and robust [20]. 6. One of the main problems associated with the conversion from conventional methods to a semi-automated platform is the tedious preparation process required for antigen conjugation. However, the availability of various chemical and enzymatic conjugation methods has made the transition from conventional methods to semi-automated platforms easier [30, 31]. 7. Another main issue related to the use of automation for panning is the effects of cross contamination. Cross contamination could occur during the colony picking with the colony picker picking colonies from plates that are too dense. Plating of the output clones is a difﬁcult step to be automated. The bottleneck with automation is the physical restrictions of multiple antigen screening in parallel. As more antigens are screened, this would mean that more dilutions of each plate would be required making it unpractical. If the dilution is not optimum, the overwhelming growth of colonies would result in small and dense colonies for picking, increasing the risk of cross- Magnetic Nanoparticle Based Panning 317

contamination. Even so, with a proper setup the cross- contamination problem would not be a significant issue [32]. 8. The phage (rescued/amplified) can be estimated by titration 10 μL droplets of each dilution series on a 2YT-K agar plate and incubate plates top-down overnight at 37 C. 9. The successful enrichment, the titter of eluted phage usually is 103–105 phage per well after the first panning round and increases two to three orders in magnitude per each additional panning round. The phage preparation after re-amplification of the eluted phage has a titer of about 1012–1014 phage/mL. The number of colonies is counted in the droplets on all the plates, and calculate from these the colony-forming units using the formula: C.F.U.: 1⁄ 4 number of colonies dilution factor 100. On average, phage preparations in microtiter plates (200 μL culture volume) produce 1010–1011c.f.u. Compare the c.f.u. values obtained on 2YT-AG and 2YT-K agar plates for each phage library. The helperphage genome containing population should be a minimum of 4–5 orders of magnitude smaller than the antibody fragment containing phagemid population. 10. Store some of the eluted phage after each round of panning at À80 C for future reference where there is any repeat on panning for specific panning round required. This avoids starting from the beginning for time wise. 11. Filter tips should be used throughout all the experiment involving phage particle which will contaminate the pipette. 12. All reagents, buffers, and solutions are aliquot in a small tube and discarded after use to prevent cross-contaminant during the experiment. 13. It is recommended to spot blank TG1 and PBS on 2YT þ Kan and 2YT þ amp plates in each round of panning to check for cross contamination. 14. For the ease of use and avoiding pipetting errors, use an eight- channel micropipette. 15. Washing with PBST is a must step to remove unspecific binders (antibody). The wash procedure increases by panning round. The automated washing procedure reduces human error as skill mastered varied most of the time. 16. The substrates used to detect positive clones which bind specifically to the antigens. In this case, we use Horseradish peroxidase substrates that are light sensitive. Thus, it will turn green if ABTS is used, the intensity of the substrates determined the binding affinities between both antigens and its binders. The intensities of the substrates can be quantified by using ELISA plate reader at wavelength of 405 nm. Other than 318 Angela Chiew Wen Ch’ng et al.

ABTS, TMB (3, 30,5,50-Tetramethylbenzidine) can be used as substrates in detection. 17. Loading the TG1 culture to released beads can be done manually without the Kingﬁsher 96 during infection.

Acknowledgment

The authors would like to acknowledge the support of USM Research University Individual grant (1001/CIPPM/812173) and Malaysian Ministry of Education through the Higher Institu- tion Centre of Excellence (HICoE) Grant (Grant No: 311/CIPPM/44001005).

References

1. Chin CF, Ler LW, Choong YS, Ong EBB, 10. Lim BN, Chin CF, Choong YS, Ismail A, Lim Ismail A, Tye GJ, Lim TS (2016) Application TS (2016) Generation of a naı¨ve human single of streptavidin mass spectrometric immunoas- chain variable fragment (scFv) library for the say tips for immunoaffinity based antibody identification of monoclonal scFv against Sal- phage display panning. J Microbiol Methods monella Typhi Hemolysin E antigen. Toxicon 120:6–14. https://doi.org/10.1016/j.mimet. 117:94–101. https://doi.org/10.1016/j. 2015.11.007 toxicon.2016.04.032 2. Cahill DJ (2001) Protein and antibody arrays 11. Rahumatullah A, Ahmad A, Noordin R, Lim and their medical applications. J Immunol TS (2015) Delineation of BmSXP antibody Methods 250(1):81–91 V-gene usage from a lymphatic filariasis based 3. Marx U, Embleton MJ, Fischer R, Gruber FP, immune scFv antibody library. Mol Immunol Hansson U, Heuer J, De Leeuw WA, 67(2, Part B):512–523. https://doi.org/10. Logtenberg T, Merz W, Portetelle D (1997) 1016/j.molimm.2015.07.040 Monoclonal antibody production. ATLA Not- 12. Hanes J, Schaffitzel C, Knappik A, Pluckthun€ A tingham 25:121–138 (2000) Picomolar affinity antibodies from a 4. Ma JK, Drake PM, Christou P (2003) The fully synthetic naive library selected and production of recombinant pharmaceutical evolved by ribosome display. Nat Biotechnol proteins in plants. Nat Rev Genet 4 18(12):1287–1292 (10):794–805 13. Rothe C, Urlinger S, Lo¨hning C, Prassler J, 5. Azzazy HM, Highsmith WE (2002) Phage dis- Stark Y, J€ager U, Hubner B, Bardroff M, play technology: clinical applications and Pradel I, Boss M (2008) The human combina- recent innovations. Clin Biochem 35 torial antibody library HuCAL GOLD com- (6):425–445 bines diversification of all six CDRs according 6. Hoogenboom HR (2002) Overview of anti- to the natural immune system with a novel body phage-display technology and its applica- display method for efficient selection of high- tions. Methods Mol Biol 178:1–37 affinity antibodies. J Mol Biol 376 (4):1182–1200 7. Kehoe JW, Kay BK (2005) Filamentous phage 14. Knappik A, Ge L, Honegger A, Pack P, display in the new millennium. Chem Rev 105 ` (11):4056–4072 Fischer M, Wellnhofer G, Hoess A, WoElle J, Pluckthun€ A, Virnek€as B (2000) Fully synthetic 8. Ling MM (2003) Large antibody display human combinatorial antibody libraries libraries for isolation of high-affinity antibo- (HuCAL) based on modular consensus frame- dies. Comb Chem High Throughput Screen works and CDRs randomized with trinucleo- 6(5):421–432 tides. J Mol Biol 296(1):57–86 9. Bahara NHH, Chin ST, Choong YS, Lim TS 15. Lee CV, Liang W-C, Dennis MS, Eigenbrot C, (2016) Construction of a semisynthetic human Sidhu SS, Fuh G (2004) High-affinity human VH single-domain antibody library and selec- α antibodies from phage-displayed synthetic Fab tion of domain antibodies against -crystalline libraries with a single framework scaffold. J Mol of Mycobacterium tuberculosis. J Biomol Biol 340(5):1073–1093 Screen 21(1):35–43 Magnetic Nanoparticle Based Panning 319

16. Sidhu SS, Li B, Chen Y, Fellouse FA, 24. Barbas CF, Kang AS, Lerner RA, Benkovic SJ Eigenbrot C, Fuh G (2004) Phage-displayed (1991) Assembly of combinatorial antibody antibody libraries of synthetic heavy chain com- libraries on phage surfaces: the gene III site. plementarity determining regions. J Mol Biol Proc Natl Acad Sci 88(18):7978–7982 338(2):299–310 25. Ward R, Clark M, Lees J, Hawkins N (1996) 17. Strachan G, McElhiney J, Drever M, Retrieval of human antibodies from phage- McIntosh F, Lawton L, Porter A (2002) display libraries using enzymatic cleavage. J Rapid selection of anti-hapten antibodies Immunol Methods 189(1):73–82 isolated from synthetic and semi-synthetic anti- 26. Hallborn J, Carlsson R (2002) Automated body phage display libraries expressed in screening procedure for high-throughput gen- Escherichia coli. FEMS Microbiol Lett 210 eration of antibody fragments. BioTechniques (2):257–261 33:30–37 18. Lim BN, Tye GJ, Choong YS, Ong EBB, 27. Walter G, Konthur Z, Lehrach H (2001) High- Ismail A, Lim TS (2014) Principles and appli- throughput screening of surface displayed gene cation of antibody libraries for infectious products. Comb Chem High Throughput diseases. Biotechnol Lett 36(12):2381–2392. Screen 4(2):193–205. https://doi.org/10. https://doi.org/10.1007/s10529-014- 2174/1386207013331228 1635-x 28. Liu B, Huang L, Sihlbom C, Burlingame A, 19. Winter G, Griffiths AD, Hawkins RE, Hoogen- Marks JD (2002) Towards proteome-wide pro- boom HR (1994) Making antibodies by phage duction of monoclonal antibody by phage dis- display technology. Annu Rev Immunol 12 play. J Mol Biol 315(5):1063–1073 (1):433–455. https://doi.org/10.1146/ 29. Schwenk JM, Lindberg J, Sundberg M, annurev.iy.12.040194.002245 Uhleń M, Nilsson P (2007) Determination of 20. Konthur Z, Wilde J, Lim TS (2010) Semi- binding specificities in highly multiplexed automated magnetic bead-based antibody bead-based assays for antibody proteomics. selection from phage display libraries. Antibody Mol Cell Proteomics 6(1):125–132 Eng 2010:267–287 30. Behrens CR, Liu B (2014) Methods for site- 21. Hamilton S (2002) Introduction to screening specific drug conjugation to antibodies. MAbs automation. High throughput screening. 1:46–53 Methods Mol Biol 190:169–189 31. Ta HT, Peter K, Hagemeyer CE (2012) Enzy- 22. Kala M, Bajaj K, Sinha S (1997) Magnetic bead matic antibody tagging: toward a universal bio- enzyme-linked immunosorbent assay (ELISA) compatible targeting tool. Trends Cardiovasc detects antigen-specific binding by phage- Med 22(4):105–111 displayed scFv antibodies that are not detected 32. Turunen L, Takkinen K, So¨derlund H, Pulli T with conventional ELISA. Anal Biochem 254 (2009) Automated panning and screening pro- (2):263–266 cedure on microplates for antibody generation 23. Konthur Z, Walter G (2002) Automation of from phage display libraries. J Biomol Screen. phage display for high-throughput antibody https://doi.org/10.1177/ development. Targets 1(1):30–36 1087057108330113 Chapter 17

Phage Display and Selections on Cells Wieland Fahr and Andre´ Frenzel

Abstract

Antibody identification by phage display on protein or peptide targets is well established and many protocols are available. But there are many targets that cannot be expressed recombinantly or, like peptides, do not reflect correct folding of the protein. Most of these targets are cell surface receptors. Here, we describe a protocol for a panning strategy on cells to obtain specific binders to cell surface receptors. A depletion step is included to prevent enrichment of antibodies that bind to unwanted targets. Each step of the protocol is explained and variations of this protocol are given.

Key words Phage display, Antibody engineering, Human antibodies, Recombinant antibodies, scFv, Therapeutic antibodies, Flow cytometry, Conformational epitope

1 Introduction

Currently, several methods exist for the generation and identiﬁca- tion of monoclonal antibodies: The well-known hybridoma technology [1] can be used with human cells or humanized mice [2, 3] to directly generate human antibodies without the necessity of laborious and time-consuming humanization [4]. The major challenge using mice or rats containing a human immunoglobulin repertoire is that the animals have to be immunized in order to develop antibodies. For some targets, an immune response may not be developed, e.g., if the antigen has high homology to a mouse protein or if human-mouse cross-species reactivity is desired. Anti- gens that are toxic or lead to severe disease of the animal may also be problematic. Here, in vitro technologies such as ribosomal display [5] or antibody phage display [6–8] may provide a solution as they do not depend on an intact immune system of a living organism, and are insusceptible to toxic substances. Although it is possible to use DNA immunization or immunization with whole cells [9, 10], in vitro technologies also allow better adjustment of the parameters during the selection process.

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_17, © Springer Science+Business Media LLC 2018 321 322 Wieland Fahr and Andre´ Frenzel

Antibody selection on cells (either by screening of hybridoma clones derived from cell-immunized mice or by using in vitro technologies) is preferred in case of cell-surface receptors such as G-protein coupled receptors, as antibodies that are generated against recombinant antigens or peptides representing extracellular loops may not be able to bind the native protein on the surface of the target cells. Several methods have been established to enrich and identify antibody fragments such as single chain variables (scFv) or peptides from naive or immune antibody gene libraries [11–14] on cells. Other possibilities of identifying binders to native cell surface receptors have been described by using phage display and panning on cancer tissue [15]. Major challenge during the selection of a phage display gene library on cells (or tissue) is the high number of different receptors and proteins that are expressed on the cell surface. Therefore, the depletion of the library with a negative cell line is preferred to remove phage particles that bind to these unwanted targets [14, 16]. The depleted library can then be used for the selection step on antigen positive cells. It is obvious that a high expression level of the target antigen is preferred and that the cell line that is used for the “negative selection” should not contain the target antigen but should express a very similar portfolio of receptors on the cell surface. Therefore, overexpression of the receptor on the cell surface of HEK or CHO cells can be used [16], since non-transfected cells can be used for library depletion and transiently expressing cells usually have a very high expression level of the transgene. Alternatively, receptor expressing cell lines can be used where the gene of interest has been knocked out, e.g., by CRISPR/Cas9. Of course, expression level of the target should be checked beforehand, either by flow cytometry using a commercial antibody (if available) or at least by real-time PCR. The protocol described in this chapter can be used as a starting point for selecting scFv antibodies that bind to cell surface receptors. An initial depletion step is included to remove scFv-phage that bind to unwanted targets or epitopes. After depletion, the selection step follows for the target molecule. Using this strategy, even complex, multimeric, fully functional cell membrane proteins can be targeted without alteration of the structure due to purification or immobilization. Furthermore, the functional properties of the receptor remain untouched, enabling also the change of the physi- ological status of the receptor, e.g., by the addition or removal of a special ligand or co-receptor that binds to the receptor under certain circumstances. The phage particles enriched under these modified conditions may be able to recognize (non-)stimulated cells or epitopes that are exposed (or hidden) by the binding of a ligand or under special salt conditions. Phage Display and Selections on Cells 323

The antibody phage libraries exemplarily used in this protocol are Human/Hust Antigen Libraries HAL 7/8 and HAL 9/10, which were described before [17, 18].

2 Materials

1. 2Â YT media, pH 7.0: 1.6% (w/v) tryptone, 1% (w/v) yeast extract, 0.5% (w/v) NaCl. 2. 2Â YT-GA: 2Â YT, containing 100 μg/mL ampicillin, 100 mM glucose. 3. Escherichia Coli XL1-Blue MRF0. 4. M13KO7 helper phage. 5. 2Â YT-AK: 2Â YT, containing 100 μg/mL ampicillin, 50 μg/ mL kanamycin. 6. PEG/NaCl solution: 20% (w/v) polyethylene glycol 6000, 2.5 M NaCl. 7. Phage Dilution Buffer: 10 mM Tris–HCl pH 7.5, 20 mM NaCl, 2 mM EDTA. 8. Cells for depletion of nonspecific antigens. 9. Cells for specific selection of phage antibody libraries. 10. Cell culture media (cell type specific). 11. Trypsin–EDTA: 0.025% (w/v) Trypsin and 1 mM EDTA in PBS, cell culture grade. 12. Fetal Bovine Serum (FBS). 13. 0.1% (w/v) trypan blue solution.

14. Phosphate-buffered saline (PBS), pH 7.4: 4 mM KH2PO4, 16 mM Na2HPO4, 115 mM NaCl. 15. MPBS: 2% Milk in PBS. 16. 10 μg/mL trypsin in PBS (stock solution: trypsin type XIII from bovine pancreas, made up in 50 mM Tris–HCl pH 7.4, 1 mM CaCl2 and stored at À20 C). 17. Escherichia coli TG1. 18. PP Deepwell 96-well multititer plates. 19. 10Â GA: 1 M Glucose, 1 mg/mL Ampicillin in 2Â YT medium. 20. 2Â YT-T: 2Â YT, containing 20 μg/mL Tetracyclin. 21. 10 cm Petri dishes. 22. 2Â YT-GA agar plates: 2Â YT-GA, 1.2% (w/v) agar. 23. Sterile 96-well U-bottom microtiter plates made for bacterial culture. 324 Wieland Fahr and Andre´ Frenzel

24. Glycerol. 25. Potassium Phosphate Buffer: 2.31% (w/v) (0.17 M) KH2PO4 þ 12.54% (w/v) (0.72 M) K2HPO4. 26. Buffered 2Â YT-SA: 2Â YT, containing 10% (v/v) Potassium Phosphate Buffer, 50 mM Saccharose þ 100 μg/mL Ampicillin.

27. IPTG (Stock 0.84 M ¼ 20% (w/v) IPTG in dH2O). 28. 96-Well V-bottom microtiter plates. 29. FC-Buffer: 1% (w/v) BSA in PBS (bovine serum albumin). 30. Fluorescently conjugated secondary monoclonal antibody, e.g., 9E10 anti-mycTag FITC-conjugated. 31. Flow cytometer sheath ﬂuid (according to the manufacturer’s description).

3 Methods

3.1 Production 1. Inoculate 2Â YT-GA medium with E. coli XL1-Blue MRF0, of Phage Antibodies pre-infected with a representative aliquot of your library. Use 10- to 100-fold more bacteria than your library diversity. An overnight culture is advised. 2. Inoculate 50 mL 2Â YT-GA with your overnight culture (resulting OD600 < 0.1). Grow until OD600 reaches 0.5 at 37 C, 250 rpm. 3. Infect 2 mL of culture (~109 bacteria) with helper phage (M13KO7) to reach a ratio of 10–20 helper phage per cell. Incubate for 30 min at 37 C without shaking, and subsequently for 30 min at 37 C, 250 rpm. 4. Spin for 10 min at 3000 Â g, RT. Remove the supernatant completely (residual glucose will inhibit phage production). 5. Resuspend pellet in 30 mL 2Â YT-AK. 6. Grow overnight at 30 C, 250 rpm. 7. Centrifuge bacterial culture in a 50 mL tube for 15 min at 3000 Â g,4C. Transfer the supernatant into a fresh 50 mL tube. 8. Precipitate phage particles by adding 1/5 volume of PEG/- NaCl. Mix by inversion and incubate for 1 h on ice. 9. Centrifuge for 1 h at 3000 Â g,4C and thoroughly discard the supernatant (turn tube upside-down on paper towels). 10. Resuspend the pellet in 300 μL phage dilution buffer and transfer to a fresh 2 mL Eppendorf tube. 11. Centrifuge for 1 min at 16,000 Â g to remove bacterial residue. 12. Transfer the supernatant into a fresh 2 mL screw-cap tube. Phage Display and Selections on Cells 325

13. Phage can be stored at 4 C.

3.2 Preparation 1. Depletion has to be performed on cells devoid of the speciﬁc of Cells antigen, and selection on cells displaying the antigen (in the best situation: the same cell line, transfected or not with the antigen). 2. Adherent cells are enzymatically detached with Trypsin–EDTA solution to get a single-cell suspension. Keep trypsin incubation as short as possible. Add medium containing 10% (v/v) FBS to inhibit trypsin and to prevent further proteolytic degradation of surface molecules. Alternatively, use a cell scraper to detach cells from the cell culture ﬂask. 3. Count cells (the vitality of the cells can be determined by trypan blue exclusion staining). 4. Centrifuge cell suspension for 5 min at 300 Â g,4C(see Note 1). 5. Resuspend cells in 10 mL of cold PBS. 6. Centrifuge for 5 min at 300 Â g,4C. 7. Use 1–5 Â 107 cells of each cell type for next step.

3.3 Depletion 1. Use aliquots of your antibody phage library. The number of of Nonspeciﬁc Phage phage should exceed the diversity of the library by 100-fold. Incubate phage in 1 mL 2% MPBS for saturation on a rotator for 1 h at 4 C. This corresponds to the input of your selection. 2. Saturate antigen-negative cells by incubation with 5 mL 2% MPBS on a rotator for 1 h at 4 C(see Note 1). 3. Centrifuge cell suspension for 5 min at 300 Â g,4C. Discard the supernatant and add phage solution to the cells for depletion. Incubate on a rotator for 2 h at 4 C(see Note 1). 4. Pellet cell-phage suspension for 5 min at 300 Â g,4C and transfer the supernatant to a new tube.

3.4 Selection 1. Saturate antigen-positive cells by incubation with 5 mL 2% of Speciﬁc Phage MPBS on a rotator for 1 h at 4 C(see Note 1). 2. Centrifuge cell suspension for 5 min at 300 Â g,4C and add the depleted phage library to the cells (supernatant of the depletion). Incubate on a rotator for 2 h at 4 C(see Note 2 for the selection of internalizing antibodies; see Note 3 for a more stringent selection procedure). 3. Pellet cell-phage suspension for 5 min at 300 Â g,4C and remove the supernatant. Wash cells with 1 mL of PBS. 4. Centrifuge for 5 min at 300 Â g,4C. 326 Wieland Fahr and Andre´ Frenzel

5. Repeat this washing procedure (steps 3 and 4) ten times. (Intensity of selection can be adjusted, e.g., through number of washing steps or by adding a competitor: see Note 3).

3.5 Elution 1. Elute phage by resuspending cells in 500 μLof10μg/mL of Speciﬁc Phage trypsin in PBS on a rotator for 30 min at RT. (You can add DNase if your elution is too viscous.) 2. Add 500 μL of PBS to have a ﬁnal volume of 1 mL, corresponding to the output of your selection. 3. If necessary, input and output phage can be stored at 4 C for up to 4 weeks. 4. If subsequent panning rounds are planned, proceed with Sub- headings 3.6 and 3.7. After the last panning round, proceed with Subheadings 3.7 and 3.8.

3.6 Infection of TG1- 1. Keep 10 μL of the eluted phage for titration. Inoculate 50 mL Tr E. coli 2Â YT medium with E. coli TG1, incubate until OD600 ¼ 0.5 at with the Selected 37 C, 250 rpm (TG1 can be stored on ice until infection, since Phage (for Next F-pili are stable at 4 C). Panning Round) 2. Pour 150 μL of eluted phage into a Polypropylene deep-well plate.

3. Add 150 μL of TG1 at OD600 ¼ 0.5. 4. Incubate for 30 min at 37 C without shaking, and subsequently for 30 min at 37 C, 650 rpm. 5. Add 1000 μL2Â YT medium and 150 μL10Â GA. Incubate for 1 h at 37 C, 650 rpm (OD600 should reach 0.4–0.5 at the end of the incubation step, equivalent to 5 Â 108 cells). 6. Infect cells with M13KO7 helper phage in a 20-fold surplus (e.g., 1010 phage for 5 Â 108 bacterial cells). 7. Incubate for 30 min at 37 C without shaking, and subsequently for 30 min at 37 C, 650 rpm. 8. Centrifuge for 10 min at 3000 Â g, RT. Remove the supernatant completely (be careful not to harm the pellet). 9. Add 950 μL2Â YT-AK. Incubate overnight at 30 C, 650 rpm. 10. Pellet bacteria for 10 min at 3000 Â g,RT. 11. Transfer the supernatant into a fresh tube and use for next panning round (proceed from Subheading 3.2).

3.7 Phage Titration 1. Inoculate 5 mL 2Â YT-T with XL1-Blue MRF0 and incubate overnight at 37 C, 250 rpm. 2. Inoculate 50 mL 2Â YT-T with 500 μL of the overnight culture. Incubate at 37 C, 250 rpm until OD600 ¼ 0.5 is reached. Phage Display and Selections on Cells 327

3. Add 5 μL of your input and your output to 495 μL of PDB or À PBS. This is the 10 2 dilution of your phage. Make serial À dilution of your phage until 10 12 for the input and until À 10 8 for the output. 4. Use 10 μL of the dilutions to infect 50 μL of XL1-Blue at OD600 ¼ 0.5 in an Eppendorf tube. 5. Incubate without shaking at 37 C for 30 min. 6. Plate each dilution of the bacteria suspension on 2Â YT-GA agar plates (10 cm Petri dish). 7. Grow overnight at 37 C. 8. Count the colonies and calculate the cfu or cfu/mL titer according to the dilution.

3.8 Preparation 1. Fill each well of a 96-well U-bottom polypropylene microtiter of Master Plate plate with 150 μLof2Â YT-GA. 2. Pick 92 clones with sterile tips from the plate created in Sub- heading 3.7 and inoculate each well. Also inoculate two wells with a positive control (e.g., XL1-Blue::pHAL14-D1.3 anti- lysozyme scFv). Seal the plate with a breathable sealing ﬁlm. Keep two wells without clones as negative control. 3. Incubate in a microtiter plate shaker overnight at 37 C, 850 rpm. The next day, use 10 μL of supernatant for subsequent steps. 4. Add glycerol solution to the overnight culture to have a ﬁnal concentration of 15–30% glycerol. Mix and store the master plate at À80 C.

3.9 Production The pHAL vector design places a myc-tag and an amber stop codon of Soluble Antibody between the scFv antibody and the M13 pIII gene, allowing soluble in 96-Well Microtiter antibody production and detection of the antibody in subsequent Plates detection experiments. 1. Fill each well of a 96-well U-bottom polypropylene microtiter plate with 180 μLof2Â YT-GA. 2. Transfer 10 μL of each overnight culture into the corresponding well of the fresh plate. 3. Incubate for 2 h at 37 C, 850 rpm. 4. Pellet cultures for 10 min at 3000 Â g, RT. Remove the supernatant by turning over the plate and carefully beating out the liquid (alternatively, remove the supernatant carefully by pipetting). 5. Resuspend the pellet in 180 μL buffered 2Â YT-SA containing 50 μM IPTG (IPTG will induce expression of the pHAL lac promoter). 328 Wieland Fahr and Andre´ Frenzel

6. Incubate overnight at 30 C, 850 rpm. 7. Pellet cultures for 10 min at 3000 Â g, RT. Transfer the supernatant into fresh plates, use for ﬂow cytometry or other analytical methods.

3.10 Analysis 1. Adherent cells are enzymatically detached with Trypsin–EDTA of Soluble Antibodies solution to get a single-cell suspension. Keep trypsin incuba- by Flow Cytometry tion as short as possible. Add medium containing 10% (v/v) FBS to inhibit trypsin and to prevent further proteolytic degradation of surface molecules (alternatively, use a cell scraper to dispatch cells from the cell culture ﬂask and pipet to generate a single-cell suspension). 2. Count the cells (the vitality of the cells can be determined by trypan blue exclusion staining). 3. Centrifuge cell suspension for 5 min at 300 Â g,4C(see Note 1). 4. Discard the supernatant completely. 5. Resuspend the cells in cold PBS at 2 Â 106 cells/mL and transfer the cells into a V-bottom 96-well plate, 100 μL per well. 6. Centrifuge the cells for 5 min at 300 Â g at 4 C. 7. Discard the supernatant completely (plate should be poured out immediately after centrifugation by turning the microtiter plate head-over and discard the supernatant with one push). 8. Put microtiter plate on ice and resuspend the cells in 30 μL per well of FC-Buffer and 30 μL of the soluble antibodies generated in Subheading 3.9, for 1 h at 4 C with gentle mixing (see Note 1). 9. Wash the cells two times with 150 μL/well PBS (add PBS, mix cells, spin down, and discard supernatant). 10. Put microtiter plate on ice and resuspend the cells in 50 μL per well of secondary-ﬂuorescent monoclonal antibody for 30 min at 4 C with gentle mixing (e.g., 9E10 antibody, FITC- coupled). 11. Wash cells two times with 150 μL/well PBS (add PBS, mix cells, spin down, and discard the supernatant). 12. Resuspend the cells in PBS, 150 μL/well. 13. For measurement by cytometry and data analysis, follow the manufacturer’s protocols. Phage Display and Selections on Cells 329

4 Notes

1. Internalization of antigen: to avoid the internalization of your target antigen and thus its loss during the depletion, selection, or screening, it is essential to perform all the procedures involving cells at 4 C. Prolonged incubation of cells at room temperature also causes cell death, causing unspecific interaction with all antibodies and thus contaminating the flow cytometry signal. 2. Selection of internalizing antibodies: The ability of bacteriophage to undergo receptor-mediated endocytosis indicates that phage libraries might be selected not only for cell binding but also for internalization into mammalian cells. This approach would be useful for generating ligands, which could deliver drugs into a cell for therapeutic applications. Note that an internalized antibody will not be detected by flow cytometric analysis. Therefore, the antibody-phage construct has to be fluorescently labeled before applying to the cells. For the selection, the cells are adherent in flask. Incubation of the cells with phage has to be performed at 4 C for 2 h. Subsequently, the cells are incubated for 30 min at 37 Cina 5% CO2 gassed incubator to allow internalization of the receptor. After incubation, the cells are washed three times with buffer to remove noninternalized binders (100 mM glycine, 150 mM NaCl, pH 2.5). Finally, the cells are detached from the flask for the remaining procedure. All other steps are equal to those described before. 3. Competitive elution: Competitive elution with a second mAb or a ligand can allow selecting antibodies directly against the desired epitope or antibodies that are competing with a known ligand of the receptor. Generally, 100 μM (if possible) of the competitive molecule for elution is incubated with the cells for 2 h on ice. It is essential to note that the eluted phage-antibodies have an affinity lower than the affinity of the molecule used for the elution. In order to select antibodies with enhanced affinity, you have to perform additional rounds of selection, but it is unlikely to select antibodies with a better affinity than the one of the competing molecule.

References

1. Ko¨hler G, Milstein C (1975) Continuous cul- their immunoglobulin genes generate antibo- tures of fused cells secreting antibody of pre- dies as efficiently as normal mice. Proc Natl defined specificity. Nature 256:495–497 Acad Sci U S A 111:5153–5158. https://doi. 2. Murphy AJ, Macdonald LE, Stevens S et al org/10.1073/pnas.1324022111 (2014) Mice with megabase humanization of 330 Wieland Fahr and Andre´ Frenzel

3. Macdonald LE, Karow M, Stevens S et al selection using an integrated vector system. (2014) Precise and in situ genetic humaniza- BMC Biotechnol 12:62. https://doi.org/10. tion of 6 Mb of mouse immunoglobulin genes. 1186/1472-6750-12-62 Proc Natl Acad Sci U S A 111:5147–5152. 12. Cao J, Zhao P, Miao XH et al (2003) Phage https://doi.org/10.1073/pnas.1323896111 display selection on whole cells yields a small 4. Bruggemann€ M, Osborn MJ, Ma B et al peptide specific for HCV receptor human (2015) Human antibody production in trans- CD81. Cell Res 13:473–479. https://doi. genic animals. Arch Immunol Ther Exp org/10.1038/sj.cr.7290190 63:101–108. https://doi.org/10.1007/ 13. Shukla GS, Krag DN (2005) Phage display s00005-014-0322-x selection for cell-specific ligands: development 5. Hanes J, Jermutus L, Weber-Bornhauser S et al of a screening procedure suitable for small (1998) Ribosome display efficiently selects and tumor specimens. J Drug Target 13:7–18. evolves high-affinity antibodies in vitro from https://doi.org/10.1080/ immune libraries. Proc Natl Acad Sci 10611860400020464 95:14130–14135. https://doi.org/10.1073/ 14. Eisenhardt SU, Schwarz M, Bassler N, Peter K pnas.95.24.14130 (2007) Subtractive single-chain antibody 6. Breitling F, Dubel€ S, Seehaus T et al (1991) A (scFv) phage-display: tailoring phage-display surface expression vector for antibody screen- for high specificity against function-specific ing. Gene 104:147–153 conformations of cell membrane molecules. 7. McCafferty J, Griffiths AD, Winter G, Chiswell Nat Protoc 2:3063–3073. https://doi.org/ DJ (1990) Phage antibodies: filamentous 10.1038/nprot.2007.455 phage displaying antibody variable domains. 15. Larsen SA, Meldgaard T, Fridriksdottir AJR Nature 348:552–554 et al (2016) Raising an antibody specific to 8. Hoogenboom HR, Winter G (1992) breast cancer subpopulations using phage dis- By-passing immunisation. Human antibodies play on tissue sections. Cancer Genomics Pro- from synthetic repertoires of germline VH teomics 13:21–30 gene segments rearranged in vitro. J Mol Biol 16. Jones ML, Alfaleh MA, Kumble S et al (2016) 227:381–388 Targeting membrane proteins for antibody dis- 9. Funahashi S-I, Suzuki Y, Nakano K et al (2017) covery using phage display. Sci Rep 6:26240. Generation and characterization of monoclo- https://doi.org/10.1038/srep26240 nal antibodies against human LGR6. J Bio- 17. Hust M, Meyer T et al (2011) A human scFv chem (Tokyo) 161(4):361–368. https://doi. antibody generation pipeline for proteome org/10.1093/jb/mvw077 research. J Biotechnol 152(4):159–170. 10. Khademi F, Mostafaie A, Parvaneh S et al https://doi.org/10.1016/j.jbiotec.2010.09. (2017) Construction and characterization of 945 monoclonal antibodies against the extracellular 18. Kugler€ J, Wilke S et al (2015) Generation and domain of B-lymphocyte antigen CD20 using analysis of the improved human HAL9/10 DNA immunization method. Int Immuno- antibody phage display libraries. BMC Biotech- pharmacol 43:23–32. https://doi.org/10. nol 15:10. https://doi.org/10.1186/s12896- 1016/j.intimp.2016.11.035 015-0125-0 11. Yoon H, Song JM, Ryu CJ et al (2012) An efficient strategy for cell-based antibody library Chapter 18

Combine Phage Antibody Display Library Selection on Patient Tissue Specimens with Laser Capture Microdissection to Identify Novel Human Antibodies Targeting Clinically Relevant Tumor Antigens

Yang Su, Scott Bidlingmaier, Nam-Kyung Lee, and Bin Liu

Abstract

A functional approach to generate tumor-targeting human monoclonal antibodies is through selection of phage antibody display libraries directly on tumor cells. Although technically convenient, the use of cancer cell lines for the selection has limitations as those cell lines often undergo genetic and epigenetic changes during prolonged in vitro culture and alter their cell surface antigen expression proﬁle. The key is to develop a technology that allows selection of phage antibody display libraries on tumor cells in situ residing in their natural tissue microenvironment. Laser capture microdissection (LCM) permits the precise procurement of tumor cells from human cancer patient tissue sections. Here, we describe a LCM-based method for selecting phage antibodies against tumor cells in situ using both fresh frozen and parafﬁn-embedded tissues. To restrict the selection to antibodies that bind internalizing epitopes, the method utilizes a polyclonal phage population pre-enriched for internalizing phage antibodies. The ability to recognize tumor cells in situ residing in their natural tissue microenvironment and to deliver payload intracellularly makes these LCM-selected antibodies attractive candidates for the development of targeted cancer therapeutics.

Key words Laser capture microdissection, Phage antibody library, Internalizing human monoclonal antibody, Macropinocytosis, Solid tumor, Natural cell surface epitope, Tissue microenvironment, Human cancer specimen, Targeted therapy, Intracellular payload delivery

1 Introduction

Tumor cell surface antigens are excellent targets for antibody-based therapy development. Identification of novel tumor-specific or tumor-associated cell surface antigens can result in significant improvements in detection and treatment of malignant tumors. The antigenic epitope space at the tumor cell surface is highly complex and consists of extensive posttranslational modifications [1–5]. Monoclonal antibodies (mAbs) can recognize with high affinity and specificity a wide range of antigenic determinants and

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_18, © Springer Science+Business Media LLC 2018 331 332 Yang Su et al.

discern subtle differences in antigen structure and conformation, which can be used to effectively map the tumor cell surface epitope space independent of gene expression analysis [6–14]. Phage antibody display technology has been widely used to develop novel human monoclonal antibodies [15–26]. Phage antibody display libraries serve as a source of random shape repertoire that can be used to probe neoplastic alterations on cancer cell surface [6–8]. Selecting phage antibody libraries directly on cancer cell lines has enabled the identification of tumor-targeting antibodies without prior knowledge of target antigens [6, 7, 13, 27]. However, when tumor cells are removed from their natural environment, they undergo genetic and epigenetic changes yielding different surface antigens than those seen in actual cases of cancer. Laser capture microdissection (LCM) under direct microscopic visualization has allowed small clusters of tumor cells to be isolated and removed from heterogeneous human tissue sections [28–34]. This technology, when combined with either phage peptide display [35–37]or phage antibody display [7, 38], permits the selection of phage binding specifically to tumor cells in their native tissue environment. We have previously developed the LCM-based phage antibody display library selection technique and used it to select non-immune human antibody phage display libraries on cancer patient tissues to identify novel human antibodies targeting clinically relevant tumor epitopes [7]. To further identify internalizing antibodies for tumor-specific intracellular payload delivery, we generated sublibraries enriched for internalizing phage antibodies as input for LCM-based selection [7]. We hereby describe this LCM selection protocol for the identification and characterization of tumor-specific internalizing phage antibodies (Fig. 1). Detailed methods are provided for phage sublibrary construction, antibody selection, and identification on tissue specimen using LCM, validation of internalization with single-chain variable fragment antibodies (scFvs), and tumor-targeted intracellular payload delivery.

2 Materials

2.1 Sublibrary 1. Cell growth medium: DMEM supplemented with 10% fetal Construction bovine serum and 100 μg/ml penicillin-streptomycin. 2. Phosphate-buffered saline (PBS), pH 7.4. 3. 0.25% trypsin/EDTA. 4. Glycine wash buffer: 100 mM glycine, pH 2.8, 150 mM NaCl. 5. Phage elution buffer: 100 mM trimethylamine (TEA). 6. Neutralizing buffer: 1 M Tris–HCl, pH 6.8. 7. Escherichia coli TG1 (Lucigen, Middleton, WI, USA). 8. 2ÂYT (1 L): 16 g tryptone, 10 g yeast extract, 5 g NaCl. LCM-Based Phage Antibody Library selection on Cancer Patient Tissues 333

Fig. 1 Outline of the LCM-based method that allows selection of phage antibody display library on cancer patient specimen to identify novel antibodies targeting tumor cells in situ. Non-immune phage antibody display libraries were counter-selected on a panel of normal cells to remove binders to normal cell surface molecules, then selected on cancer cell lines and/or primary tumor cells to enrich for tumor specificity. When the selection is performed under internalizing conditions, a sublibrary enriched for tumor-specific internalizing antibodies is created and used as input for LCM-based selection on patient tissue specimen. Following incubation with slides containing sectioned tumor tissues, a small cluster of tumor cells along with tumor- bound phages were procured by LCM and collected on the cap of a PCR tube (step 1). Genes of scFv-coding regions were amplified by PCR (step 2) and spliced into a phage display vector to create secondary libraries (step 3) that were used for screening (step 4) or additional rounds of LCM-based selection (step 5). Ctr control, MW molecular weight (Adopted from a figure that was originally published in reference [7] in Mol Cell Proteomics)

9. Bacteria/phage growth media: 2ÂYT, 50ug/ml tetracycline. 10. Bacteria/phage plate: YT, 15 μg/ml tetracycline. 11. PEG8000/NaCl solution (5Â): 20% (w/v) polyethylene glycol 8000, 2.5 M NaCl in PBS. 12. 0.45 μm sterile syringe ﬁlter (Corning, New York, USA). 13. Phage or bacteria storage buffer: 25% glycerol (v/v) in PBS, pH 7.4. 334 Yang Su et al.

2.2 Selection 1. Leica Membrane Slides (MicroDissect, Mittenaar, Germany). of Phage Antibodies 2. Fixation buffer 1: acetone. Targeting Tumor Cells 3. Fixation buffer 2: 4% paraformaldehyde (PFA). In Situ by LCM 4. Blocking buffer: 3% H2O2 in PBS. 5. Hematoxylin (H-3401, Vector Laboratories, Burlingame, CA, USA). 6. Microcentrifuge PCR tubes (BioExpress, Kaysville, UT, USA). 7. Leica AS LMD (Leica Microsystems GmbH, Wetzlar, Germany).

2.3 PCR Recovery 1. PCR kit (Lucigen). of scFv Genes from 2. Primers for recovery and transfer of scFv gene to fd phage LCM-Procured Tissue display vector: Pieces Fd2 (TTTTTGGAGATTTTCAAC). Fdseq (GAATTTTCTGTATGAGG). 3. QIAquick PCR purification kit (Qiagen, Germantown, MD, USA). 4. QIAquick gel extraction kit (Qiagen). 5. 10Â digest buffer (New England Biolabs, Ipswich, MA, USA). 6. SfiI restriction enzyme (New England Biolabs). 7. NotI restriction enzyme (New England Biolabs). 8. 10Â T4 DNA ligation buffer (New England Biolabs). 9. T4 DNA ligase (New England Biolabs). 10. Chemically competent TG1 (Lucigen). 11. BstNI restriction enzyme (New England Biolabs). 12. Spectrophotometer (NanoDrop from ThermoScientific, Wal- tham, MA, USA).

2.4 FACS Analysis 1. 96-well polystyrene round-bottom plates (Corning). of Selection Output 2. 96-well polypropylene v-bottom plates (Corning). 3. Binding buffer: PBS, pH 7.4, 1% BSA. 4. Biotinylated anti-M13 antibody (Sigma, St. Louis, MO, USA). 5. Streptavidin-phycoerythrin (Invitrogen, Carlsbad, CA, USA). 6. Flow cytometer (Accuri™ C6 from BD Biosciences, San Jose, CA, USA).

2.5 ScFv 1. pSyn1 scFv expression vector. Construction, 2. NcoI restriction enzyme (New England Biolabs). Production, and Biotin 3. NotI restriction enzyme (New England Biolabs). Labeling 4. Bacteria/scFv-pSyn1 plate: YT, 50 μg/ml ampicillin. LCM-Based Phage Antibody Library selection on Cancer Patient Tissues 335

5. Overnight medium: 2ÂYT, 2% (w/v) glucose, 50 μg/ml ampicillin. 6. Bacterial growth medium: 2ÂYT, 0.1% (w/v) glucose, 100 mM ampicillin. 7. Induction solution: 1 mM IPTG (1:1000 diluted from 1 M stock). 8. Periplasmic prep buffer (PPB): 200 mg/ml sucrose, 1 mM EDTA, 30 mM Tris–HCl, pH 8.0, ﬁlter-sterilized.

9. Osmotic shock buffer: 5 mM MgSO4. 10. Ni-NTA agarose resin (Qiagen). 11. Wash buffer: PBS containing 20 mM imidazole. 12. Elution buffer: PBS containing 250 mM imidazole. 13. High-speed centrifuge tubes (Nalgene). 14. Zeba™ spin desalting columns (#87766, ThermoFisher). 15. EZ-Link Sulfo-NHS-LC-Biotin (#21327, ThermoFisher) 16. 1 M Tris–HCl, pH 8.0. 17. Ultra-15 concentrator tube (Millipore, Billerica, MA, USA).

2.6 Analysis 1. Unmasking solution (Vector Laboratories). of Selection Output by 2. Blocking buffer: 2% (v/v) goat serum in PBS, pH 7.4. Immuno- 3. Avidin/biotin blocking kit (Vector Laboratories). histochemistry 4. R.T.U. ABC REAGENT (Vector Laboratories). 5. DAB solution (ThermoFisher).

6. Acid rinse solution: 2% (v/v) glacial acetic acid in ddH2O.

7. Bluing solution: 1.5 ml NH4OH (30% stock) with 98.5 ml of 70% ethanol. 8. VectaMount™ Mounting Medium (H-5000, Vector Laboratories). 9. Keyence BZ-9000 digital microscope (Keyence-America, Itasca, IL, USA).

2.7 Microscopic 1. Lab-Tek II glass chambers (ThermoFisher). Detection of Antibody 2. Permeabilization buffer: PBS, 1% BSA, 0.1% TritonX-100. Internalization 3. Streptavidin-Cy3 (1 mg/ml) (Jackson immunoresearch, West Grove, PA, USA). 4. Fix/counterstain buffer: 4% PFA in PBS, 1:10,000 Hoechst. 5. Confocal microscope (Fluoview FV10i from Olympus- America, San Jose, CA, USA). 336 Yang Su et al.

2.8 Immunotoxin 1. 96-well tissue culture-treated ﬂat-bottom plates (Corning). Delivery (Functional 2. Streptavidin-conjugated saporin (SA-ZAP, Advanced Target- Internalization) ing Systems, San Diego, CA, USA). 3. CCK-8 solution: (Dojindo, 10 μl CCK-8 solution mixed with 90 μl PBS just before use). 4. Plate reader (Synergy HT from Biotek, Winooski, VT, USA).

3 Methods

A summary of the experimental workflow of the protocols described below is as follows: First, a phage antibody display sublibrary enriched for antibodies that bind tumor cell-specific internalizing cell surface epitopes is generated (described in Subheading 3.1). This sublibrary is created by first depleting a naive phage antibody display library against a mixed panel of normal cell lines, followed by selection for binding to a mixed panel of tumor cell lines under internalizing conditions. Next, this sublibrary is selected on cancer patient tissue sections (frozen or paraffin- embedded) and the tumor cells and associated binding phage antibodies are precisely excised by LCM (described in Subheading 3.2). The tumor-binding scFv sequences are recovered from the LCM-procured tumor cells by PCR and re-cloned into the phage display vector (described in Subheading 3.3). The LCM-selected phage antibodies are then screened for binding to tumor and normal cell lines by FACS and the tumor-specific binding clones are sequenced (described in Subheading 3.4). The tumor-specific antibodies are then cloned into a scFv expression vector, produced as scFvs, and biotinylated (described in Subheading 3.5). Finally, the biotinylated scFvs are used for immunohistochemistry analysis of tumor-specific binding to cancer patient tissue samples (described in Subheading 3.6), microscopic analysis of tumor- specific internalization using tumor and normal cell lines (described in Subheading 3.7), and tumor-specific intracellular delivery of immunotoxins (described in Subheading 3.8).

3.1 Generation 1. To prepare cells for counter-selection, culture normal human of Phage Antibody fibroblasts, non-cancerous epithelial lines RWPE-1, BPH-1, Sublibrary Enriched MCF10A, and human mammary epithelial cells (HMEC) in for Internalizing three 10 cm diameter round cell culture-treated plates each in Antibody Binding cell growth medium to approximately 80% confluence (see to Tumor Cell Surface Note 1). Epitopes 2. Remove the cell growth medium and wash cells once with 5 ml PBS. Add enough trypsin/EDTA solution to cover the cells and incubate at 37 C, roll the flask gently to detach the cells from the flask. Add an equal volume of cell growth medium LCM-Based Phage Antibody Library selection on Cancer Patient Tissues 337

into the flask, pipette to mix and transfer the cells to centrifuge tubes. 3. Spin down and wash the cells three times with cell growth medium by centrifugation at 500 Â g for 5 min. 4. Pool the cells (~108 total cells) by resuspending in 3 ml ice-cold cell growth medium, add 1 ml 1012 phage particles in PBS to the cells, and incubate at 4 C for 4 h, with rotation. 5. Spin down the cells by centrifugation at 500 Â g for 5 min. Carefully transfer the supernatant containing phage antibodies to a new tube. Spin again at 6000 Â g for 5 min and filter the supernatant with a 0.45 μm sterile syringe filter. 6. Prepare 106 target tumor cells (e.g., a panel of prostate cancer cell lines, PC3 and Du-145) in 1 ml cell growth medium as described in steps 1–4 (one 10 cm dish for each cell line will be sufficient). Add phage particles from step 5 to the target tumor cells and incubate in a humidified atmosphere of 95% air and 5% CO2 at 37 C for 2 h (see Note 2). 7. Spin down and wash cells twice with 5 ml 100 mM glycine (pH 2.8) in the presence of 150 mM NaCl, followed with washing once with 10 ml PBS, pH 7.0 (see Note 3). 8. Lyse the cells by adding 0.5 ml fresh 100 mM trimethylamine (TEA) to the cell pellet. Mix with pipette and rotate at room temperature for 5 min. 9. Add 250 μl Tris–HCl, pH 6.8 to neutralize the cell lysate. Mix gently. 10. Add total volume of neutralized cell lysate to 10 ml of exponentially growing TG1 (OD600 ¼ 0.7), mix, and incubate without shaking at 37 C for 30 min. 11. Titer the phage by making a tenfold dilution of the culture in 2ÂYT/tet and plate 100 μl of each dilution (10, 1, 0.1 μlof original culture) on YT/tet plates. 12. Spin down the remaining bacterial culture at 4000 Â g for 15 min, resuspend in 0.5 ml of 2ÂYT/tet, plate on a 150 mm YT/tet plate, and incubate overnight at 37 C. 13. The next day, add 3–5 ml of 2ÂYT/tet to the plate, scrape the bacteria, and mix with the 2ÂYT/tet. Add glycerol to a final concentration of 25% (v/v), aliquot, and store at À80 C. 14. To prepare the phage sublibrary for the next selection step, inoculate 100 ml 2ÂYT with 0.1% of the panning output from step 13, culture with shaking (250 rpm) at 37 C overnight (see Note 4). 15. Centrifuge the bacteria culture at 6000 Â g for 20 min at 4 C. 16. Collect and transfer the supernatant to a new centrifuge tube, add 30 ml 20% PEG 8000/NaCl solution, mix and incubate 338 Yang Su et al.

on ice for 2 h. Phage precipitation should be visible as the supernatant should become cloudy. 17. Centrifuge the precipitated phage at 6000 Â g for 20 min at 4 C. Remove as much supernatant as possible. Resuspend the phage sediment in 30 ml PBS and transfer to a new tube. 18. Add 10 ml 20% PEG 8000/NaCl solution to resuspended phage, mix, and incubate on ice for 1 h. 19. Centrifuge the precipitated phage at 6000 Â g for 20 min at 4 C. Remove as much supernatant as possible. Centrifuge at 6000 Â g again for 5 min at 4 C and remove the residual supernatant. 20. Add 5 ml PBS to resuspend the phage. Transfer the supernatant to new tubes and centrifuge at 10,000 Â g for 10 min at 4 C. 21. Filter the supernatant with a 0.45 μm sterile syringe ﬁlter and store the phage for further selection (see Note 5).

3.2 Selection Selections were performed on both frozen and paraffin-embedded of Antibodies prostate cancer tissues. Targeting Tumor Cells For selection on frozen tissue slides: In Situ by LCM 1. Cut cryostat sections of prostate cancer specimens at 5 μm and mount on Leica Membrane Slides. Fix with ice-cold acetone at À20 C for 20 min (see Note 6). 2. Air-dry the sections using a fan for 30 min at room temperature to prevent sections from falling off the slides during phage antibody incubations. 3. Add ice-cold 4% PFA to each tissue section, fix at 4 C for 10 min. 4. Add one drop of Hematoxylin to the slide directly and counterstain for 3 min. Rinse slide with running tap water until rinse water is colorless. 5. Tap slides into PBS buffer, then incubate with the phage sublibrary (0.5 ml of 5 Â 1011 colony forming unit (c.f.u.)/ml stock) from Subheading 3.1, step 21 at room temperature for 1h. 6. Proceed to step 12. For the selection on paraffin-embedded tissue slides: 7. Deparaffinize paraffin tissue slides by immersing slides in xylene overnight. 8. Next day, immerse slide in a new xylene solution for 10 min. 9. Rehydrate the slides by sequential incubation in 100%, 95%, and 70% ethanol, followed by ddH2O for 5 min each. Wash the slides twice with PBS for 5 min each. LCM-Based Phage Antibody Library selection on Cancer Patient Tissues 339

10. Incubate slides with blocking solution at room temperature for 1 h. Wash slides three times with PBS for 5 min each. 11. Incubate with the phage sublibrary (0.5 ml of 5 Â 1011 cfu/ml stock) from Subheading 3.1, step 21 at room temperature for 1h. 12. Wash slides three times for 5 min each with PBS to remove unbound phage. Dehydrate the slides by sequential incubation in 70%, 95%, and 100% ethanol. Incubate for 5 min for each step (see Note 7). 13. Insert the dried slide into the specimen holder with section face down. 14. Open the cap of a 0.5 ml microcentrifuge PCR tube and place it in the collection area. 15. Focus on regions containing tumor cells of interest, draw a laser path around the target area. 16. Activate the laser, cut specimen along the predeﬁned laser path, and drop the excised cells into the cap of the collection tube by electrostatic force and gravity (see Note 8). 17. Proceed immediately to PCR ampliﬁcation or store the tissue pieces at À80 C until analysis.

3.3 Recovery 1. PCR amplify the genes encoding scFv fragments from of Phage Antibody LCM-collection tubes from Subheading 3.2 step 16 using from LCM-Procured the following PCR cycling conditions: Tumor Cells Initial denature 95 C 5 min Denature 94 C 1 min Annealing 55 C45s Extending 68 C 1 min Number of cycles 30 cycles Final extending 72 C 10 min

2. Purify the PCR products using a Qiagen PCR purification kit and digest the amplified fragments and fd phage display vector with SfiI and NotI at 37 C for 2–4 h. 3. Run the restriction digested products on a 1% agarose gel. Cut out target bands (approximately 800 bp for amplified scFv and 9 kb for fd phage vector) with a clean razor blade and isolate the PCR products using a Qiagen gel isolation kit. Elute in ddH2O and measure the concentration by a spectrophotometer. 4. Ligate precut PCR products into the fd phage display vector using T4 DNA ligase at room temperature for 15 min. Trans- form ligation products into chemically competent TG1. 340 Yang Su et al.

Culture at 37 C at 225 rpm for 45 min and plate all the bacteria on YT/tet plate. 5. Make the LCM-selected phage sublibrary as described in Sub- heading 3.1, steps 14–21. 6. Amplify the scFv gene by colony PCR from phage-infected bacteria, digest the PCR products with BstNI, and analyze on a 1% agarose gel to estimate the diversity of recovered scFv sequences (see Note 9).

3.4 Analysis 1. Inoculate individual phage-infected bacteria in 96-well U-bot- of Selection Output tom plates by picking single colonies using sterile pipette tips or by FACS toothpicks and dipping into 120 μlof2ÂYT/tet per well, leaving one or more mock well/plate without bacteria as a contamination control. 2. Culture the plates at 37 C with shaking at 200 rpm for 18 h. 3. Transfer 50 μl of bacterial/phage culture per well into a new 96-well microtiter plate. Add 50 μlof2ÂYT/tet containing 50% glycerol to each well, pipette up and down to mix, and store at À80 C as master plates. 4. Centrifuge the remaining bacteria/phage in the U-bottom plate at 4000 Â g for 30 min for screening on cells by FACS (see below). 5. Prepare 105 cells/ml (10 ml for each 96-well plate) of prostate cancer (PC3 and Du-145) or non-tumorigenic control (BPH-1) cells as described in Subheading 3.1, steps 1–3. 6. Resuspend cells in PBS with 1% BSA. Add 100 μl cells per well into 96-well V-bottom plates. Transfer 30 μl of supernatant containing phage particles (5 Â 1011 cfu/ml) in Subheading 3.4, step 4 into the V-bottom 96-well plate containing tissue culture cells. Incu- bate at 4 C for 1 h with rocking. 7. Centrifuge the cells at 500 Â g for 5 min and remove the supernatant. Wash cells three times with PBS containing 1% BSA. 8. Resuspend cells in 100 μl of PBS with 1% BSA and biotinylated anti-M13 antibody (Sigma, diluted 1:1000). Incubate at 4 C for 1 h with rocking. 9. Centrifuge the cells at 500 Â g for 5 min, remove the supernatant. Wash the cells three times with PBS containing 1% BSA. 10. Resuspend the cells in 100 μl of PBS with 1% BSA and streptavidin-phycoerythrin (diluted 1:1000). Incubate at 4 C for 1 h with rocking. LCM-Based Phage Antibody Library selection on Cancer Patient Tissues 341

11. Centrifuge the cells at 500 Â g for 5 min and remove the supernatant. Wash the cells three times with PBS containing 1% BSA. Resuspend the cells in 150 μl PBS. 12. Analyze cell ﬂuorescence by FACS. 13. Sequence the phage antibodies that bind to tumor cells but not normal cells (see Note 10).

3.5 Construction, 1. Analyze the sequences of phage antibodies and design primers Expression, for PCR amplification of scFv genes with NcoI and NotI Purification, restriction digest site (see Note 11). and Biotinylation 2. PCR amplify the genes encoding scFv fragments using the of scFv Fragments cycling conditions described in Subheading 3.3, step 1. 3. Purify the PCR products using a Qiagen PCR purification kit and digest the amplified fragments and pSyn-1 vector with NcoI and NotI at 37 C for 2–4 h. 4. Run the restriction-digested products on a 1% agarose gel. Cut out target bands with a clean razor blade and isolate the PCR products using a Qiagen gel isolation kit. Elute in ddH2O and measure the concentration by a spectrophotometer. 5. Ligate precut PCR products into pSyn-1 vectors using T4 DNA ligase at room temperature for 15 min. Transform ligation products into chemically competent TG1. Culture at 37 C at 225 rpm for 45 min and plate all the bacteria on YT/amp plate. 6. Miniprep and sequence to verify scFv-pSyn1 expression clones. 7. Pick and culture single-bacterial colony overnight at 37 Cin 5 ml of overnight medium (defined in Subheading 2.5). 8. Add 1 ml of overnight cultured bacterial to 400 ml growth medium (defined in Subheading 2.5). Culture bacteria to OD600 ~ 0.7 at 37 C with shaking at 250 rpm. 9. Cool down the bacterial culture to room temperature. Add 1 M IPTG to culture to a final concentration of 1 mM. Con- tinue culturing the bacteria at 30 C for 16 h. 10. Collect and centrifuge the bacteria stock at 5000 Â g for 20 min at 4 C. Remove all the supernatant. 11. Resuspend the bacterial pellet in 12.5 ml PPB. Keep the bacterial solution on ice for 20 min. 12. Centrifuge the bacteria at 5000 Â g for 15 min at 4 C. Transfer the supernatant to a high-speed centrifuge tube. 13. Osmotically shock the cells by resuspending the pellet in 12.5 ml of 5 mM MgSO4 and incubate on ice for 20 min. 14. Combine osmotic shock prep with the periplasmic prep from step 7 and centrifuge at 10,000 Â g for 15 min at 4 C. 342 Yang Su et al.

15. Transfer the supernatant to a new 50 ml Falcon tube. Add 25 ml PBS and 500 μl pre-washed Ni-NTA agarose resin beads to the tube. Incubate at 4 C by rotating for 2 h (see Note 12). 16. Spin down the beads at 3000 Â g for 15 min at 4 C and carefully discard the liquid, and wash three times with 50 ml of PBS. 17. Transfer the beads to a new 2 ml Eppendorf tube. Spin down and wash the beads three times with wash buffer for 5 min each. 18. Add 1 ml of elution buffer to the beads and incubate at room temperature for 5 min. Spin down and transfer the liquid to a new Eppendorf tube. 19. Concentrate and buffer exchange the scFv antibody to PBS by using Millipore ultra-15 concentrator tube and Zeba™ spin desalting columns. 20. Measure the concentration of scFv antibody by using Nano- Drop according to the manufacturer’s instruction (see Note 13). 21. To biotin-label antibody, prepare a 10 mM EZ-Link Sulfo- NHS-LC-Biotin solution using ultrapure water immediately before use. Add 27 μl of 10 mM biotin solution to 1 ml of 2 mg/ml puriﬁed scFv antibody in PBS at pH 7.4, rotate the mixture at room temperature for 45 min. 22. Add 20% (v/v) of 1 M Tris–HCl (pH 8.0) to quench the reaction and mix gently by pipetting up and down. 23. Buffer-exchange to PBS and remove nonreacted biotin using Zebaspin desalting columns according to the manufacturer’s instruction. Biotin-labeled antibody can be stored at À20 C for months or 4 C for a couple of weeks until use in immunotoxin assays. 24. FACS analyze the biotinylated scFv antibody binding to prostate cancer lines as described in Subheading 3.4 by using 10 μg/ml biotinylated scFv antibody followed detection with streptavidin-phycoerythrin.

3.6 Analysis For frozen tissue slides: of Selection Output by 1. Prepare the frozen tissue sections as described in Subheading Immuno- 3.2 steps 2 and 3. Proceed to step 4. histochemistry For paraffin-embedded tissue slides: 2. Deparaffinize and rehydrate the paraffin-embedded tissue as described in Subheading 3.2 steps 7–9. LCM-Based Phage Antibody Library selection on Cancer Patient Tissues 343

3. Place the slide from step 2 into a glass jar filled with unmasking solution (diluted 1:100). Incubate the jar at 95–100 C for 10 min in a pressure cooker. Remove the jar to room temperature and allow the slides to cool to room temperature (in about 20 min). Rinse the slide twice with PBS for 5 min each. 4. Use “liquid blocker” pen to demarcate the tissue. 5. Block endogenous peroxidase activity by incubating the slide in 3% H2O2 in PBS, 10 min. Rinse the slide three times in PBS for 5 min each. 6. Incubate the slide with 2% goat serum at room temperature for 30 min. Rinse briefly with PBS and incubate with avidin solution for 15 min. Rinse briefly with PBS followed by incubating with biotin solution 15 min. Rinse briefly with PBS. 7. Add 50 μg/ml of biotinylated scFv antibody in PBS with 2% goat serum, incubate for 1 h at room temperature, rinse three times in PBS for 5 min each. 8. Add R.T.U. ABC REAGENT and incubate the slides for 30 min at room temperature. Rinse three times in PBS for 5 min each. 9. Add the DAB solution to the slide and check the reaction under a microscope within 5 min. Rinse in PBS to stop the reaction. 10. Add hematoxylin to the slide directly for counter staining for 3 min. Rinse slide with running tap water until rinse water is colorless. 11. Dip slides ten times in acid rinse solution, followed by ten dips in tap water. 12. Incubate slides in bluing solution for 1 min followed by ten dips in tap water. 13. Dehydrate in 75% ethanol, 95% ethanol, 100% ethanol for 5 min each. Then clear the slides by incubating in xylene twice for a total of 15 min and allow the slide to air dry. 14. Add VectaMount™ mounting medium and apply coverslip. Analyze antibody staining under a microscope.

3.7 Microscopic 1. Grow prostate cancer (DU145) cells in Lab-Tek II glass cham- Analysis of Antibody bers to 50–60% conﬂuence. Internalization 2. Wash the cells once with pre-warmed fresh growth medium. 3. Add cell growth medium (deﬁned in Subheading 2.1) containing 15 μg/ml of biotinylated scFv and 50 μg/ml of ND70-TR to cells, incubate at 37 C for 3 h. 4. Wash the cells three times with PBS for 5 min each. 5. Fix the cells with 4% PFA in PBS for 15 min at 4 C. 6. Wash the cells three times with PBS for 5 min each. 344 Yang Su et al.

7. Permeabilize the cells with permeabilization buffer for 20 min at room temperature. 8. Add 1:300 diluted streptavidin-Cy3 in 1:5 PBS-diluted permeabilization buffer, incubate at room temperature for 30 min. 9. Wash the cells three times with PBS for 5 min each, ﬁx and counterstain cells with 4% PFA containing Hoechst in PBS for 10 min at room temperature. 10. Image and analyze the staining using a confocal microscope.

3.8 Intracellular 1. Seed 3000 cells per well of prostate cancer (PC3 or Du-145) or Delivery non-tumorigenic control (BPH-1) in 96-well ﬂat bottom of Immunotoxin plates with 50 μl cell growth medium and culture overnight at 37 C with 5% CO2. 2. Prepare the immunotoxin by mixing biotinylated scFv with SA-ZAP at a molar ratio of 1:1, incubate on ice for 30 min. 3. Add 50 μl of serially diluted immunotoxin in PBS to each well and incubate for 96 h at 37 C with 5% CO2 (see Note 14). 4. Carefully remove the cell growth medium from each well. 5. Add 100 μl of diluted CCK-8 to each well in the 96-well plates, incubate for 1–4 h at 37 Cin5%CO2 (see Note 15). 6. Measure the absorbance at 450 nm using a microtiter plate reader and determine the EC50 value by curve ﬁtting using appropriate software (e.g., GraphPad Prism).

4 Notes

1. The list of normal cells can be expanded to additional non-tumorigenic cell lines and normal primary cells when available. 2. Incubation at 37 C in cell growth media allows for internalization of phage antibodies. 3. Wash with glycine for no longer than 5 min. 4. The inoculated media should look very slightly turbid (initial OD600 ¼ 0.05–0.1). 5. The sublibrary contained 1–5 Â 105 copies of about 106 independent clones at the concentration of 1–5 Â 1011 cfu/ml. 6. The suggested cryostat temperature is between À15 and À23 C. Slides can be stored unfixed for several months at À80 C. Frozen tissue samples saved for later analysis should be stored intact. 7. All the slides should be reviewed by a board-certified patholo- gist and regions containing clusters of tumor cells should be confirmed and marked. LCM-Based Phage Antibody Library selection on Cancer Patient Tissues 345

8. Laser microdissection microscope that uses a UV pulse laser to excise selected cells from surrounding tissues. Typically, 20–50 tumor cells were procured at a time by generating a closed laser path around the group of cells of interest. 9. Each LCM selection library contained >105 independent clones. The number of unique phage antibodies was determined by patterns of BstNI digestion [6, 7, 26]. When restriction digestion patterns showed ambiguity, phage antibody genes were sequenced to determine their uniqueness. 10. More than 600 clones from various LCM-derived sublibraries were screened. Only those clones that bound to both PC3 and Du-145 cells but not BPH-1cells were chosen for further analysis because they were more likely to recognize tumor cell surface antigens as opposed to artifacts associated with a particular tissue slide. 0 11. Two forms of soluble antibody fragments, scFv and (scFv )2, can be produced [6, 7, 39, 40]. The scFv gene was subcloned into the secretion vector pUC119mycHis, adding a c-Myc epitope tag and hexahistidine tag at the c-terminus of the 0 scFv. To create the (scFv )2 dimer for immunoliposome studies, the c-Myc epitope tag was removed, and a free cysteine was introduced at the c-terminus of the scFv preceding the hexahistidine tag. 12. Purification can be done with Ni-NTA agarose beads/gravity method or GE HisTrap column/FPLC method. 13. The purified scFv should also be analyzed by SDS-PAGE. The molecular weight of monomeric scFv is about 27 kDa. 14. For initial assessment, 1:10 serial dilutions are often used to find the linear range of activity. Once the linear range has been determined, 1:3 serial dilutions can be used to improve the accuracy of the EC50 measurement. 15. Remove any air bubbles in the well, as they interfere with the absorbance measurement.

Acknowledgment

References

1. Kobata A, Amano J (2005) Altered glycosyla- 12. Zhu X, Bidlingmaier S, Hashizume R et al tion of proteins produced by malignant cells, (2010) Identification of internalizing human and application for the diagnosis and immuno- single-chain antibodies targeting brain tumor therapy of tumours. Immunol Cell Biol sphere cells. Mol Cancer Ther 9:2131–2141 83:429–439 13. Sherbenou DW, Aftab BT, Su Y et al (2016) 2. Birkle S, Zeng G, Gao L et al (2003) Role of Antibody-drug conjugate targeting CD46 tumor-associated gangliosides in cancer pro- eliminates multiple myeloma cells. J Clin Invest gression. Biochimie 85:455–463 126:4640–4653. https://doi.org/10.1172/ 3. Hakomori S (2001) Tumor-associated carbo- JCI85856 hydrate antigens defining tumor malignancy: 14. Bidlingmaier S, Zhu X, Liu B (2008) The util- basis for development of anti-cancer vaccines. ity and limitations of glycosylated human Adv Exp Med Biol 491:369–402 CD133 epitopes in defining cancer stem cells. 4. Hanisch FG (2001) O-glycosylation of the J Mol Med (Berl) 86:1025–1032. https://doi. mucin type. Biol Chem 382:143–149 org/10.1007/s00109-008-0357-8 5. Ugorski M, Laskowska A (2002) Sialyl Lewis 15. Marks J, Hoogenboom H, Bonnert T et al (a): a tumor-associated carbohydrate antigen (1991) By-passing immunization. Human involved in adhesion and metastatic potential antibodies from V-gene libraries displayed on of cancer cells. Acta Biochim Pol 49:303–311 phage. J Mol Biol 222:581–597 6. Liu B, Conrad F, Cooperberg MR et al (2004) 16. McCafferty J, Griffiths A, Winter G et al Mapping tumor epitope space by direct selec- (1990) Phage antibodies: filamentous phage tion of single-chain Fv antibody libraries on displaying antibody variable domains. Nature prostate cancer cells. Cancer Res 64:704–710 348:552–554 7. Ruan W, Sassoon A, An F et al (2006) Identifi- 17. O’Connell D, Becerril B, Roy-Burman A et al cation of clinically significant tumor antigens (2002) Phage versus phagemid libraries for by selecting phage antibody library on tumor generation of human monoclonal antibodies. cells in situ using laser capture microdissection. J Mol Biol 321:49–56 Mol Cell Proteomics 5:2364–2373. https:// 18. Hoogenboom HR, Winter G (1992) doi.org/10.1074/mcp.M600246-MCP200 By-passing immunisation. Human antibodies 8. An F, Drummond DC, Wilson S et al (2008) from synthetic repertoires of germline VH Targeted drug delivery to mesothelioma cells gene segments rearranged in vitro. J Mol Biol using functionally selected internalizing human 227:381–388 single-chain antibodies. Mol Cancer Ther 19. de Haard HJ, Reurs A, Hufton SE, Roovers 7:569–578. https://doi.org/10.1158/1535- RC, Henderikx P, de Bruine AP, Arends JW, 7163.MCT-07-2132 Hoogenboom HR (1999) A large 9. Bidlingmaier S, He J, Wang Y et al (2009) non-immunized human Fab fragment phage Identification of MCAM/CD146 as the target library that permits rapid isolation and kinetic antigen of a human monoclonal antibody that analysis of high affinity antibodies. J Biol Chem recognizes both epithelioid and sarcomatoid 274:18218–18230 types of mesothelioma. Cancer Res 20. Clackson T, Hoogenboom HR, Griffiths AD 69:1570–1577. https://doi.org/10.1158/ et al (1991) Making antibody fragments using 0008-5472.CAN-08-1363 phage display libraries. Nature 352:624–628. 10. Bidlingmaier S, Su Y, Liu B (2015) Combining https://doi.org/10.1038/352624a0 phage and yeast cell surface antibody display to 21. Winter G, Griffiths A, Hawkins R et al (1994) identify novel cell type-selective internalizing Making antibodies by phage display technol- human monoclonal antibodies. Methods Mol ogy. Annu Rev Immunol 12:433–455 Biol 1319:51–63. https://doi.org/10.1007/ 22. Barbas CF 3rd, Kang AS, Lerner RA et al 978-1-4939-2748-7_3 (1991) Assembly of combinatorial antibody 11. Ha KD, Bidlingmaier SM, Zhang Y et al (2014) libraries on phage surfaces: the gene III site. High-content analysis of antibody phage- Proc Natl Acad Sci U S A 88:7978–7982 display library selection outputs identifies 23. Huston JS, George AJ (2001) Engineered antitumor selective macropinocytosis-dependent bodies take center stage. Hum Antibodies rapidly internalizing antibodies. Mol Cell Pro- 10:127–142 teomics. https://doi.org/10.1074/mcp. 24. Janda KD, Lo CH, Li T et al (1994) Direct M114.039768 selection for a catalytic mechanism from LCM-Based Phage Antibody Library selection on Cancer Patient Tissues 347

combinatorial antibody libraries. Proc Natl 34. Johann DJ, Rodriguez-Canales J, Mukherjee S Acad Sci U S A 91:2532–2536 et al (2009) Approaching solid tumor hetero- 25. Sheets MD, Amersdorfer P, Finnern R et al geneity on a cellular basis by tissue proteomics (1998) Efficient construction of a large nonim- using laser capture microdissection and mune phage antibody library: the production biological mass spectrometry. J Proteome Res of high-affinity human single-chain antibodies 8:2310–2318. https://doi.org/10.1021/ to protein antigens. Proc Natl Acad Sci U S A pr8009403 95:6157–6162 35. Lu H, Jin D, Kapila YL (2004) Application of 26. Liu B, Marks JD (2000) Applying phage anti- laser capture microdissection to phage display bodies to proteomics: selecting single chain Fv peptide library screening. Oral Surg Oral Med antibodies to antigens blotted on nitrocellu- Oral Pathol Oral Radiol Endod 98:692–697. lose. Anal Biochem 286:119–128. https:// https://doi.org/10.1016/ doi.org/10.1006/abio.2000.4788 S1079210404006134 27. Liu B, Huang L, Sihlbom C et al (2002) 36. Kubo N, Akita N, Shimizu A et al (2008) Towards proteome-wide production of mono- Identification of oligopeptide binding to clonal antibody by phage display. J Mol Biol colon cancer cells separated from patients 315:1063–1073 using laser capture microdissection. J Drug 28. Bonner RF, Emmert-Buck M, Cole K et al Target 16:396–404. https://doi.org/10. (1997) Laser capture microdissection: molecu- 1080/10611860802088796 lar analysis of tissue. Science 278:1481,1483 37. Sun Y, Shukla GS, Kennedy GG et al (2009) 29. Emmert-Buck MR, Bonner RF, Smith PD et al Biopanning phage-display libraries on small tis- (1996) Laser capture microdissection. Science sue sections captured by laser capture microdis- 274:998–1001 section. J Biotechnol Res 1:55–63 30. Best CJ, Emmert-Buck MR (2001) Molecular 38. Sun Y, Shukla GS, Weaver D et al (2009) profiling of tissue samples using laser capture Phage-display selection on tumor histological microdissection. Expert Rev Mol Diagn specimens with laser capture microdissection. J 1:53–60. https://doi.org/10.1586/ Immunol Methods 347:46–53. https://doi. 14737159.1.1.53 org/10.1016/j.jim.2009.06.002 31. Mukherjee S, Rodriguez-Canales J, Hanson J 39. Roth A, Drummond DC, Conrad F et al et al (2013) Proteomic analysis of frozen tissue (2007) Anti-CD166 single chain antibody- samples using laser capture microdissection. mediated intracellular delivery of liposomal Methods Mol Biol 1002:71–83. https://doi. drugs to prostate cancer cells. Mol Cancer org/10.1007/978-1-62703-360-2_6 Ther 6:2737–2746 32. Espina V, Wulfkuhle JD, Calvert VS et al 40. Iyer AK, Su Y, Feng J et al (2011) The effect of (2006) Laser-capture microdissection. Nat internalizing human single chain antibody frag- Protoc 1:586–603. https://doi.org/10. ment on liposome targeting to epithelioid and 1038/nprot.2006.85 sarcomatoid mesothelioma. Biomaterials 32:2605–2613. https://doi.org/10.1016/j. 33. Espina V, Milia J, Wu G et al (2006) Laser biomaterials.2010.11.073 capture microdissection. Methods Mol Biol 319:213–229. https://doi.org/10.1007/ 978-1-59259-993-6_10 Chapter 19

Antibody Isolation From a Human Synthetic Combinatorial and Other Libraries of Single-Chain Antibodies

Almog Bitton, Limor Nahary, and Itai Benhar

Abstract

Antibody libraries came into existence 25 years ago when the accumulating sequence data of immunoglobulin genes and the advent of the PCR technology made it possible to clone antibody gene repertoires. Phage display (most common) and additional display and screening technologies were applied to pan out desired binding speciﬁcities from antibody libraries. “Synthetic” or “semisynthetic” libraries are from naive—non- immunized source and considered to be a source for many different targets, including self-antigens. As other antibody discovery tools, phage display is not an off-the-shelf technology and not offered as a kit but rather requires experience and expertise for making it indeed very useful. Here we present application notes that expand the usefulness of antibody phage display as a very versatile and robust antibody discovery tool.

Key words Synthetic library, Phage display, Single-chain antibodies, scFv—single-chain variable fragment, VH—variable region of antibody heavy chain, VL—variable region of antibody light chain, FR—variable frameworks region, CDRs—complementarity determining regions

1 Introduction

Antibody phage display was the first and is still the most popular tool to access antibody libraries (reviewed in [1, 2]). This method, in its most common format, is based on the expression of functional antibody fragments (scFvs or Fabs) fused with the minor coat protein (g3p) of the filamentous phage [3], was demonstrated for the first time in 1990, and provided the way to quickly isolate recombinant antibodies from antibody libraries on the basis of antigen-binding by individual library clones [4]. In such systems, the genetic information encoding for the displayed molecule is physically linked to its product via the displaying phage particle. The most popular antibody formats present in libraries were the single-chain variable fragment (scFv), as pioneered by the groups of Sir Gregory Winter at the Medical Research Council, Cambridge, UK [5] and Melvyn Little’s group at Heidelberg, Germany [6], and

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_19, © Springer Science+Business Media LLC 2018 349 350 Almog Bitton et al.

the Fab’, as pioneered by the group of Burton and Lerner at the Scripps Research Institute, La Jolla, CA, USA [3]. The first libraries that were built from natural sources of sequence diversity, namely animal or human B cells, were soon followed by libraries into which sequence diversity was inserted artificially [7]. Antibody libraries are classified as “immune libraries” when the source for antibody genes is an immunized donor [3, 7], or as “non-immune” or “naı¨ve” libraries when the source for antibody genes is a donor (animal or human) that was not intentionally immunized for the purpose of library construction [8]. When diversity is inserted artificially (as done by inserting random sequences into the antigen-combining site), the result is a “semi-synthetic” or a “synthetic” antibody library [2, 9, 10]. The advantage of synthetic libraries is that, from a sufficiently large library, scFv/Fab antibodies can be isolated against any desired target [8, 10–14]. Phage libraries are enriched for specific-binding clones by subjecting the phage to repetitive rounds of selection (also known as panning). In 2009, we published within the MMB series a protocol for designing a human synthetic combinatorial library (“Ronit 1 library”) of scFvs [15]. In that chapter, we described the construction of a human synthetic scFv antibody phage-display library, using the n-CoDeR principle [12, 14, 16], as was done in our lab [17]. According to this approach, scFvs were constructed based on a single master framework (FR) for each variable domain (VH, V-kappa and V-lambda), using shuffled CDR1–CDR3 sequences that originated from many different in vivo-formed V-genes. The DP-47, DPL-3, and DPK-22 were used as master frameworks for VH, V-lambda, and V-kappa, respectively. Here, we describe an updated version of the protocol, with a focus on antibody isolation from phage display libraries, which contains application notes that significantly increase the probability of isolating specific, stable, high-affinity antibodies from, in principle, any large and diverse antibody phage display library. The actual construction of the original library [17] will not be described in this chapter; however, we describe it schematically in Fig. 1 and for additional details we refer the reader to the original chapter [15], the article where the construction of the “Ronit 1 library” was initially described [17] or to request a copy from the corresponding author.

2 Materials

Prepare all solutions using ultrapure water (prepared by purifying deionized water, to attain a sensitivity of 18 mΩ-cm at 25 C) and analytical grade reagents. Prepare and store all reagents at room temperature (unless indicated otherwise). Diligently follow all waste disposal regulations when disposing waste materials. Fig. 1 An outline of the “Ronit 1” human-synthetic library construction. (a) Human cDNA libraries from spleens, lymph nodes, and peripheral blood leukocytes are used as templates for PCR ampliﬁcation of each CDR individually into CDR pools. (b) The ampliﬁed CDR pools are mixed with oligonucleotides encoding 352 Almog Bitton et al.

1. Bacterial glycerol stock of a phage display library (in house production). 2. Glycerol. 3. Bacterial growth media: (a) YTAG: 2ÂYT medium: 16 g/L tryptone, 10 g/L yeast extract, 5 g/L NaCl, A: 100 μg/L ampicillin, G:1%D- glucose. (b) YTAK: 2ÂYT medium, A: 100 μg/L ampicillin; K: 50 μg/L kanamycin. (c) Difco™ LB Broth, Lennox (BD, USA): 4. Helper phage: A variety of helper phages are available for the rescue of phagemid libraries, such as VCS-M13 (Stratagene, La Jolla, CA, USA) and M13KO7 (Bio-Rad Laboratories, Hercu- les, CA, USA). 5. Filtrap—Filter System 0.45 μm CA (Corning, NY, USA).

6. PEG/NaCl (PEG6000–8000 200 g/L (Sigma, Israel); NaCl 146.1 g/L). 7. Phosphate buffers: (a) 10Â phosphate-buffered saline (PBS) was purchased from Sigma, Israel. (b) PBST: PBS supplemented with 0.05% Tween-20 detergent (Sigma, Israel). (c) 3% MBPS: 3% skim milk powder in PBS. 8. For capture of his-tagged proteins—magnetic nickel beads: Dynabeads® His-Tag Isolation & Pulldown (Life Technolo- gies Ltd., UK). For capture of biotinylated proteins and peptides: Dynabeads® M-280 Streptavidin (Thermo Fisher Scientiﬁc). 9. 24-well cell culture plates. 10. Bovine serum albumin (BSA). 11. Triethylamine. 12. 1.5 M Tris(HCl) solution pH 7.0.

13. E. coli strains: XL-1 Blue and TG-1 (see Note 1). ä

Fig. 1 (continued) framework regions, and intact cassettes encoding variable domains (VH and VL) are assembled using a two-step overlap-extension PCR. (c) The ampliﬁed variable domain pools are mixed 0 0 with oligonucleotides containing restriction sites (NcoI and the 5 of the VH; NotI at the 3 of the VL), and intact cassettes encoding scFv are assembled using overlap-extension PCR. (d) The newly assembled scFv cassettes are cloned into pCC phagemid vector in frame by the NcoI and NotI restriction sites Antibody Isolation From a Human Synthetic Combinatorial and Other... 353

14. HRP-conjugated anti-M13 antibody (GE Healthcare Life Sciences). 15. Isopropyl β-D-1-thiogalactopyranoside (IPTG). 16. Triton-X100 (a nonionic surfactant).

3 Methods

3.1 Afﬁnity Selection 1. Inoculate an aliquot of the bacterial library glycerol stock of scFv Displaying (~1 Â 1010 clones) into 100 mL YTAG. Phages 2. Grow with 250 RPM shaking at 37 C until on Immobilized OD600nm ¼ 0.4–0.6. Antigen 3. Infect the cells with helper phage at the multiplicity of infection (MOI) of 20 (number of helper phage particles/number of 3.1.1 Growth and Helper 8 target bacteria, taking into account that 1 OD600nm ~2 Â 10 Phage Rescue bacteria/mL) (see Note 1) and shake for a few seconds. of the Library (See Note 2) 4. Incubate at 37 C for 30 min without shaking and then for additional 30 min with 250 RPM shaking. 5. Spin the infected cells at 3300 Â g for 10 min and resuspend the pellet in 200 mL of YTAK medium. Incubate overnight at 30 C with 250 RPM shaking. 6. Spin the culture at 8000 Â g for 10 min at 4 C and ﬁlter the supernatant with a 0.45 μm Filtrap. 7. Add 1/5 volume PEG/NaCl to the supernatant (50 mL PEG/NaCl to 200 mL YTAK). Mix well and keep on ice for 1h. 8. Spin at 10,800 Â g for 30 min at 4 C. Resuspend the pellet in 0.5–5 mL of sterile PBS. 9. Store the phage supernatant at 4 C for short-term storage (and skip to step 11), or add sterile glycerol (15% v/v) for long-term storage at À80 C. “Rescued” phages that have been stored for no longer than 1 week at 4 C should be used as input for panning. 10. Before the panning procedure, precipitate phages using PEG/- NaCl as described above, to remove the glycerol, and resuspend in sterile PBS. 11. To titer the phage stock make serial tenfold dilutions of the phages in sterile PBS. Seed logarithmic E. coli cells (see Note 1) in a sterile 96-well plate (90 μL/well) and infect with 10 μLof diluted phages (infect with the 107–1013 dilutions). Mix by pipetting up and down and incubate at 37 C for 1 h. Plate the infected cells on YTAG plates and grow overnight at 37 C. Phage stock titer should be 1012–1013/mL. 354 Almog Bitton et al.

3.1.2 Affinity Selection There are various approaches for antigen immobilization (e.g., (Panning) on Immobilized plastic plates, polystyrene beads, immunotubes, and magnetic Antigen pull-down beads) that enable the enrichment of binders from a phage display library by applying sequential affinity-selection (panning) cycles. We found that using alternating phage capture approaches (i.e., a different antigen immobilization method in every other cycle) helps depleting phage clones that bind the surfaces of the solid phases used for protein immobilization (see Note 3). As a routine, we advise using two alternating complexes: (1) magnetic beads in the 1st and 3rd cycles; and (2) a 24-well plate in the second and fourth cycles. Due to space limitations we describe here a selection method using magnetic nickel beads (IMAC) for immobilizing a 6Â His-tagged antigen. For biotinylated antigens (proteins and peptides) we use streptavidin magnetic beads in the bead-capture cycles. Please refer to our previously published protocol [15] for the description of using a 24-well plate for antigen immobilization (see Note 4). A scheme of library construction and affinity selection is shown in Fig. 2. The selection efficiency depends on many factors, such as: the selection condition (an immobilized antigen on a plate or beads, or a cell-displayed antigen); the antigen’s concentration in a solution or its density on the surface of a solid phase; the number of washes and the duration of each. In order to preserve rare binders, we recommend performing the initial panning cycles using relatively high antigen concentrations and short washes, and to employ more stringent washing conditions in later selection cycles (see Note 5). All incubations described below are performed at RT (room temperature, about 25 C) unless mentioned otherwise. Blocking

1. Using the suitable magnet, capture 1 mg of magnetic nickel beads (insure excess of beads over antigen) for 1 min, remove the sup, wash with 1 mL PBS, and capture again. Remove the sup and resuspend the beads in 1 mL of blocking solution (PBS þ 2% BSA) and incubate for 1 h in a rotating platform (see Notes 6 and 7). 2. Suspend 1012 phages in 1 mL of blocking solution (this is the 1st panning input) and incubate for 1 h in a rotating platform. This blocking step decreases nonspeciﬁc binding of phages to the beads. Binding

3. If you wish to deplete phages that bind speciﬁc regions in your antigen, refer to Note 8. Otherwise, proceed to step 4. 4. Transfer the blocked phages to an Eppendorf tube containing 10 μgofa6Â His-tagged antigen. Incubate for 1 h in a rotating platform. Antibody Isolation From a Human Synthetic Combinatorial and Other... 355

Fig. 2 Scheme of library construction and affinity selection. Steps 1–3 are the library construction and phage preparation steps. Steps 4–7 describe an affinity selection cycle which should be repeated about four times to obtain sufficient enrichment of antigen binders allowing characterization of monoclonal phage clones (step 8)

5. Capture the blocked beads (from step 1) and remove the supernatant. Transfer the antigen-phage mix to the beads. Incubate for 30 min in a rotating platform. In this step, phage-antigen-beads complexes are formed. Washing

6. Capture the beads for 1 min on the magnet, remove the sup, add 1 mL of PBST, and incubate 5 min. Repeat 9 more times with PBST and 3 times with PBS (see Note 5). Elution

7. Remove the excess PBS from the beads and elute phages by adding 1 mL of 100 mM triethylamine pH 13.0 (14 μL 356 Almog Bitton et al.

trimethylamine (7.18 M) in 1 mL ultrapure water, diluted on the day of use) and incubate for 25 min on a rotating platform. 8. Add the eluted 1 mL phages into a 13 mL polypropylene culture tube containing 1 mL of 1.5 M Tris (HCl) pH 7.0. Neutralized phages can be stored for several days at 4 Cor (better) used to immediately infect E. coli cells (see Notes 1 and 9)asinstep 9. The neutralized phages solution is the ﬁrst panning output. Infection

9. Add 1 mL of the neutralized output phages (store the other half at 4 C) to 5 mL of an exponentially growing culture (OD600nm ¼ 0.4–0.6) of E. coli XL-1 Blue cells in 2ÂYT medium (see Note 1). Mix well and incubate at 37 C for 30 min without shaking and then for additional 30 min with shaking at 250 RPM. You may perform step 11 during this incubation time. 10. Centrifuge the infected E. coli XL-1 Blue culture at 3300 Â g for 10 min. Resuspend the pellet in 1 mL YTAG and spread on two 15 cm YTAG plates. Grow overnight at 37 C. Titration

11. Transfer exponential E. coli XL-1 Blue cells into a 96-well plate (90 μL/well). Infect with serial tenfold dilutions of the input (1010–1012) and output phages (103–109). Incubate at 37 C for 1 h without shaking. Plate on YTAG plates. Grow overnight at 37 C to determine the panning input and output sizes. Output titer should be 104 and 107/mL. Amplification of output phages and further selection cycles 12. The first selection cycle is the most important one. Any errors made at this point will only be amplified in the following selection cycles. You should get back at least 104 phages as cycle 1 panning output. If you obtain less it is probable that a mistake had occurred. Repeat the infection of the remaining 1 mL of eluted, neutralized phages (see Notes 10 and 11); otherwise, continue to further selection cycles. 13. Using a cell scraper, scrape the output cells (from step 10) into 10 mL of YTAG medium. Plate serial tenfold dilutions onto YTAG plates to determine how much the library was amplified during the overnight growth. Pre- pare glycerol stocks (15% v/v) and store 1 mL aliquots at À80 C. Antibody Isolation From a Human Synthetic Combinatorial and Other... 357

14. Once you know the titer of the scraped bacteria, inoculate (in 100 mL YTAG medium) an amount of cells that yields at least 20 copies of phage output, i.e. (scraped bacteria titer)/(phage output titer) 20. For instance, having a bacterial titer of 109/mL and phage titer of 106/mL, inoculate 20 μL of scraped bacteria (2 Â 107 cells). 15. Continue with phage rescue as in Subheading 3.1.1, steps 2–11. 16. Use 1011 phages as input for the next panning cycle. Store the remaining phages at 4 C(see Note 12). 17. Repeat the selection for a total of 3–4 cycles. In each cycle, decrease the size of phage input and the antigen’s concentration (a factor of 5–10 is reasonable) (see Notes 5, 12, and 13). 18. Monitor the ratio between panning input and panning output in each cycle. With successful enrichment of binders you should observe a descending input/output ratio.

3.2 Identiﬁcation Phage ELISA serves as the primary method for screening scFv- of Antigen Binders displaying phage clones that speciﬁcally bind the antigen. There- fore, for reliable results, high phage titers are critical as well as the number of scFv molecules displayed on an average phage. We addressed those issues by comparing phages that were produced in various strains of E. coli and found that phages rescued from E. coli TG-1 cells yield the strongest signals in phage ELISA. Consequently, TG-1 cells should be infected with output phages of the last panning cycle (XL-1 Blue cells should also be infected for preparing glycerol stocks and for phage rescue of further panning cycles, if required) (see Note 14). Single-phage clones preparation

1. Use 50–100 μL of phages from the last output (either those eluted and neutralized at the end of the last panning cycle, or rescued phages from XL-1 Blue cells) to infect a 5 mL exponential E. coli TG-1 culture growing in 2ÂYT medium. Mix well and incubate at 37 C for 30 min without shaking, and then for additional 30 min shaking at 250 RPM. 2. Spread dilutions of infected cells on YTAG plates to obtain single, isolated colonies and grow overnight at 37 C. 3. On the following day, use sterile inoculation loops or tips to pick single colonies into single wells in a 96-well plate containing 100 μL/well of YTAG. Keep one well sterile for blank control. Grow overnight at 37 C with gentle 358 Almog Bitton et al.

shaking (100–150 RPM, to avoid contamination between wells). This is the master plate. Phage rescue

4. On the next day, dilute the cells 1/10 by transferring 10 μL from each well of the master plate into a new 96-well plate containing 90 μL/well of YTAG. Grow to mid-log at 37 C shaking at 150 RPM (~2 h for TG-1 cells). 5. Initiate rescue by adding 11 μLof1010/mL helper-phage per well. Incubate at 37 C for 30 min without shaking, and then for additional 30 min shaking at 150 RPM. 6. Spin at 3200 Â g for 10 min at 14 C. Discard the supernatant quickly and add 150 μL/well of YTAK. Grow overnight at 30 C shaking at 150 RPM. This is the rescue plate. Proceed to step 7 on this day. Phage ELISA

7. Coat two ELISA plates (100 μL per well) at 5 μg/mL of antigen (1) or control protein (2) diluted in PBS, overnight at 4 C. 8. On the following day, wash the ELISA plates with PBST and block with 300 μL/well of 3% MPBS for 1 h. 9. Wash three times with PBST and add 100 μL/well of PBST to the control-protein coated wells. 10. Spin the rescue plate at 3200 Â g for 10 min at 14 C. Transfer 100 μL/well of supernatant into the control- protein-coated wells (already containing 100 μL PBST). Mix well by pipetting up and down and transfer 100 μL into the antigen-coated wells (see Note 15). 11. Complete phage ELISA by adding anti-phage secondary antibodies (e.g., HRP-conjugated anti-M13) and developing with an appropriate substrate. 12. Repeat the procedure at least once (including rescue from the same master plate) to discriminate false positives and to conﬁrm speciﬁcity of initial binders (use a different control protein each time). Use the master plate to inoculate validated clones on YTAG plates (to obtain well isolated, single colonies). 13. Proceed to step 14 to perform another screening phase by expressing soluble scFv antibodies (see Note 16). Otherwise, skip to step 21. Antibody Isolation From a Human Synthetic Combinatorial and Other... 359

Soluble ELISA

14. Prepare 5 mL LB þ 100 μg/mL ampicillin starters by inoculating single colonies of phage ELISA-veriﬁed binders and grow overnight at 37 C shaking at 250 RPM. 15. Per clone, keep 1 mL for glycerol stock (15% v/v) and 3 mL for plasmid DNA preparation. Add the remaining 1ml into 9 mL LB þ 0.4% D-glucose þ100 μg/mL ampicillin in a 50 mL tube and grow at 37 C shaking at 250 RPM. 16. At OD600nm ¼ 0.8, cool cells to 30 C and induce scFv expression by adding IPTG to 0.5 mM. Incubate for 3–4 h at 30 C shaking at 250 RPM. 17. Collect the cells by centrifugation at 3300 Â g for 10 min. Resuspend the pellet(s) in 1 mL of PBS þ 0.1% Triton- X100 and lyse the cells preferably by sonication. 18. Spin the cell extracts at 12,000 Â g for 150 at 4 C. Collect the soluble fractions (supernatants); these fractions contain the soluble scFv molecules (see Note 17). 19. Dilute the soluble fractions 1:1 in PBST and perform ELISA (plates coated with 1–5 μg/mL of antigen and control proteins). Use anti-tag antibodies for detection (see Note 18). 20. If an assessment of scFv expression levels is required, carry out a western blot analysis with 10 μL of soluble fraction alongside a series of dilutions from a reference scFv protein of known concentration. Evaluation of antibody diversity 21. Prepare plasmid preparations for all positive, antigen- speciﬁc clones. The diversity can be assessed by sequencing the scFv domain.

4 Notes

1. Filamentous phages infect F+ E. coli via the sex pili. For sex pili production and efﬁcient infection by phage, E. coli must be grown at 37 C and be in the exponential (logarithmic) growth phase (OD600 nm of 0.4–0.6). 2. All glassware that had been used for phage work should be immersed in a diluted solution (5%) hypochlorite (chlorine bleach) before being sterilized by autoclaving. 3. Some phages in the library stick to the surfaces that are used to immobilize the antigen. These phages will be eluted with the 360 Almog Bitton et al.

antigen-bound phages and therefore they will be amplified. Although sticky phages can be discriminated in the screening phase, it is recommended to deplete them sooner by not repeating the same immobilization method in the following cycle. 4. The detailed affinity-selection process relates to immobilized antigens or to panning using soluble antigen followed by capture using magnetic beads. This approach is suitable for protein and peptide antigens. In addition to protein and peptides [18–20], we successfully isolated from the “Ronit 1 library” and from immune scFv phage libraries antibodies that bind different antigens such as hapten-carrier conjugates [21], antigen-expressing mammalian cells (with counter selection on antigen-negative cells) [22] crystalline facets of semicon- ducting materials [23] and whole fungal cells (unpublished data). 5. Decrease antigen concentration on progressing cycles to enrich the high-affinity binders’ population. Standard antigen concentrations for the first four cycles: 10, 5, 1, and 0.1 μg/mL. Perform short washes (3–5 min each) in the first cycle and increase the durations in the following cycles (10 min or longer). 6. When using polystyrene surfaces (such as 24-well plates) for immobilization, the antigen must be carrier-free (CF) to prevent amplifying phages that bind the carrier protein. The antigen does not have to be CF when using affinity-based immobilizing methods. 7. To deplete sticky phages that could bind BSA, use a different blocking solution in the second cycle (for instance: 3% MPBS). You may use BSA again in the third cycle, and so on. Using the same blocking reagent for all cycles usually leads to the isolation of blocker-specific antibodies at the expense of antigen- specific ones. 8. Some antigens contain regions or domains that are common in other proteins, such as immunoglobulin domains or conserved regions between homologous proteins from different organisms. This might result in amplification of clones that bind those regions and fewer clones that bind antigen-specific regions. To isolate antigen-specific clones, perform a depletion step in which a control protein is immobilized. Allow phage-control protein complexes to form during 1 h incubation with rotation and collect the unbound phages before exposing them to the desired antigen. By doing this we were able to isolate anti-idiotype antibodies that specifically bind the CDRs of Remicade (anti human TNF-α), when Avastin (anti Antibody Isolation From a Human Synthetic Combinatorial and Other... 361

human VEGF-A) had served as the control protein (unpublished data). 9. E. coli strains such as XL-1 Blue, but not TG1, which possess the recA1 genotype, are less likely to insert DNA mutations that result from recombination. Therefore, use XL-1 Blue cells for amplification of output phages and storage. 10. Few or no colonies on plates after first panning cycle may indicate that the cells lost the F pilus and were not infected by output phages, or that antigen coating was not efficient— start a cell culture from a single colony on a minimal plate. Grow the cells at no lower than 37 C. Optimize coating and blocking buffers and conditions of the wells. 11. Too many colonies (>107) after the first panning cycle: This may be due to inadequate blocking of wells—optimize coating and blocking conditions of the wells; inadequate blocking of phages—block the phages with the same blocking solution used to block the wells; insufficient washing—increase the number of washes. 12. When your library is sufficiently large (>109 clones), you should be able to isolate high-affinity binders against most À À antigens (affinity of 10 8–10 9 M). To preferably isolate the high-affinity binders, apply “off-rate selection” by prolonging the washing time: after 20 PBST washes (Subheading 3.1.2, step 6) fill the well again with PBS supplemented with 1% BSA and drain it after 15 min; repeat several times so that accumu- lated washing time is from 1 h to overnight (and even longer). It may be advantageous to run several panning wells in parallel, each with a different total duration of washing to determine the optimal conditions for your library and particular antigen. It is also suggested to decrease the input by a factor of 10 for each progressive panning cycle. 13. If no positive binders are identified after 3–4 panning cycles, the enrichment is insufficient. Perform additional panning cycles or start from scratch using a different panning approach. 14. As an alternative to phage ELISA, next-generation sequencing of the enriched phage population can be used to interrogate the phage population and potential binders can be identified by their relative high abundance within the enriched library [21]. 15. When performing a phage ELISA (as described in [24]), we usually coat half a plate (columns 1–6) with antigen and the other half with a control protein such as BSA. After coating and blocking, the control half of the plate is filled with 100 μL/well of PBST. To these wells, 100 μL of rescued phages from the picked clones are added, mixed, and then 100 μL are transferred to the antigen-coated wells. The plate is further developed by incubation with anti-phage and secondary antibodies 362 Almog Bitton et al.

and the appropriate substrate. This gives an important specificity control and helps avoid carrying nonspecific (“sticky”) phage clones to further validation and characterization steps. 16. Some phage-displayed antibodies lose their ability to bind the antigen when expressed in soluble formats (scFv, IgG, etc.), and some of them bind it nonspecifically. Therefore, a soluble scFv screening step is recommended to eliminate such mislead- ing phage clone hits. 17. To prevent degradation of the soluble scFv, it is strongly recommended to add a protease inhibitor cocktail to the soluble fractions, and to store them at 4 C (short-term) or at À20 C (long-term). 18. Select the detection antibody according to the tag of your library. While in our pCC phagemid there is a CBD tag, more common phagemid vectors have a hexa-histidine tag, a myc tag, or both.

Acknowledgments

In 2009, we published in “Methods in Molecular Biology” a chapter describing the construction of a large human synthetic single- chain Fv antibody library displayed on phage, where in vivo formed complementarity determining regions (CDRs) were shufﬂed com- binatorially onto germline-derived human variable region frameworks [15]. The present chapter is a revision and update of that chapter, not including the part of library construction. We are grateful to past and present members of the Benhar Lab for their contributions in optimizing the antibody phage display protocols described herein.

References

1. Benhar I (2007) Design of synthetic antibody 5. Marks JD, Hoogenboom HR, Bonnert TP, libraries. Expert Opin Biol Ther 7(5):763–779 McCafferty J, Griffiths AD, Winter G (1991) 2. Hoogenboom HR (2005) Selecting and By-passing immunization. Human antibodies screening recombinant antibody libraries. Nat from V-gene libraries displayed on phage. J Biotechnol 23(9):1105–1116 Mol Biol 222(3):581–597 3. Burton DR, Barbas CF 3rd, Persson MA, 6. Breitling F, Dubel€ S, Seehaus T, Koenig S, Chanock RM, Lerner RA (1991) A Klewinghaus I, Little M (1991) A surface large array of human monoclonal antibodies to expression vector for antibody screening. type 1 human immunodeficiency virus from Gene 104(2):147–153 combinatorial libraries of asymptomatic sero- 7. Clackson T, Hoogenboom HR, Griffiths AD, positive individuals. Proc Natl Acad Sci U S A Winter G (1991) Making antibody fragments 88(22):10134–10137 using phage display libraries. Nature 352 4. McCafferty J, Griffiths AD, Winter G, Chiswell (6336):624–628 DJ (1990) Phage antibodies: filamentous 8. Vaughan TJ, Williams AJ, Pritchard K, phage displaying antibody variable domains. Osbourn JK, Pope AR, Earnshaw JC, Nature 348(6301):552–554 McCafferty J, Hodits RA, Wilton J, Johnson Antibody Isolation From a Human Synthetic Combinatorial and Other... 363

KS (1996) Human antibodies with 17. Azriel-Rosenfeld R, Valensi M, Benhar I sub-nanomolar affinities isolated from a large (2004) A human synthetic combinatorial non-immunized phage display library. Nat Bio- library of arrayable single-chain antibodies technol 14(3):309–314 based on shuffling in vivo formed CDRs into 9. Barbas CF 3rd, Bain JD, Hoekstra DM, Lerner general framework regions. J Mol Biol RA (1992) Semisynthetic combinatorial anti- 335:177–192 body libraries: a chemical solution to the diver- 18. Nahary L, Trahtenherts A, Benhar I (2009) sity problem. Proc Natl Acad Sci U S A Isolation of scFvs that inhibit the NS3 protease 89:4457–4461 of hepatitis C virus by a combination of phage 10. Griffiths AD (1993) Production of human display and a bacterial genetic screen. Methods antibodies using bacteriophage. Curr Opin Mol Biol 562:115–132. https://doi.org/10. Immunol 5(2):263–267 1007/978-1-60327-302-2_9 11. de Haard HJ, van Neer N, Reurs A, Hufton SE, 19. Trahtenherts A, Benhar I (2009) An internaliz- Roovers RC, Henderikx P, de Bruine AP, ing antibody specific for the human asialogly- Arends JW, Hoogenboom HR (1999) A large coprotein receptor. Hybridoma (Larchmt) 28 non-immunized human fab fragment phage (4):225–233. https://doi.org/10.1089/hyb. library that permits rapid isolation and kinetic 2009.0019 analysis of high affinity antibodies. J Biol Chem 20. Ofir K, Berdichevsky Y, Benhar I, Azriel- 274(26):18218–18230 Rosenfeld R, Lamed R, Barak Y, Bayer EA, 12. Holt LJ, Bussow K, Walter G, Tomlinson IM Morag E (2005) Versatile protein microarray (2000) By-passing selection: direct screening based on carbohydrate-binding modules. Pro- for antibody-antigen interactions using protein teomics 5(7):1806–1814. https://doi.org/10. arrays. Nucleic Acids Res 28(15):E72 1002/pmic.200401078 13. Knappik A, Ge L, Honegger A, Pack P, 21. Saggy I, Wine Y, Shefet-Carasso L, Nahary L, Fischer M, Wellnhofer G, Hoess A, Wolle J, Georgiou G, Benhar I (2012) Antibody isola- Pluckthun A, Virnekas B (2000) Fully synthetic tion from immunized animals: comparison of human combinatorial antibody libraries phage display and antibody discovery via V (HuCAL) based on modular consensus frame- gene repertoire mining. Protein Eng Des Sel works and CDRs randomized with trinucleo- 25(10):539–549. https://doi.org/10.1093/ tides. J Mol Biol 296(1):57–86 protein/gzs060 14. Soderlind E, Strandberg L, Jirholt P, 22. Shimoni M, Herschhorn A, Britan-Rosich Y, Kobayashi N, Alexeiva V, Aberg AM, Kotler M, Benhar I, Hizi A (2013) The isola- Nilsson A, Jansson B, Ohlin M, Wingren C, tion of novel phage display-derived human Danielsson L, Carlsson R, Borrebaeck CA recombinant antibodies against CCR5, the (2000) Recombining germline-derived CDR major co-receptor of HIV. Viral Immunol 26 sequences for creating diverse single- (4):277–290. https://doi.org/10.1089/vim. framework antibody libraries. Nat Biotechnol 2012.0029 18(8):852–856 23. Artzy Schnirman A, Zahavi E, Yeger H, 15. Nahary L, Benhar I (2009) Design of a human Rosenfeld R, Benhar I, Reiter Y, Sivan U synthetic combinatorial library of single-chain (2006) Antibody molecules discriminate antibodies. Methods Mol Biol 525:61–80. between crystalline facets of a gallium arsenide https://doi.org/10.1007/978–1–59745- semiconductor. Nano Lett 6(9):1870–1874. 554-1_3 https://doi.org/10.1021/nl0607636 16. Jirholt P, Ohlin M, Borrebaeck CA, Soderlind 24. Benhar I, Reiter Y (2001) Phage display of E (1998) Exploiting sequence space: shuffling single-chain antibodies (scFvs). In: Colligan J in vivo formed complementarity determining (ed) Current protocols in immunology, vol regions into a master framework. Gene 215 Chapter 10.19B. John Wiley & Sons Inc, (2):471–476 New York, pp 10.19B.11–10.19B.39 Chapter 20

Screening Phage-Display Antibody Libraries Using Protein Arrays

Ricardo Jara-Acevedo, Paula Dıéz, Marıá Gonza´lez-Gonza´lez, Rosa MarıáDe´gano, Nieves Ibarrola, Rafael Gońgora, Alberto Orfao, and Manuel Fuentes

Abstract

Phage-display technology constitutes a powerful tool for the generation of specific antibodies against a predefined antigen. The main advantages of phage-display technology in comparison to conventional hybridoma-based techniques are: (1) rapid generation time and (2) antibody selection against an unlimited number of molecules (biological or not). However, the main bottleneck with phage-display technology is the validation strategies employed to confirm the greatest number of antibody fragments. The development of new high-throughput (HT) techniques has helped overcome this great limitation. Here, we describe a new method based on an array technology that allows the deposition of hundreds to thousands of phages by micro-contact on a unique nitrocellulose surface. This setup comes in combination with bioinformatic approaches that enables simultaneous affinity screening in a HT format of antibody-displaying phages.

Key words Phage display, Array, Antibodies, High-throughput screening, Antibodies, scFv

1 Introduction

Antibody phage display is an in vitro technology commonly used to select recombinant antibodies. The selection process aims at sequential enrichment of clones from an antibody phage-display library that recognizes the target of interest or antigen as the library undergoes successive rounds of selection. This means that a great number of clones have to be screened to isolate clones with specific features such as predefined binding with the antigen or its affinity. However, the selection steps often lead to the enrichment of a mixed population of specific and nonspecific ligands which are not directed against the desired target (e.g., streptavidin ligands selected efficiently by direct binding during the panning process). For this reason, the screening procedures that make it possible to

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_20, © Springer Science+Business Media LLC 2018 365 366 Ricardo Jara-Acevedo et al.

Fig. 1 (a) scFv display on phage protein III. (b) Schematic overview of a phagemid (pIT2). Abbreviations: lacZ promoter: promoter of the bacterial lac operon; RBS ribosome binding site; pelB signal peptide sequence of bacterial pectate lyase Erwinia caratovora, mediating secretion into the periplasmic space; gIII: gene coding for the phage protein III; amber: amber stop codon; myc tag: polypeptide protein tag derived from the c-myc gene; ochre: ochre stop codon

carry out the screening of the individual clones in a fast and robust way are crucial [1, 2]. In order to accelerate the screening process, phagemid vectors [3] have been developed to which an amber codon has been inserted between the antibody and pIII gene (3gp) (Fig. 1). Allow- ing, from the same construction, the antibody fragments display monovalently as well as the production of soluble antibody fragments for screening, without the need to carry out subcloning of the V gene of the antibody [4]. Taking into account the selection method employed, a large number of different screening methods have been developed. Regardless of the isolation method used, the antibodies must be expressed in a recombinant format for screening, characterization, and downstream applications. Although both complete antibody and antibody fragments (Fabs [4], Fvs, scFvs [5], and diabodies [6]) (Fig. 2) can be expressed in eukaryotic cells (e.g., yeast, mam- mal, plants, and insect cells [7, 8]), this is time-consuming and cost- intensive. For this reason, bacterial expression and particularly bacterial periplasm secretion is an inexpensive alternative and amenable for the screening and characterization of a large number of antibody variants [9]. However, a drawback for the ﬁrst analysis post-selection is that the level of antibody expression in E. coli Selection on Proteins Arrays 367

Fig. 2 The antibody fragments range from whole immunoglobulin IgG (a), Fab (b), Fv (variable heavy (VH) þ variable light (VL) domains) and single-chain variable fragment (scFv) (c), diabody (dsFv) (d), to a single-domain antibody (dAb) (e). The estimated sizes are given in kilodalton (kDa)

depends basically on its primary sequence and format (e.g., Fv vs Fab), rather than the expression system, and can be extremely varied (10 μg to 100 mg/L). Unless antibody fragment expression presents sufﬁciently high yields, the subcloning of the antibody to other expression systems [10] must be considered, such as Pichia pastoris particularly for diabodies [11, 12] or mammalian cells for transient or stable expression, of scFvs, Fabs, or whole IgGs [13, 14]. Likewise, the end use of the antibody in functional assays often requires re-cloning of the antibody genes for the expression in an IgG isotype [15]. To do this, the re-cloning methods are based on the selection of rare restriction sites in the human V genes that allows a single-step cloning process of the genes to the expression vectors. This is a very important aspect, due to the large number of clones that need to be screened.

1.1 High-Throughput Screening by ELISA of antibodies to individual antigens is relatively Screening Methods straightforward with the conventional 96 multiwell plate process but it only allows screening of a small percentage of selected clones [16]. The scaling up of this process to meet the sample volume in proteomics is considerably more complicated as manual screening requires much effort. In this regard, several efforts have been implemented in high-throughput screening (HTS) where a large number of phage antibodies can be simultaneously screened in an automated fashion. Like most selection processes, HTS has four 368 Ricardo Jara-Acevedo et al.

main elements: (a) an arrayed antibody phage-display library; (b) a configurable method of selection for automation, e.g., procedures on solid support; (c) a robotic platform consisting of a colony picker, liquid handling robot, and incubators; (d) a computer sta- tion and a software to manage data [17, 18]. In attempts to automate the selection, especially when you are working with various antigens in parallel, 96 multiwell plates are considered the standard format. Miniaturization of the process using 384 multiwell plates has also been tried, where high expression levels of antibody in E.coli are essential. A possible drawback is the possibility of using non-purified phage antibodies or antibody fragments present in the crude supernatant or periplasmic space extract in the screening assays. In addition, the ELISA protocol is well suited for automation as described in different studies utilizing robotics-based HTS [19–22]. These automated systems are configured to screen tens of thousands of clones per day to target antigens in various formats, facilitating the rapid information generation about specific antibody repertoire present in the library. A different HTS approach focuses on the use of antibody fragment filter screening secreted from individual bacterial colonies. This method is based on the use of two membranes and helps to circumvent the more time-consuming phage-display stage that is normally required for affinity selection of antigen-positive antibody clones. Antibody fragments are detected in a filter assay where bacterial colonies are grown on a master filter that is in contact with a second, antigen-coated filter. This allows for antibody fragments to diffuse onto the second filter and bind the antigen directly and specifically prior to detection with a monoclonal antibody- enzyme conjugated [23, 24]. Colonies on the first membrane filter remain viable and can be re-grown for new selection rounds or antibody purification. This type of procedure has been carried out in a multitude of studies, allowing the simultaneous analysis of thousands of antibody clones and, more importantly, can be used with crude detergent-solubilized cell extracts, permitting discovery of antibodies that bind integral membrane proteins present in heterogeneous mixtures of antigens [25]. A modification of this procedure is the termed capture lift, consisting of immobilization of greater quantities of antibody fragments with a decreased binding of unrelated host proteins, resulting in a more sensitive plaque lift assay by subtractive colony to recognize unrelated proteins [26, 27]. Despite filter-based screening techniques allowing screening of a large number of clones against a single antigen, it is difficult to identify the genuine positive clones and isolate them from neigh- boring negative clones, because of the highly dense propagation rate of bacteria. Furthermore, it is difficult to rescue viable clones from bacterial colonies that have been induced for antibody expression. Finally, these techniques are not suitable for screening against Selection on Proteins Arrays 369

different antigens, due to the difficulty of comparing duplicate filter signals [28]. For these reasons, development of protein and antibody microarrays is particularly useful for HT analysis of antibodies based on specificity and affinity. It can also facilitate the detection and quantification of thousands of binding reactions simultaneously with a small amount of sample. With appropriate instrumentation like a picking/spotting robot, protein microarrays can be a common resource for HT screening of receptor-ligand interactions. In this context, a large number of studies have used microarrays for HTS of antibodies. A direct screening bypasses the selection process by using array of proteins [29], while in a secondary screening a more exhaustive analysis of the positive clones by other techniques [30]. The use of a phage-display array also allows easier identification of different antibodies fragments against known proteins [31].

1.2 Phage-Display The development of protein and antibody microarrays is particu- Screening in Array larly useful for HT analysis of antibody specificity and affinity, Format facilitating detection and quantification of thousands of binding reactions in a simultaneous manner and with a small amount of sample. With appropriate equipment, a picking/spotting robot, protein microarrays should be a common resource for screening HT receptor-ligand interactions. The open microarray architecture is one of the major advantages of microarray technology allowing several antibodies to be screened on several antigens at the same time which requires complicated liquid handling. This is highlighted by the multiple spotting techniques (MIST), which comprises immobilization of a binder onto a surface and subsequent spotting of the second compound on the same spot, on the top of the immobilized binder [32, 33]. A major advantage of microarray technology is the production of functional proteins with methods such as the protein in situ microarray (PISA). PISA allows protein microarrays to be rapidly generated in a single step, directly from DNA templates, by cell-free protein expression with simultaneous in situ immobilization on the array surface [34] or with the high-density self- assembling protein microarray, based on the nucleic programmable acid protein array (NAPPA) concept. The NAPPA concept displays thousands of proteins that are produced and captured in situ from immobilized cDNA templates devoid of protein purification [35]. However, for expression, NAPPA requires plasmids containing the gene of interest as a GST-fusion protein. This necessitates time-consuming cloning of cDNAs, besides the immobilization of the plasmid. To overcome these problems, MIST based a method that requires the spotting of a DNA template in a first spotting and the transfer of a cell-free transcription and translation mixture on the top of the very same spot in a second spotting run [36]. 370 Ricardo Jara-Acevedo et al.

Furthermore, different studies have shown that the possible lack of specificity of the capture reagents limits the use of microarrays since the cross-reactivities of some antibodies to unrelated proteins prevent the use of protein microarrays for specific antibody screening against whole protein libraries [37]. Likewise, assays have been defined for determining antibody affinity and specificity, on the basis that the amount of an antibody captured by an immobilized antigen is directly dependent on the affinity constant governing the binding reaction and the concentration of antibody present in the sample [38]. Recent developments in HTS technologies have had a profound influence on the widespread use of NGS for antibody sequence analysis from phage-display libraries. This allows rapid and high-throughput characterization of the library in terms of diversity and quality, as well as being applied for in vitro selection and screening [39–41]. In this sense, the ability to obtain sequence information about all clones in each round would provide a virtually complete analysis of the outcome selection process [42]. Addi- tionally, NGS-based library phage-display characterization is more cost effective than the traditional cloning due to fewer selection rounds required, allowing identifying positive phages after a first selection round and enhancing significantly the discovery of new clones and restricting the number of false positives [43]. In the application of sequence frequency information, it is possible to rescue antibody clones that are normally missed by traditional in vitro screening techniques. Due to the accuracy, running time, diversity coverage, and cost-effectiveness, a tandem approach to screening involving quantitative real-time PCR quantification and next-generation DNA sequencing could be a gold standard for the antibody phage-display screening in the future [44]. In this context, we have demonstrated that affinity screening of antibody- displaying phages based on protein arrays containing scFv fragments as capture agents immobilized onto a hydrophobic nitrocellulose surface is robust, relatively fast and is able to detect up to 80% of the clones considered positive by ELISA [8, 31]. Herein, we describe a detailed protocol for preparing antibody-displaying phage microarrays, including the optimization of a number of crucial steps that would facilitate the incorporation of this approach by other researchers in the field.

2 Materials

In this protocol, we employed the Tomlinson I þ J libraries (MRC HGMP Resource Centre, University of Cambridge, UK) that were built in the pIT2 vector (derived from pHEN1). The pIT2 vector contains a pelB promoter located upstream of the VH-(G4S)3-VL insert followed by His and Myc tags, a stop codon, and the G3P Selection on Proteins Arrays 371

anchor gene. These libraries are based on a single-human framework for VH and Vκ containing 18 positions in the CDRH2, CDRH3, CDRL2, and CDRL3 regions randomized with NNK codons to achieve a diversity with ~1.4 Â 108 unique clones.

2.1 Immobilization 1. N-hydroxysuccinimide (NHS)-biotin solution (Sigma-Aldrich, of Antigen by St. Louis, USA): 10 mg NHS-biotin in 1 mL DMSO. Biotinylation 2. Streptavidin-agarose (Sigma-Aldrich, St. Louis, USA). 3. Neutravidin agarose (Thermo Fisher Scientiﬁc, Waltham, USA).

4. Phosphate-buffered saline (PBS) pH 7.4: 3.6 g Na2HPO4, 0.2 g KCl, 0.24 g KH2PO4, 8 g NaCl in 1 L. 5. MT blocking buffer: PBS. 5% (w/v) marvel milk powder, 2% (v/v) Tween20.

2.2 Binding 1. Empty disposable PD-10 columns (GE healthcare Life and Elution of Phage Sciences, Chicago, USA). 2. Phosphate-Buffered Saline Tween-20 (PBST): 0.1% Tween-20 to PBS. 3. Tris-buffered saline, with Calcium (TBSC): 10 mM Tris pH 7.4, 137 mM NaCl, 1 mM CaCl2. 4. Trypsin solution (Sigma-Aldrich, St. Louis, USA): 10 mg/mL trypsin powder in TBSC (trypsin stock). Freeze in 20 mL aliquots in liquid nitrogen. For the experiment, 100 mL trypsin stock in 10 mL TBSC (trypsin solution). 5. M9 minimal medium glucose plates: 15 g of agar in 800 mL of deionized water. Autoclave. Cool down to 50 C and add 200 mL of 5_M9 salts, 10 mL of 20%(wt/vol) glucose, 1 mL of 1 M MgSO4, 100 mL of 1 M CaCl2, and 1 mL of 1 mg mL_1 VitB1. For M9 salts (5_ solution), add 64 g of Na2HPO4, 15 g of KH2PO4,5gofNH4Cl, and 2.5 g of NaCl to 1 L of deionized water and autoclave. Pour plates. 6. 2ÂTY medium; 6 g of bacto-tryptone, 10 g of yeast extract, and 5 g of NaCl in 1 L of deionized water. Autoclave. Cool to room temperature (25 C) and add antibiotic solutions and glucose solution as required. 7. TYE ampicillin glucose agar plates: 15 g of agar, 8 g of NaCl, 10 g of bacto-tryptone, and 5 g of yeast extract in 800 mL of deionized water. Autoclave. Cool down to 50 C and add 1 mL of ampicillin solution and 200 mL of glucose solution. Pour plates. 8. Ampicillin solution (Sigma-Aldrich, St. Louis, USA): Dissolve ampicillin powder at 100 mg/mL in deionized water. Filter through 0.2 mM ﬁlter. Aliquot in 1 mL portions. 372 Ricardo Jara-Acevedo et al.

9. Kanamycin solution (Sigma-Aldrich, St. Louis, USA): Dissolve kanamycin powder at 50 mg/mL in deionized water. Filter through 0.2 mM filter. Aliquot in 1 mL portions. Thawed aliquots should be recently diluted 1000-fold into medium or agar. 10. Glucose solution: 20% glucose solution. 200 g of glucose in 1 L of deionized water. Filter through 0.2 mM filter. 11. PEG solution (Sigma-Aldrich, St. Louis, USA): 20% PEG, 2.5 M NaCl. 100 g of PEG-6000 and 73 g of NaCl in 500 mL of deionized water. Filter through 0.2 mM filter.

2.3 Array Nitrocellulose-coated FAST slides (Schleicher & Schuell Whatman, Preparation Sanford, USA).

2.4 Quality Control 1. Anti-cMyc antibody (Sigma-Aldrich, St. Louis, USA). of the Microarray 2. Anti-mouse IgG HRP (Sigma-Aldrich, St. Louis, USA). Printing 3. Tyramide signal ampliﬁcation (TSA) reagent (PerkinElmer, Waltham, USA). 4. Tris-buffered saline with Tween 20 (TBST): Tris-buffered saline, 0.1% Tween 20.

2.5 Functional Assay 1. Super Block™ Blocking Buffer in PBS (Thermo Fisher Scien- of scFvs Phage- tiﬁc, Waltham, USA). Display Library 2. Streptavidin-Cy3 (Sigma-Aldrich, St. Louis, USA). in Array Format 3. Corning hybridization chambers (Sigma-Aldrich, St. Louis, USA)

3 Methods

The following protocols describe HT antibody selection process from phage-display libraries, paying attention to characteristics such as sensitivity, reproducibility, and easy implementation of the method in other laboratories.

3.1 Growth The growth and puriﬁcation of phage antibody repertoire were and Puriﬁcation carried out as reported previously [28, 45]. of Phage Antibody Repertoire

3.2 Immobilization 1. Use 500 mg of protein at 1 mg/mL in PBS, pH 7.4 (see Note 1). of Antigen by 2. Add 5 mL of 10 mg/mL of NHS-biotin solution and mix well. Biotinylation 3. Incubate at room temperature (RT) for 30 min. 4. Add 50 mL of glycine solution to stop the reaction and mix well. Selection on Proteins Arrays 373

5. Incubate at RT for 10 min. Check molecular weight by mass spectrometry. Use 10 mg of non-biotinylated and biotinylated protein for analysis (see Note 2). 6. Dialyze by disposable PD MidiTrap G-25 using gravity protocol at least three times. 7. After dialysis, determine protein concentration in a UV-visible spectrophotometer at 280 nm. 8. Wash twice with PBS a total of 100 mL of 4% cross-linked agarose microbeads (4BCL) coated with streptavidin agarose resin: centrifuge at 376 Â g for 5 min at 4 C in an Eppendorf tube and carefully remove PBS. Streptavidin agarose should only be used in the ﬁrst round of selection, whereas neutravidin agarose is used in all subsequent rounds. This prevents the selection of streptavidin binders. 9. Add 50 mg of protein in PBS to prewashed resin. 10. Incubate for 30 min. 11. Wash twice with PBS (as in step 8). 12. Add additional resin (2Â PBS-washed) up to 2 mL to increase resin volume to a manageable level. For this purpose, streptavidin agarose (or neutravidin agarose in subsequent selection rounds), G25 sepharose or 0.2 mm glass beads can be used. 13. Add 2 mL of MT blocking buffer. 14. Incubate for 30 min. After incubation resin is ready for phage selection, next step.

3.3 Phage Selection 1. Add 5 Â 1012 phages to 6 mL of MT buffer and incubate at RT for 30 min. 3.3.1 Binding and Elution 2. Mix phage and antigen bound to streptavidin agarose resin and of Phages (See Note 3) rotate at 4 C overnight in a 15 mL Falcon tube. 3. Pour resin into empty disposable PD-10 columns. 4. Wash column 10Â with 10 mL of PBST buffer and twice with 10 mL of PBS buffer. 5. Close the column with plug. Add 2 mL of trypsin solution and incubate for 1 h at RT. 6. Remove the plug and collect the ﬂow-through. Add an additional 1 mL of trypsin solution, collect ﬂow-through and combine.

3.3.2 Infection of TG1 1. Streak TG1 bacteria from glycerol stock on an M9 minimal Bacteria with Eluted Phage medium plate and incubate for 36 h at 37 C. Grow overnight and Subsequent Rounds culture in 5 mL of 2-TY medium from a single colony at 37 1C of Selection and 250 rpm. Dilute culture 100-fold into 2-TY medium. Grow at 37 C and 250 rpm. 374 Ricardo Jara-Acevedo et al.

2. Take 1.75 mL of TG1 at an OD 600 of 0.4 and add 250 μLof the eluted phage (the remaining 250 μL should be stored at 4 C). Incubate for 30 min at 37 C in a water bath without shaking. 3. Spot 10 μL, 10 μL of a 1:102 dilution and 10 μL of a 1:104 dilution on TYE plates containing 100 μg/mL ampicillin and 1% glucose and grow overnight at 37 C to titer the phage.

3.3.3 Subsequent Scrape cells from agar plates using 5 mL of 2-TY medium per plate Rounds of Selection (See and a glass spreader. Mix cells thoroughly by vortexing in a 50 mL Note 4) Falcon tube. Dilute with 500 mL of 2-TY medium supplemented with 4% (wt/vol) glucose and 100 mg/mL of ampicillin to an OD600 of 0.1. Grow to OD60 0 ¼ 0.5 at 37 C and 250 rpm. Then infect with helper phage, grow overnight, and purify phage by PEG puriﬁcation as reported previously [28, 45].

3.4 Screening After three rounds of selection, individual colonies from the dilu- of Clones by tion series can be tested for antigen binding. In this step, the phage Monoclonal Phage- clones can be tested by the phage-display arrays. Display Array

3.4.1 Phage Supernatant 1. Pick colonies using sterile toothpicks or pipette tips into a Production 96-well round-bottomed plate containing 200 mL of 2-TY medium supplemented with 100 mg/mL of ampicillin and 4% (wt/vol) glucose. Pick positive and negative control clones into A1/A2 wells from freshly streaked TYE plates supplemented with 100 mg/mL of ampicillin and 4% (wt/vol) glucose. Grow overnight at 37 C and 250 rpm (see Note 5). 2. The next day, a fresh 96-well round-bottomed plate containing 200 mL of 2-TY medium supplemented with 100 mg/mL of ampicillin and 4% (wt/vol) glucose should be inoculated with 5 mL of the overnight culture. The recently inoculated plate should be shaken at 37 C and 250 rpm for 3 h. 3. After 3 h, add 50 mL of 2-TY medium supplemented with 4 Â 108 KM13 helper phages to each well. Mix by gentle agitation. Incubate the plate at 37 C without shaking for 1 h. Spin at 3200 Â g in a plate centrifuge for 10 min at room temperature. Discard the supernatant by quickly inverting the plate. Resuspend pellets in 200 mL of 2-TY medium supplemented with 100 mg/mL of ampicillin, 50 mg/mL of kanamycin, and 0.1% (wt/vol) glucose by gentle agitation. Grow overnight at 25 C and 250 rpm for 16–24 h. 4. The next day, spin the plate at 3200 Â g for 10 min at room temperature in a plate centrifuge and transfer the supernatant to a new 96-well plate and store at 4 C. The phage clones can now be tested by phage-display array. Selection on Proteins Arrays 375

3.4.2 Array Preparation 1. Spin 96-well plates at 3200 Â g for 10 min at RT in a plate centrifuge and transfer the phage supernatant to new 384-well plates and store at 4 C until use. Reformatting 96-well plates into 384-well plates is described in Fig. 3. 2. Centrifuge 384-well round-bottom plate at 300 Â g for 17 min at 4 C Spot ~25 nanoliters of each clone on the nitrocellulose- coated slides by using a 16-pin custom-built micro-contact printer (MicroGrid II, Biorobotics, De Meern, The Nether- lands) with a 200 μm steel pin diameter in a 45–50% humidiﬁed environment yielding protein spots of 200 μm(see Note 6). In order to avoid cross-contamination and cross-talking effects, spots were printed at 375 μm apart from each other in two sub-arrays per slide. Steel pins were washed three times with 1% (w/v) SDS and 96% (v/v) ethanol. The array design is shown in Fig. 3 (see Note 7). After printing arrays were kept in a vacuum desiccator at RT overnight.

Fig. 3 Graphic description of the array construction, from the 96-well plates to the ﬁnal slide 376 Ricardo Jara-Acevedo et al.

3.4.3 Quality Control 1. Block the array surface with Super Block™ Blocking buffer for of the Microarray Printing 1 h at RT. 2. Wash the arrays with distilled water. 3. Add anti-cMyc primary antibody diluted in blocking solution (dilution 1:200). 4. Incubate at RT for 1 h in a humidified chamber. 5. Wash three times for 5 min with PBS 1Â. 6. Add anti-mouse IgG HRP-linked, as the secondary antibody, diluted in blocking solution (dilution 1:200). 7. Incubate at RT for 1 h in a humidified chamber. 8. Wash three times for 5 min with PBS 1Â. 9. Add fluorophore-conjugated TSA® Plus amplification reagent (dilution 1:50). 10. Apply 200 μL per slide and incubate for 10 min at RT in a humidified chamber, protected from light. 11. Wash arrays with PBS 1Â twice for 5 min, with gentle agitation and protect from light. 12. Wash each array with distilled water. Detect the signal at 532 nm using a scanner (GenePix 4000B, Axon Instruments, Molecular Devices, NJ/USA).

3.4.4 Functional Assay 1. Block printed arrays with Super Block™ Blocking buffer for 1 h at of scFvs Phage-Display mild stirring and RT. After this period, wash with distilled water. Library in Array Format 2. Add 2 mL of a solution containing the biotinylated query protein of interest diluted in milk solution (5% (w/v) skimmed milk in PBST supplemented with 0.2% (v/v) Tween®20) to each slide. Incubate each slide in an individual slide incubation chamber overnight in the dark at RT (Fig. 4). 3. After the incubation time, wash the arrays twice with milk solution for 5 min with mild stirring at 60 rpm, followed by three washes with distilled water (see Note 8). 4. To reveal the presence of scFv phage-display, add labeled streptavidin-Cy3 (1:1000 dilution in Super Block® Blocking buffer) and incubate for 10 min. 5. After the incubation time, wash slides three times with distilled water for 5 min and mild-stirring. Then, dry each slide with ﬁltered compressed air. 6. Detect the signal by scanning each slide with a scanner (Scanner GenePix 4000B. Axon Instruments, Molecular Devices, NJ/USA) applying Cy3 settings. The image is analyzed by GenePix®Pro 4.0.1.27 software (Molecular Devices, Sunny- vale/CA, USA). The GenePix Array List ﬁle (.gal) generated is used as a reference grid to locate and identify the scFvs clone spots on the surface of the arrays. Selection on Proteins Arrays 377

Fig. 4 Graphic description of array display from 384-well plates to array display in multiple sub-arrays including number of replicates

3.5 Data Processing 1. Perform background noise correction from microarray assay by subtracting the absorbance and intensities values from the control spots. 2. Calculate mean and standard deviation (SD) of the data set. 3. Normalize according to Dı´ez et al. [46].

4 Notes

1. Biotinylation reaction requires lysine and arginine residues accessible in the protein. Avoid buffers that contain amines as the presence of free amines will quench the reaction. Amine- containing buffers include Tris buffers and buffers with glycine. 2. Biotinylation of proteins should be moderate, as excess biotinylation may alter epitopes and induce aggregation. 3. The ﬁrst round of selection is the most important, as any loss of diversity is ampliﬁed in subsequent rounds. 4. Neutravidin agarose (rather than streptavidin agarose) should be used in subsequent selection rounds to prevent the enrichment of streptavidin binders. 378 Ricardo Jara-Acevedo et al.

5. During phage supernatant production, the 96-well plate can be secured inside the box with pieces of foam or with paper towels. The plate lid should be removed; however, the plastic box should be closed. Carefully keep the box horizontal to avoid spills and cross-contamination. 6. Protein arrays for phage-display screening must be constructed under controlled humidity and temperature to ensure the reproducibility of the assays. 7. Including several replicates in the array of the same phage- display antibody will increase the robustness of the technique. 8. Washing steps during protein array assays are crucial for obtaining optimal results.

Acknowledgments

We gratefully acknowledge ﬁnancial support from the Carlos III Health Institute of Spain (FIS PI14/01538), Fondos FEDER (EU), Junta Castilla-Leon (BIO/SA07/15), and Fundacio´n Solo´r- zano FS/23-2015. The proteomics Unit belongs to ProteoRed- ISCIII, PRB2-ISCII, supported by grant PT13/001. P.D. is supported by a JCYL-EDU/346/2013 Ph.D. scholarship.

References

1. Smith G (1985) Filamentous fusion phage: 348:552–554. https://doi.org/10.1038/ novel expression vectors that display cloned 348552a0 antigens on the virion surface. Science 6. McGuinness B, Walter G, FitzGerald K, 228:1315–1317. https://doi.org/10.1126/ Schuler P, Mahoney W, Duncan A, Hoogen- science.4001944 boom H (1996) Phage diabody repertoires for 2. Winter G, Griffiths A, Hawkins R, Hoogen- selection of large numbers of bispecific anti- boom H (1994) Making antibodies by phage body fragments. Nat Biotechnol display technology. Annu Rev Immunol 14:1149–1154. https://doi.org/10.1038/ 12:433–455. https://doi.org/10.1146/ nbt0996-1149 annurev.iy.12.040194.002245 7. Frenzel A, Hust M, Schirrmann T (2013) 3. Breitling F, Dubel€ S, Seehaus T, Expression of recombinant antibodies. Front Klewinghaus I, Little M (1991) A surface Immunol 4:217. https://doi.org/10.3389/ expression vector for antibody screening. fimmu.2013.00217 Gene 104:147–153. https://doi.org/10. 8. Waldmann T (2003) Immunotherapy: past, 1016/0378-1119(91)90244-6 present and future. Nat Med 9:269–277. 4. Hoogenboom H, Griffiths A, Johnson K, https://doi.org/10.1038/nm0303-269 Chiswell D, Hudson P, Winter G (1991) 9. Skerra APluckthun A (1988) Assembly of a Multi-subunit proteins on the surface of fila- functional immunoglobulin Fv fragment in mentous phage: methodologies for displaying Escherichia coli. Science 240:1038–1041. antibody (fab) heavy and light chains. Nucleic https://doi.org/10.1126/science.3285470 Acids Res 19:4133–4137. https://doi.org/10. 10. Verma R, Boleti E, George A (1998) Antibody 1093/nar/19.15.4133 engineering: comparison of bacterial, yeast, 5. McCafferty J, Griffiths A, Winter G, Chiswell insect and mammalian expression systems. J D (1990) Phage antibodies: filamentous phage Immunol Methods 216:165–181. https:// displaying antibody variable domains. Nature doi.org/10.1016/s0022-1759(98)00077-5 Selection on Proteins Arrays 379

11. Eldin P, Pauza M, Hieda Y, Lin G, 20. Turunen L, Takkinen K, So¨derlund H, Pulli T Murtaugh M, Pentel P, Pennell C (1997) (2009) Automated panning and screening pro- High-level secretion of two antibody single cedure on microplates for antibody generation chain Fv fragments by Pichia pastoris. J Immu- from phage display libraries. J Biomol Screen nol Methods 201:67–75. https://doi.org/10. 14:282–293. https://doi.org/10.1177/ 1016/s0022-1759(96)00213-x 1087057108330113 12. FitzGerald K, Holliger P, Winter G (1997) 21. Crameri RKodzius R (2012) The powerful Improved tumour targeting by disulphide sta- combination of phage surface display of bilized diabodies expressed in Pichia pastoris. cDNA libraries and high throughput screen- Protein Eng Des Sel 10:1221–1225. https:// ing. Comb Chem High Throughput Screen doi.org/10.1093/protein/10.10.1221 4:145–155. https://doi.org/10.2174/ 13. Persic L, Roberts A, Wilton J, Cattaneo A, 1386207013331237 Bradbury A, Hoogenboom H (1997) An 22. Hallborn J, Carlsson R (2002) Automated integrated vector system for the eukaryotic screening procedure for high-throughput gen- expression of antibodies or their fragments eration of antibody fragments. Biotechniques after selection from phage display libraries. Suppl:30–37 Gene 187:9–18. https://doi.org/10.1016/ 23. Skerra A, Dreher M, Winter G (1991) Filter s0378-1119(96)00628-2 screening of antibody Fab fragments secreted 14. Helfrich W, Haisma H, Magdolen V, Luther T, from individual bacterial colonies: specific Bom V, Westra J, van der Hoeven R, detection of antigen binding with a Kroesen B, Molema G, de Leij L (2000) A two-membrane system. Anal Biochem rapid and versatile method for harnessing scFv 196:151–155. https://doi.org/10.1016/ antibody fragments with various biological 0003-2697(91)90131-c effector functions. J Immunol Methods 24. Dreher M, Gherardi E, Skerra A, Milstein C 237:131–145. https://doi.org/10.1016/ (1991) Colony assays for antibody fragments s0022-1759(99)00220-3 expressed in bacteria. J Immunol Methods 15. Jostock T (2004) Rapid generation of func- 139:197–205. https://doi.org/10.1016/ tional human IgG antibodies derived from 0022-1759(91)90189-m Fab-on-phage display libraries. J Immunol 25. Watkins J, Beuerlein G, Wu H, McFadden P, Methods 289:65–80. https://doi.org/10. Pancook J, Huse W (1998) Discovery of 1016/s0022-1759(04)00119-x human antibodies to cell surface antigens by 16. Lou J, Marzari R, Verzillo V, Ferrero F, Pak D, capture lift screening of phage-expressed anti- Sheng M, Yang C, Sblattero D, Bradbury A body libraries. Anal Biochem 256:169–177. (2001) Antibodies in haystacks: how selection https://doi.org/10.1006/abio.1997.2523 strategy influences the outcome of selection 26. Wu H, Pancook JD, Greg Beuerlein H (2002) from molecular diversity libraries. J Immunol Cloning, isolation and characterization of Methods 253:233–242. https://doi.org/10. human tumor in situ monoclonal antibodies. 1016/s0022-1759(01)00385-4 Cancer Immunol Immunother 51:79–90. 17. Hoogenboom H (2005) Selecting and screen- https://doi.org/10.1007/s00262-001-0258- ing recombinant antibody libraries. Nat Bio- y technol 23:1105–1116. https://doi.org/10. 27. Robert R, Jacobin-Valat M, Daret D, Miraux S, 1038/nbt1126 Nurden A, Franconi J, Clofent-Sanchez G 18. Walter G, Konthur Z, Lehrach H (2012) High- (2006) Identification of human scFvs targeting throughput screening of surface displayed gene atherosclerotic lesions: selection by single products. Comb Chem High Throughput round in vivo phage display. J Biol Chem Screen 4:193–205. https://doi.org/10. 281:40135–40143. https://doi.org/10. 2174/1386207013331228 1074/jbc.m609344200 19. Krebs B, Rauchenberger R, Reiffert S, 28. de Wildt RM, Mundy CR, Gorick BD, Tom- Rothe C, Tesar M, Thomassen E, Cao M, linson IM (2000) Antibody arrays for high- Dreier T, Fischer D, Ho¨ß A, Inge L, throughput screening of antibody-antigen Knappik A, Marget M, Pack P, Meng X, interactions. Nat Biotechnol 18(9):989–994. Schier R, So¨hlemann P, Winter J, Wo¨lle J, https://doi.org/10.1038/79494 Kretzschmar T (2001) High-throughput gen- 29. Holt L (2000) By-passing selection: direct eration and engineering of recombinant human screening for antibody-antigen interactions antibodies. J Immunol Methods 254:67–84. using protein arrays. Nucleic Acids Res 28: https://doi.org/10.1016/s0022-1759(01) e72. https://doi.org/10.1093/nar/28.15. 00398-2 e72 380 Ricardo Jara-Acevedo et al.

30. Kibat J, Schirrmann T, Knape M, Helmsing S, microarrays for antibody profiling: specificity Meier D, Hust M, Schro¨der C, Bertinetti D, and affinity determination on a chip. Proteo- Winter G, Pardes K, Funk M, Vala A, Giese N, mics 5:2402–2411. https://doi.org/10. Herberg F, Dubel€ S, Hoheisel J (2016) Utilisa- 1002/pmic.200401299 tion of antibody microarrays for the selection 39. Fischer N (2011) Sequencing antibody reper- of specific and informative antibodies from toires: the next generation. MAbs 3(1):17–20 recombinant library binders of unknown qual- 40. Glanville J, Zhai W, Berka J, Telman D, ity. New Biotechnol 33:574–581. https://doi. Huerta G, Mehta G, Ni I, Mei L, Sundar P, org/10.1016/j.nbt.2015.12.003 Day G, Cox D, Rajpal A, Pons J (2009) Precise 31. Dıéz P, Jara-Acevedo R, Gonza´lez- determination of the diversity of a combinato- Gonza´lez M, Casado-Vela J, Dasilva N, rial antibody library gives insight into the Lećrevisse Q, Bartolome´ R, Claros J, human immunoglobulin repertoire. Proc Natl Gonza´lez A, Lo´pez R, Orfao A, Fuentes M Acad Sci 106:20216–20221. https://doi.org/ (2015) High-throughgput phage-display 10.1073/pnas.0909775106 screening in array format. Enzym Microb 41. Turner K, Naciri J, Liu J, Anderson G, Technol 79-80:34–41. https://doi.org/10. Goldman E, Zabetakis D (2016) Next- 1016/j.enzmictec.2015.06.016 generation sequencing of a single domain anti- 32. Angenendt P, Glo¨kler J, Konthur Z, body repertoire reveals quality of phage display Lehrach H, Cahill D (2003) 3D protein micro- selected candidates. PLoS One 11:e0149393. arrays: performing multiplex immunoassays on https://doi.org/10.1371/journal.pone. a single Chip. Anal Chem 75:4368–4372. 0149393 https://doi.org/10.1021/ac034260l 42. Ravn U, Gueneau F, Baerlocher L, Osteras M, 33. Angenendt P, Wilde J, Kijanka G, Baars S, Desmurs M, Malinge P, Magistrelli G, Cahill D, Kreutzberger J, Lehrach H, Farinelli L, Kosco-Vilbois M, Fischer N Konthur Z, Glo¨kler J (2004) Seeing better (2010) By-passing in vitro screening--next through a MIST: evaluation of monoclonal generation sequencing technologies applied to recombinant antibody fragments on microar- antibody display and in silico candidate selec- rays. Anal Chem 76:2916–2921. https://doi. tion. Nucleic Acids Res 38:e193–e193. org/10.1021/ac035357a https://doi.org/10.1093/nar/gkq789 34. He M (2001) Single step generation of protein 43. ‘t Hoen P, Jirka S, ten Broeke B, Schultes E, arrays from DNA by cell-free expression and in Aguilera B, Pang K, Heemskerk H, Aartsma- situ immobilisation (PISA method). Nucleic Rus A, van Ommen G, den Dunnen J (2012) Acids Res 29:e73. https://doi.org/10.1093/ Phage display screening without repetitious nar/29.15.e73 selection rounds. Anal Biochem 35. Ramachandran N, Raphael J, Hainsworth E, 421:622–631. https://doi.org/10.1016/j.ab. Demirkan G, Fuentes M, Rolfs A, Hu Y, 2011.11.005 LaBaer J (2008) Next-generation high-density 44. Dias-Neto E, Nunes D, Giordano R, Sun J, self-assembling functional protein arrays. Nat Botz G, Yang K, Setubal J, Pasqualini R, Arap Methods 5:535–538. https://doi.org/10. W (2009) Next-generation phage display: inte- 1038/nmeth.1210 grating and comparing available molecular 36. Angenendt P (2006) Generation of high den- tools to enable cost-effective high-throughput sity protein microarrays by cell-free in situ analysis. PLoS One 4:e8338. https://doi.org/ expression of unpurified PCR products. Mol 10.1371/journal.pone.0008338 Cell Proteomics 5:1658–1666. https://doi. 45. Lee C, Iorno N, Sierro F, Christ D (2007) org/10.1074/mcp.t600024-mcp200 Selection of human antibody fragments by 37. Lueking A, Horn M, Eickhoff H, Bussow€ K, phage display. Nat Protoc 2:3001–3008. Lehrach H, Walter G (1999) Protein microar- https://doi.org/10.1038/nprot.2007.448 rays for gene expression and antibody screen- 46. Dıéz P, Dasilva N, Gonza´lez-Gonza´lez M, ing. Anal Biochem 270:103–111. https://doi. Matarraz S, Casado-Vela J, Orfao A, Fuentes org/10.1006/abio.1999.4063 M (2012) Data analysis strategies for protein 38. Poetz O, Ostendorp R, Brocks B, Schwenk J, microarrays. Microarrays 1:64–83. https://doi. Stoll D, Joos T, Templin M (2005) Protein org/10.3390/microarrays1020064 Chapter 21

Antibody Selection on FFPE Tissue Slides

Andre ten Haaf, Stefan Gattenlo¨hner, and Mehmet Kemal Tur

Abstract

Standard antibody phage-display panning uses purified proteins, antigen-transfected cells, or tumor cell lines as target structure to generate specific antibodies. Here, we describe a method for the selection of specific antibodies by phage panning against routine formalin-fixed paraffin-embedded (FFPE) tissue biopsies immobilized on glass slides. Selected antibody fragments recognize disease-associated antigens in its native conformation, suitable for the development of targeted diagnostic and therapeutic agents.

Key words On-slide selection, Formalin-fixed paraffin-embedded (FFPE) tissue-specific antibodies, Phage-display technology

1 Introduction

Monoclonal antibodies play a major role in diagnostic immunohistochemistry and targeted cancer therapy [1, 2]. Therefore, the process of identification of target-specific antibodies is very important. An effective method for the isolation of new, highly specific antibodies is the phage-display technology, in which a repertoire of different antibody fragments displayed on a filamentous bacteriophage is used for the selection of target-specific antibodies [3]. Standard selection strategies use recombinant proteins, antigen-transfected cells, or tumor cell lines as target structure. However, the functional expression and purification of recombinant proteins could be very challenging and a suitable target cell line is not always available. Therefore, we and others [3–6] have established a phage antibody selection strategy that can be applied directly to FFPE tissue biopsies routinely taken from patients and immobilized on glass slides. Formalin fixation and paraffin embed- ding is a worldwide standard preservation method for pathological tissue samples [7]. FFPE tissues can be stored at room temperature for many years thus avoiding the need for liquid nitrogen or a freezer, and this provides an advantage compared to the storage of frozen sections. In FFPE tissue the target cells remain in their

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_21, © Springer Science+Business Media LLC 2018 381 382 Andre ten Haaf et al.

original environment with many of their original clinical characteristics, thus making tissue sections an interesting source for antibody selection [6]. There is no need for the production and purification of recombinant proteins or cultivation of cells for panning procedure. Furthermore, all target-binding FFPE-selected antibodies are able to penetrate paraffin-embedded tissue, which makes them suitable for in situ immunohistochemical diagnosis. However, FFPE tissues contain cross-linked and partial denatured proteins, due to the reactions of the formaldehyde with functional groups of the proteins and can reduce accessibility of antigenic epitopes [8, 9]. Therefore, different antigen retrieval procedures should be applied before antibody selection to enhance antigen accessibility and increase selection efficacy. Using FFPE tissue slides for selection and by combination of negative and positive panning strategy, we were able to generate a panel of cell surface-binding antibodies highly specific against small-cell lung cancer (SCLC) cells with therapeutic and diagnostic potential (see Fig. 1)[4].

2 Material

All solutions are prepared in ultrapure water (deionized water, 18 MΩ-cm at 25 C) and stored at room temperature unless indicated otherwise.

2.1 Phage Rescue 1. 2ÂYT medium: 16 g tryptone, 10 g yeast extract, 5 g NaCl, add water to a volume of 1 L. Autoclave the medium before usage (see Note 1).

2. 2ÂYTAmp/Gluc medium: 2ÂYT medium, 100 μg/mL ampicillin, 1% glucose.

3. 2ÂYTAmp/Kana/IPTG medium: 2ÂYT medium, 100 μg/mL ampicillin, 50 μg/mL kanamycin, 0.25 mM IPTG. 4. 2ÂYT agar: 16 g tryptone, 10 g yeast extract, 5 g NaCl, 15 g agar, add water to a volume of 1 L. Autoclave agar before usage.

5. 2ÂYTAmp/Gluc agar: 2ÂYT agar, 100 μg/mL ampicillin, 1% glucose.

6. 2ÂYTKan/Gluc agar: 2ÂYT agar, 50 μg/mL kanamycin, 1% glucose.

7. 2ÂYTAmp/Kana agar: 2ÂYT agar, 100 μg/mL ampicillin, 50 μg/mL kanamycin.

8. M9 salt: 2.5 g NaCl, 5 g NH4Cl, 33.9 g Na2HPO4,15g KH2PO4, add water to a volume of 1 L. Autoclave before usage. FFPE Tissue-Speciﬁc Antibodies 383

B: Amplification C: Depletion

A: Infection

Helperphage/ Hyperphage scFv-presenting Bacteria Normal organ phages FFPE tissue slide

G: Infection D: Selection

Malignant FFPE tissue slide F: Elution E: Washing

Fig. 1 Reproduced from ten Haaf et al. 2015 with permission from Elsevier [4]. Schematic presentation of the FFPE tissue slide selection for the generation of specific antibodies. Bacteria containing the antibody library are infected with either helper phage for monovalent antibody display or hyperphage for polyvalent display (a). After amplification of the antibody-bearing phages (b) the rescued phage particles are depleted on normal FFPE tissue (c). The nonbinding phages are applied to the malignant FFPE tissue allowing the selection of tumor-specific antibodies (d). Stringent washing is used to remove nonbinding and weak-binding phages (e), before the specific phages are eluted in an alkaline solution (f) and used to infect fresh bacteria (g). After three rounds of panning, the antibody library is enriched for tumor-specific antibodies, which can be characterized further

9. M9 agar: 200 mL M9 salt, 15 g agar, 788 mL water, autoclave and cool to 50 C. Add 2 mL 1 M MgSO4 (sterile filtered), 0.1 mL 1 M CaCl2 (sterile filtered), 10 mL 40% glucose (sterile filtered), and 0.4 mL of a 10% thiaminehydrochloride solution (sterile filtered). 10. Phosphate-Buffered Saline (PBS, 10Â): 1.37 mM NaCl, 27 mM KCl, 0.1 mM Na2HPO4, 17.6 mM KH2PO4, pH 7.4 11. PEG/NaCl solution: 20% PEG 6000, 2.5 M NaCl, store at 4 C. 12. Glycerol stock of an antibody library (e.g., human scFv antibody libraries Tomlinson I and J), store at À80 C. 384 Andre ten Haaf et al.

13. M13KO7ΔpIII hyperphage (for ﬁrst round of panning) and M13KO7 helper phage (for second and third rounds of panning) (see Note 2). 14. E. coli TG1, store at À80 C. 15. 10% RBS: 50 mL RBS, 450 mL water (see Note 3).

2.2 Antibody 1. Four slides of antigen-positive FFPE tissue and four slides of Selection antigen-negative FFPE tissue for each round of selection (see Note 4). The slides should be prepared freshly for every round of selection and should be about 5 μm thick. If using thicker tissue you receive more than one layer of cells which can hamper the antigens or can lead to an unspeciﬁc binding of the antigens. 2. Citrate buffer: Solution A: 0.1 M citric acid monohydrate, Solution B: 0.1 M trisodium citrate dihydrate, store both at 4 C. Mix 1.8 mL solution A, 8.2 mL solution B and 90 mL water. 3. EDTA-based Epitope Retrieval Solution 2 pH 9.0 (Leica Bio- systems), store at 4 C. 4. Enzyme Pretreatment Kit (Leica Biosystems), store at 4 C. 5. Tris-Buffered Saline (TBS, 20Â): 18 g Tris-Base, 137 g HCl, 175.6 g NaCl, add water to a volume of 1 L. 6. Xylol. 7. 100% ethanol, 96% ethanol and 70% ethanol. 8. Delimiting pen for immunocytochemistry. 9. Wash buffer: 1Â PBS, 0.1% Tween 20. 10. Blocking buffer MPBS: 4% and 2% skin dry milk powder in 1Â PBS. 11. Phage elution: 0.1 M triethylamine, pH 12.2. 12. 1 M Tris–HCl: 15.76 g Tris–HCl, add water to a volume of 100 mL.

13. 2ÂTYGlycerol/Gluc/Amp medium: 2ÂYT medium, 15% glycerol, 1% glucose, 100 μg/mL ampicillin. 14. Overhead rotator. 15. Microtome. 16. Glass slides.

3 Methods

3.1 Phage Rescue The antibody-bearing phage particles are the starting material for the antibody selection. For that reason the production is a very important step and crucial for a successful selection. The antibody FFPE Tissue-Speciﬁc Antibodies 385 libraries used in this protocol are from the Medical Research Coun- cil Laboratory in Cambridge, UK and are in scFv antibody format, but different antibody libraries can also be utilized. Please note that the used phagemid carries an ampicillin resistance and the hyper- and helper phage a kanamycin resistance. If your phagemid carries a different resistance you should adapt the protocol accordingly. 1. Inoculate one 500 μL glycerol aliquot of the antibody library in 250 mL 2ÂYTAmp/Gluc medium and incubate overnight at 37 C and 250 rpm. 2. Spread 50 μL fresh E. coli TG1 on a M9 agar plate and incubate overnight at 37 C(see Note 5). 3. Pipet 200–250 μL of the overnight culture to 10 mL 2ÂYTAmp/Gluc medium and incubate at 37 C and 250 rpm. The start OD600nm should be between 0.08 and 0.1. Incubate until an OD600nm of 0.4–0.5 is reached (duration about 1.5–2 h) (see Note 6). 4. Take 5 mL of the bacteria culture and infect with 1010 phage particles (hyperphage for the ﬁrst round and helper phage for the second and third rounds). Incubate the infected bacteria for 30 min at 37 C without shaking and afterward 30 min at 37 C with shaking at 250 rpm. 5. As a control of the infection spread 50 μL of the bacteria on a 2ÂYTAmp/Kana agar plate and incubate overnight at 37 C. This plate should be full of bacteria clones on the next day (see Note 7). 6. Centrifuge the bacteria for 5 min at 4500 Â g and 4 C and discard the supernatant.

7. Resuspend the pellet in 25 mL 2ÂYTAmp/Kana/IPTG medium and incubate the bacteria for 18–20 h at 30 C and 250 rpm (see Note 8). 8. Centrifuge the bacteria for 30 min at 4500 Â g and 4 C. 9. Mix the phage containing supernatant with 5 mL ice-cold PEG/NaCl-solution and incubate for 60 min on ice. Mix the solution occasionally during the incubation. 10. Centrifuge the solution for 30 min at 15,000 Â g and 4 C and carefully discard the supernatant. Dry the pellet by placing it overhead on a paper tissue for 2 min. 11. Resuspend the phage pellet in 220 μL PBS and transfer the phages in a 1.5 mL tube. 12. Centrifuge the phages for 2 min at 13,000 Â g and 4 C and transfer the phage containing supernatant into a fresh 1.5 mL tube. Place the phages on ice or for long time storage at À20 CorÀ80 C. 386 Andre ten Haaf et al.

13. Inoculate TG1’s from the minimal agar plate in 10 mL 2ÂYT medium and incubate at 37 C and 250 rpm. The start OD600nm should be between 0.08 and 0.1. Incubate until an OD600nm of 0.4–0.5 is reached (duration about 1.5–2 h). Bac- teria can be placed on ice until use. 14. As a control of unwanted contamination spread 50 μL of the TG1’s on a 2ÂYTAmp/Gluc agar plate and a 2ÂYTKan/Gluc agar plate and incubate overnight at 37 C. These plates should be empty on the next day (see Note 9). 15. Prepare a serial dilution of the rescued phages for the determination of the input-titer. Therefore, mix the phages carefully by pipetting up and down and transfer 2 μL into a 1.5 mL tube À ﬁlled with 198 μL PBS (10 2 dilution). Mix this tube again carefully and transfer 2 μL in the next 1.5 mL tube ﬁlled with À À 198 μL PBS (1 Â 10 4 dilution). Repeat this step for a 10 6 À and 10 8 dilution.

16. Add 800 μL of the TG1’s with an OD600nm of 0.4–0.5 to each tube and incubate for 30 min at 37 C without shaking and afterward 30 min at 37 C with shaking at 250 rpm.

17. Transfer 10 μL of each dilution on a 2ÂYTAmp/Gluc agar plate and incubate the plates overnight at 37 C. 18. Count the bacteria on the agar plates and calculate the input- titer as followed (see Note 10):

colonie number Â dilution factor Titer½¼ cfu=mL dilution Â spread volume

3.2 Antibody 1. Block the 220 μL phages from Subheading 3.1, step 12 with Selection 220 μL 4% MPBS and incubate for 2 h on an overhead rotator. 2. To remove the paraffin place the FFPE-tissue slides (4 for depletion and 4 for selection) in xylol and incubate for 10 min. 3. Afterward dip the slides 5–10 times in 100% ethanol, then 5–10 times in fresh 100% ethanol. 4. Dip the slides 5–10 times in 96% ethanol, then 5–10 times in fresh 96% ethanol. 5. Dip the slides 5–10 times in 70% ethanol. 6. Dip the slides 5–10 times in deionized water and place them in Tris-buffer. The tissues should never get dry between any steps. Drying could lead to shrinkage of the tissue and thereby to a conformational change of the antigens. Subsequently, the accessibility of the antibodies can be hampered or an unspecific binding can occur. 7. Start with the epitope retrieval of the slides for depletion (see Note 11). FFPE Tissue-Specific Antibodies 387

(a) Keep one slide that is not treated with epitope retrieval technique in Tris-buffer. (b) For proteolytic epitope retrieval cover one tissue slide with enzyme from the enzyme pretreatment kit and incubate for 10 min at 37 C. Afterward wash the tissue slide 3-times in tris-buffer and place the slide in Tris-buffer until further use. (c) For heat-induced epitope retrieval place one slide in citrate-buffer and one slide in EDTA-buffer, respectively, and bring to the boil in a microwave at 800 W. Afterward heat the slides in the buffers for 7 min at 160 W. Repeat this step once. If buffer evaporates or boils over add new buffer to fully cover the tissue. Cool down the tissue slides for 20 min, wash them 3-times in Tris-buffer and place the slides in Tris-buffer until further use. 8. Revolve tissue with the delimiting pen to prevent liquids to clear off from the slides. 9. Block tissue slides with 2% MPBS for 1 h. Take care to fully cover the tissue. 10. Meanwhile, execute the epitope retrieval of the four slides for selection as described above. 11. Remove the blocking solution from the tissue and add 100 μL of the blocked phages on each slide. Incubate for 1 h. The 100 μL blocked phages should cover the complete tissue. If you use large tissues, you may need to resuspend the rescued phage particles from Subheading 3.1, step 11 in a bigger amount of PBS and block them with the same amount of 4% MPBS at Subheading 3.2, step 1. 12. Simultaneously, block the tissue slides for selection with 2% MPBS for 1 h. 13. Transfer the phages from the negative tissue slides to the tissue slides for selection and incubate for 1 h. 14. Inoculate TG1’s from the minimal agar plate in 25 mL 2ÂYT medium and incubate at 37 C and 250 rpm. The start OD600nm should be between 0.08 and 0.1. Incubate until an OD600nm of 0.4–0.5 is reached (duration about 1.5–2 h). Bac- teria can be placed on ice until use. 15. As a control of unwanted contamination spread 50 μL of the TG1’s on a 2ÂYTAmp/Gluc agar plate and a 2ÂYTKan/Gluc agar plate incubate overnight at 37 C. These plates should be empty on the next day. 16. Meanwhile, wash the tissue slides as followed (see Note 12): 388 Andre ten Haaf et al.

First round Second round Third round

3Â PBS 1Â PBST 10Â PBST 2Â PBS, shaking for 4Â PBST, shaking for 5Â PBST, shaking for 5 min 5 min 5 min 3Â PBS 1Â PBST 10Â PBST 2Â PBS, shaking for 4Â PBST, shaking for 5Â PBST, shaking for 5 min 5 min 5 min

17. For the elution of the phages add 100 μL 0.1 M triethylamine on each slide and incubate for 10 min. 18. For neutralization of the phages pipet 400 μL Tris–HCl in a 1.5 mL tube and transfer the eluted phages from the slides in the tube (4 Â 100 μL). 19. To determine the output-titer transfer 2 μL of the neutralized phages into a 1.5 mL tube ﬁlled with 198 μL PBS and prepare a serial dilution as described above for the input-titer. Place residual phages on ice.

20. Add 800 μL of the TG1’s with an OD600nm of 0.4–0.5 to each tube and incubate for 30 min at 37 C without shaking and afterward 30 min at 37 C with shaking at 250 rpm.

21. Transfer 10 μL of each dilution on a 2ÂYTAmp/Gluc agar plate and incubate the plates overnight at 37 C. 22. Add 100 μL TG1’s on each tissue slide and incubate for 30 min at 37 C(see Note 13). 23. Simultaneously, pipet the residual phages to 15 mL TG1’s and incubate for 30 min at 37 C. 24. Transfer the TG1’s from the tissue slides (4 Â 100 μL) to the other TG1’s and centrifuge for 5 min at 4500 Â g. 25. Discard the supernatant and resuspend the pellet in 200 μL 2ÂYT medium.

26. Spread the bacteria on a 15 cm 2ÂYTAmp/Gluc agar plate and incubate overnight at 37 C. 27. Count the colonies on the output-titer plates and calculate the titer as described above for the input-titer (see Note 14).

28. Add 1.5 mL 2ÂTYGlycerol/Gluc/Amp medium to the 15 cm agar plate and scratch the bacteria from the plate. 29. Transfer the bacteria into a kryo tube and store at À80 C. For the second and third rounds of selection scratch a small amount of the glycerol stock from the previous round and incubate the bacteria in 5 mL 2ÂYTAmp/Gluc medium overnight at 37 C and 250 rpm. On the next day follow the protocol as described above. FFPE Tissue-Speciﬁc Antibodies 389

After three rounds of panning the antibody library should be enriched of specific antibodies and can be analyzed for unique monoclonal antibodies, e.g., by ELISA or flow cytometry. You should always use an unrelated antigen as a control for specificity of the selected antibodies in your analysis.

4 Notes

1. Storage of the medium at room temperature helps to identify contaminations. Always check medium for any contamination. If the medium looks turbid use a fresh bottle of autoclaved medium. 2. Hyperphage is used in the first round of panning for a multivalent display of the antibodies on phage surface to select all antibodies against the given antigen. Helper phage is used in the second and third rounds of panning, thereby having a monovalent display of the antibodies. Thus, the binding strength is dependent on the affinity of the antibody to its target. There are no avidity effects affecting the binding strength in the second and third rounds which helps to select for highly affine antibodies. 3. Phages are very robust and are not inactivated by autoclaving for 20 min at 121 C. To prevent contamination with phages 10% RBS can be used to inactivate phage particles. Thus, the phage working place and pipettes should be cleaned with 10% RBS after every phage working step. Besides RBS, phage particles can be inactivated by UV exposure for 60 min. 4. Tissues contain a lot of different antigens and against most of them no antibodies are wanted. Therefore, depletion of the unwanted antibodies on antigen negative tissue is necessary. The depletion tissue should be as close as possible to the antigen positive tissue (e.g., tumor tissue for selection and corresponding healthy tissue for depletion). Ideally, the only differences are the target antigen. 5. One agar plate of TG1’s is sufficient for all three rounds of panning. The TG1’s should be sealed with parafilm or adhesive tape and stored at 4 C but separated from the phages to circumvent an unwanted infection. If there is any contamination prepare a new plate of TG1’s. 6. The phages infect the bacteria over their F-pilus. Therefore, it is important that the bacteria are in the exponential growth phase (log phase) where they express the F-pilus. If the OD600nm is too high the bacteria reach a stationary phase where an infection of the phages is not possible any more. 390 Andre ten Haaf et al.

7. The control agar plate contains ampicillin and kanamycin. Thus, only bacteria that carry the phagemid (ampicillin resistance) and are infected with the phages (kanamycin resistance) could grow on the plate. 8. The gpIII-antibody fusion gene on the phagemid is controlled by a lac promoter. The IPTG in the medium enables the expression of the gpIII-antibody gene and thus the production of antibody presenting phages. The incubation at 30 C ensures a correct folding of the antibodies on the phage surface.

9. If there are clones on the 2ÂYTAmp/Gluc agar plate the bacteria might be contaminated with the phagemid from the antibody library, whereby the bacteria receive the ampicillin resistance. If there are clone on the 2ÂYTKan/Gluc agar the bacteria are infected with hyper- or helper phage, whereby the bacteria receive the kanamycin resistance. In both cases, all steps starting from Subheading 3.1, step 13 need to be repeated with fresh bacteria. 10. Example for calculation of the input-titer: 200 μL phages þ 800 μL TG1’s (¼ dilution factor 5), 10 μL spread À on the agar plate, next day 120 colonies on the 1 Â 10 8 dilution plate. Â 120 5 ¼ Â 12 = À 6 10 cfu mL 1 Â 10 8 Â 0:01 mL Normally, the input-titer is about 1 Â 1012 cfu/mL when using hyperphage and 1 Â 1013 cfu/mL when using helper phage.

11. Formalin-fixation of the tissue can lead to cross-linked and partial denatured proteins, which can prevent antibodies from binding to its target. By the use of epitope retrieval techniques the immunoreactivity of the proteins can be restored. Since the antigen for the selection is unknown the optimal epitope retrieval technique cannot be predicted. Therefore, three different epitope retrieval techniques are used for depletion and selection to ensure that the antibodies are able to bind the antigens. In addition, one slide is not treated further, because not all antigens need to be retrieved. 12. Wash stringency is increased every round to select for highly affine antibodies. 13. High affine antibodies might not be eluted from the tissue slides by the triethylamine in Subheading 3.2, step 17. To also recover phages bearing these antibodies, fresh TG1’s are added to the tissue slides and placed for 30 min at 37 C. Hereby, the phages should infect the bacteria and can be transferred to the other eluted phages. FFPE Tissue-Specific Antibodies 391

14. A typical output-titer for the ﬁrst round of selection is about 1 Â 106–1 Â 107 cfu/mL, for the second round 1 Â 107–1 Â 108 cfu/mL and for the third round 1 Â 107–1 Â 109 cfu/mL. The output-titer depends on the tissue, the antigen accessibility, the used library and the input- titer. By dividing the output-titer over the input-titer you can determine the enrichment factor for each round. Ideally, the ratio increases after each round, which is an indication for an enrichment of speciﬁc antibodies. However, in the event you are not seeing a calculated enrichment, you should proceed with standard analysis of monoclonal binders.

Acknowledgment

This work was supported by the “Verein zur Fo¨rderung der Krebs- forschung in Giessen e. V.”

References

1. Fleuren ED, Versleijen-Jonkers YM, Heskamp S, of vimentin-specific antibodies from the HuCAL van Herpen CM, Oyen WJ, van der Graaf WT, phage display library by subtractive panning on Boerman OC (2014) Theranostic applications formalin-fixed, paraffin-embedded tissue. Biol of antibodies in oncology. Mol Oncol 8 Chem 388(6):651–658. https://doi.org/10. (4):799–812. https://doi.org/10.1016/j. 1515/BC.2007.070 molonc.2014.03.010 6. Sun Y, Shukla GS, Kennedy GG, Warshaw DM, 2. Schofield DJ, Lewis AR, Austin MJ (2014) Weaver DL, Pero SC, Floyd L, Krag DN (2009) Genetic methods of antibody generation and Biopanning phage-display libraries on small tis- their use in immunohistochemistry. Methods sue sections captured by laser capture microdis- 70(1):20–27. https://doi.org/10.1016/j. section. J Biotech Res 1:55–63 ymeth.2014.02.031 7. Tanca A, Pagnozzi D, Addis MF (2012) Setting 3. Sanchez-Martin D, Sorensen MD, Lykkemark S, proteins free: progresses and achievements in Sanz L, Kristensen P, Ruoslahti E, Alvarez- proteomics of formalin-fixed, paraffin-embed- Vallina L (2015) Selection strategies for antican- ded tissues. Proteomics Clin Appl 6(1–2):7–21. cer antibody discovery: searching off the beaten https://doi.org/10.1002/prca.201100044 path. Trends Biotechnol 33(5):292–301. 8. Thavarajah R, Mudimbaimannar VK, https://doi.org/10.1016/j.tibtech.2015.02. Elizabeth J, Rao UK, Ranganathan K (2012) 008 Chemical and physical basics of routine formal- 4. Ten Haaf A, Pscherer S, Fries K, Barth S, dehyde fixation. J Oral Maxillofac Pathol 16 Gattenlohner S, Tur MK (2015) Phage display- (3):400–405. https://doi.org/10.4103/0973- based on-slide selection of tumor-specific anti- 029X.102496 bodies on formalin-fixed paraffin-embedded 9. Tanaka T, Ito T, Furuta M, Eguchi C, Toda H, human tissue biopsies. Immunol Lett 166 Wakabayashi-Takai E, Kaneko K (2002) In situ (2):65–78. https://doi.org/10.1016/j.imlet. phage screening. A method for identification of 2015.05.013 subnanogram tissue components in situ. J Biol 5. Jarutat T, Nickels C, Frisch C, Stellmacher F, Chem 277(33):30382–30387. https://doi. Hofig KP, Knappik A, Merz H (2007) Selection org/10.1074/jbc.M203547200 Chapter 22

Antibody Afﬁnity and Stability Maturation by Error-Prone PCR

Tobias Unkauf, Michael Hust, and Andre´ Frenzel

Abstract

Antibodies are the fastest growing class of pharmaceutical proteins and essential tools for research and diagnostics. Often antibodies do show a desirable specificity profile but lack sufficient affinity for the desired application. Here, we describe a method to increase the affinity of recombinant antibody fragments based on the construction of mutagenized phage display libraries. After the construction of a mutated antibody gene library by error-prone PCR, selection of high-affinity variants is either performed by panning in solution or on immobilized antigen with washing conditions optimized for off-rate-dependent selection. An additional screening protocol to identify antibodies with improved thermal stability is described.

Key words Afﬁnity maturation, Stability maturation, Error-prone PCR, Antibody phage display, Single-chain fragment variable (scFv)

1 Introduction

In the past three decades monoclonal antibodies have been the fastest growing class of pharmaceutical proteins with a global sales revenue of nearly $75 billion in 2013, representing approximately half of the total sales of all biopharmaceutical products. The combined worldwide sales are predicted to be nearly $125 billion in 2020 [1]. Antibodies are furthermore critical reagents in many fundamental biochemical methods such as affinity chromatography, enzyme linked immunosorbent assays (ELISA), immunohistochemistry, western blotting, or flow cytometry. The rapid expansion of genomics, proteomics, and other biotechnology fields has led to a growing demand for antibodies as high-affinity reagents to specifically recognize, e.g., peptides and proteins but also carbohydrates and haptens [2]. Although the more recent technologies yeast and ribosome display are becoming highly established, phage display is currently

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_22, © Springer Science+Business Media LLC 2018 393 394 Tobias Unkauf et al.

the most robust, reliable, and well-characterized method. It allows the screening of large antibody libraries with theoretical complex- ities of up to 1011 different clones [3], hence almost entirely covering the structural diversity of naturally occurring antibodies (2 Â 1012)[4]. When compared to animal-based generation of antibodies many advantages are offered by in-vitro technologies such as phage display, ribosome display, or yeast display: enhanced throughput by parallelization and miniaturization, the stringent control of selection conditions, and the possibility of using toxic proteins [5]. However, even complex naive antibody libraries sometimes show a proportion of just 1–10 binders to a given target per 107 clones [6]. Sometimes, this results in the selection of low affinity clones with KD values above 100 nM. Although exhibiting low affinities, these antibodies still can provide useful starting molecules for the construction of mutagenized libraries to screen from [7–9]. For the creation of diverse libraries from single-parental sequences many methods have been developed. Nearly all approaches generate diversity by introducing mutations at the nucleotide level. Non-stochastic techniques often use alanine-scanning or site- directed mutagenesis to generate limited collections of specific variants but require prior knowledge of the respective antibody— antigen interaction [10]. Stochastic methods like error-prone PCR [11], mutator bacterial strains [12], and site-specific saturation mutagenesis of complementary determining regions (CDR) [13, 14] are random mutagenesis methods. Mutator bacterial strains mutate the antibody gene but also the vector backbone, making a subsequent recloning of the antibody gene necessary. Directed mutagenesis of the CDR is only limited to the immediate antigen-binding sites of the antibody; hence, a limitation of the affinity maturation potential of an antibody might occur. CDRs are directly involved in target recognition, but framework regions are the foundations of the VH and VL structures and thus are of high importance for the CDRs’ presentation. Mutations within the framework regions of a given antibody can therefore stabilize and improve the antibody-antigen interactions [15]. The most common technique to introduce random mutations is error-prone PCR as it bypasses both limitations by targeting the whole antibody gene which than can be directly used for the construction of a mutagenized library. Selectivity for high-affinity binders during panning can be relatively low even when binders do differ by a factor of 10 in affinity [6]. Successful selection of high-affinity mutants is only achieved by phage-display panning approaches [16] using many harsh and long washing steps. The use of multiple forms of the target antigen in Antibody Affinity and Stability Maturation by Error-Prone PCR 395

sequential selection rounds and the inclusion of competitor proteins can drive the selected pool toward a highly specific set of epitopes. A disadvantage of classical panning on immobilized antigen is that this kind of selection often enriches binders with increased tendency for dimerization, especially with phage libraries using the scFv format. This problem can be bypassed by panning the antigen in solution [17] or by adding unbiotinylated antigen or soluble antibody fragments after antibody phage binding has occurred [18]. Monoclonal binders can be tested and affinity ranked by using crude cell supernatants from 96-well production in a competitive ELISA approach [19] or solution equilibrium titration (SET) using highly sensitive electrochemiluminescence [20]. Specificity can be validated, for example, by flow cytometry or peptide arrays such as PEPperCHIP [21]. Affinities are usually determined by surface plasmon resonance (SPR), microscale ther- mophoresis (MST), or biolayer interferometry (BLI). A rapid and easy screening procedure to rank lead candidates for thermal stability from crude E. coli production supernatant is described within these protocols.

2 Materials

2.1 Error-Prone PCR 1. Template DNA (e.g., phagemid-containing antibody fragment gene). 2. Site-speciﬁc DNA-oligo primer sets. 3. PCR Thermocycler. 4. GeneMorphII Random Mutagenesis Kit (Stratagene, Amster- dam, Netherlands). 5. PCR clean-up kit. 6. Agarose gel.

2.2 Library 1. Phage-display-compatible phagemid (e.g., pHAL14, Construction pCOMB3Â)[22, 23]. 2. Restriction enzymes NcoI, NotI. 3. Shrimp alkaline phosphatase, SAP. 4. PCR clean-up kit. 5. T4 DNA-Ligase. 6. Electrocompetent E. coli ER2738 cells (Lucigen Corporation, Middleton, USA); Genotype: [F’proA + B+ lacIq Δ(lacZ)M15 zzf::Tn10 (tetr)] fhuA2 glnVΔ(lac-proAB) thi-1Δ(hsdS-mcrB)5. 396 Tobias Unkauf et al.

7. SOC medium: 0.5% (w/v) yeast extract, 2.0% (w/v) tryptone, 10 mM NaCl, 2.5 mM KCl, 10 mM MgCl2, 20 mM MgSO4, 20 mM glucose. 8. Tabletop thermomixer.

9. 2 M Mg solution: 1 M MgCl +1 M MgSO4. 10. SOB medium pH 7.0: 2% (w/v) tryptone, 0.5% (w/v) yeast extract, 0.05% (w/v) NaCl; after autoclaving add sterile 1% (v/v) 2 M Mg solution. SOB-GA: SOB, containing 100 μg/ mL ampicillin, 100 mM glucose, 1.5% (w/v) agar. 11. 25 cm sterile plastic dishes. 12. 2YT medium pH 7.0: 1.6% (w/v) tryptone, 1% (w/v) yeast extract, 0.5% (w/v) NaCl. 13. 2YT-GA medium: 2YT, containing 100 μg/mL ampicillin, 100 mM glucose. 14. 80% (v/v) glycerol solution. 15. 1.8 mL Cryotubes.

2.3 Library 1. 9 cm sterile petri dishes. Validation 2. Phagemid-speciﬁc DNA-oligo primer set. 3. GoTaq DNA Polymerase. 4. PCR clean-up kit.

2.4 Library 1. 2YT-GA: 2YT, containing 100 μg/mL ampicillin, 100 mM Packaging glucose. 2. 100 mL shake ﬂask. 3. Helperphage M13K07 (Stratagene). 4. 2YT-AK: 2YT, containing 100 μg/mL ampicillin, 50 μg/mL kanamycin. 5. Polyethyleneglycol (PEG) solution: 20% (w/v) PEG 6000, 2.5 M NaCl in water. 6. Phage dilution buffer: 10 mM Tris–HCl pH 7.5, 20 mM NaCl, 2 mM EDTA.

2.5 Titration 1. 2YT-T: 2YT, containing 50 μg/mL tetracycline. 2. 2YT-GA: 2YT, containing 100 μg/mL ampicillin, 100 mM glucose, 1.5% (w/v) agar agar. 3. E. coli XL1 Blue MRF´ (Stratagene); Genotype: Δ(mcrA)183 Δ(mcrCB-hsdSMR-mrr)173 endA1 supE44 thi-1 recA1 gyrA96 relA1 lac [F´ proAB lacIqZΔM15 Tn10 (Tetr)].

2.6 Selection by 1. PBS pH 7.4: 8.0 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4*2H2O, Panning 0.24 g KH2PO4 in 1 l water solution. Antibody Afﬁnity and Stability Maturation by Error-Prone PCR 397

2. PBS-T: PBS + 0.1% Tween 20. 3. MPBS-T: 2% skim milk in PBST, prepare fresh. 4. Panning block solution: 2% (w/v) skim milk +2% (w/v) BSA in PBST; prepare fresh. 5. 10 μg/mL trypsin in PBS. 6. 2YT-T: 2YT, containing 50 μg/mL tetracycline. 7. 15 cm petri dishes. 8. 80% (v/v) glycerol solution.

2.7 Selection 1. Soluble antigen. in Solution 2. Biotinylated soluble antigen. 3. Streptavidin beads (Dynabeads M-280, Invitrogen, Karlsruhe, Germany). 4. Overhead shaker. 5. Magnet particle concentrator (Dynal). 6. 10 μg/mL trypsin in PBS.

2.8 Production 1. 96-well U-bottom polypropylene (PP) microtiter plates (Grei- of Soluble, Monoclonal ner BioOne, Frickenhausen, Germany). Antibody Fragments 2. AeroSeal breathable sealing ﬁlm (Excel Scientiﬁc, USA). 3. 2YT-A containing 50 μM isopropyl-beta-D-thiogalactopyranoside (IPTG). 4. Microtiter plate thermo shaker.

2.9 Monoclonal 1. Maxisorp Plates (Nunc). ELISA and Thermal 2. α-tag antibody (e.g., for pHAL vectors: Mouse α-myc-tag Stability Screening monoclonal antibody (9E10, (Sigma-Aldrich, Munich, Ger- many) or Mouse α-Penta His-tag monoclonal antibody (Qia- gen, Hilden, Germany)). 3. Goat α-Mouse IgG serum (Fab-speciﬁc), HRP conjugated (Sigma-Aldrich, Munich, Germany). 4. TMB solution A, pH 4.1: 10 g citric acid solved in 100 mL water, add 9.73 g potassium citrate, add H2O to make 1 L. 5. TMB solution B: 240 mg tetramethylbenzidine, 10 mL acetone, 90 mL ethanol, 907 mL 30% H2O2.

6. 0.5 M H2SO4. 7. Microtiter plate thermo shaker. 398 Tobias Unkauf et al.

3 Methods

3.1 Error-Prone PCR 1. Design specific primers for your phagemid flanking your gene of interest approximately 50–100 bp up and downstream respectively. To increase DNA quality for cloning and number of error-prone PCR rounds, a nested approach with a second pair of primers is feasible. Consider that the restriction sites that are used for library cloning should be included into these primers. 2. Perform PCR by using GeneMorph II Random Mutagenesis Kit (Stratagene) according to the manufacturer’s instructions. 3. Recommended: Use 10–15 ng purified template DNA and 30 cycles for error-prone PCR. The indicated amount of DNA refers to the amplified sequence only, not the total vector. For example, 100 ng À150 ng DNA of a 1000 bp plasmid will be used if the amplified sequence is 100 bp (see Note 1). 4. Validate successful amplification on agarose gel, cut out the respective band. 5. Clean up DNA. 6. 10–15 ng of this DNA can be used in an additional error-prone PCR. PCR and clean-up steps can be repeated 3–4 times until no distinct amplification band can be obtained anymore.

3.2 Library 1. Digest the PCR product and phage-display vector with suitable Construction restriction enzymes (e.g., NcoI and NotI can be used for pHAL vectors pHAL14 and pHAL30 [23]). 2. For increased ligation efﬁciency, any suitable dephosphorilation enzyme can be added to the vector digest. 3. Incubate digest according to the manufacturer’s instructions. 4. Optional: Heat inactivate all enzymes according to the manufacturer’s instructions. 5. Clean up digested DNA and determine the respective DNA concentrations of vector and PCR product (see Note 2). 6. Use approximately 1 μg vector DNA for ligation. Adjust the amount of PCR product accordingly. A molar ratio vector: insert of 1:3 is recommended. Perform ligation overnight at 16 C. 7. Optional: Heat inactivation of ligation according to ligase manufacturer’s instructions.

8. Clean up ligation by washing four times with dH2O using ® Amicon Ultra-0.5 centrifugal ﬁlter devices (Merck KGaA). Final volume should be 50 μL. Antibody Afﬁnity and Stability Maturation by Error-Prone PCR 399

9. Mix 25 μL prechilled ligation aliquot with 25 μL electrocompetent ER2738 E. coli cells (Lucigen) and perform electroporation at 1.7 kV, 4–5 ms pulse (BioRad, electroporation unit). The remaining ligation can be used for a second transformation that gets pooled with the ﬁrst one in Subheading 3.2, step 12. 10. Resuspend transformed cells immediately in 950 μL pre-warmed SOC medium, incubate for 1 h at 37 C and 650 rpm (Eppendorf thermomixer). 11. Take 10 μL of the cell suspension and make a dilution series in 2ÂYT or LB medium. For successful transformations 1 Â 108–5 Â 108 clones can be expected. Plate an aliquot of each dilution and calculate the transformation efﬁciency after overnight incubation at 30 C. The average of all dilution plates with countable colonies is the maximum theoretical diversity (size) of the library. 12. Plate the rest of the cell suspension to 25 Â 25 cm 2ÂYT GA agar plates. Incubate overnight at 30 C. 13. Add 30 mL 2ÂYT medium to each plate and incubate on a rocker for 15 min. Carefully scrape cells from the plate using a spatula. 14. Prepare glycerol stocks of your library by mixing 250 μL glycerol and 750 μL cell suspension in 1.8 mL cryotubes. Shock freeze aliquots in liquid nitrogen, store at À80 C.

3.3 Library 1. Pick up to 30 clones from the plates used for the determination Validation of transformation efﬁciency. 2. Perform colony-PCR (50 μL scale) using a primer set which ampliﬁes the gene of interest. 3. Analyze a 5 μL aliquot by agarose gel electrophoresis to check upon rate of successful insert integration. The percentage of these “positive” clones indicates the quality of the library. 4. Clean up the remaining 45 μL PCR product and use it for sequencing with appropriate primers. 5. Determine the average mutation rate by aligning the obtained sequences with the parental template sequence. Note: The percentage of silent mutations can be greatly reduced by subsequent rounds of error-prone PCR.

3.4 Library 1. Gently thaw a library glycerol stock on ice. Packaging 2. Inoculate 50 mL 2ÂYT-GA medium with cell suspension directly from the glycerol stock to an initial OD600 of about 0.1. 3. Grow the cells at 37 C and 250 rpm to an OD600 of about 0.5. 400 Tobias Unkauf et al.

4. Transfer 20 mL culture into a sterile 50 mL polypropylene tube and add 5 Â 1012 cfu M13K07. Mix gently. 5. Incubate at 37 C for 30 min without and 30 min with shaking at 250 rpm. 6. Centrifuge the suspension at 10 min at 3200 Â g to pellet the cells. Discard the supernatant to remove remaining glucose and resuspend the pellet in 50 mL 2ÂYT-AK in a 100 mL shake flask. Due to the selection with kanamycin, only M13K07 (Kanr) infected cells will survive and produce antibody-phage. 7. Incubate the cells overnight at 30 C and 250 rpm. 8. Centrifuge the culture at 10 min at 3200 Â g to pellet the cells. 9. Precipitate the phage from supernatant by adding 1/5 volume ice-cold PEG/NaCL solution. Incubate for 2 h on ice with gentle shaking. 10. Pellet the phage by centrifugation for 1 h at 12,000 Â g and 4 C. 11. Put the tubes upside down on tissue paper to remove the PEG solution completely. 12. Resuspend the phage pellet in 10 mL phage dilution buffer and filter the solution through a 0.45 μm filter (Whatman syringe filter). 13. Precipitate phage a second time by adding 1/5 volume ice-cold PEG/NaCl and incubation on ice for 2 h. 14. Pellet phage particles by centrifugation at 4 C and 20,000 Â g for 30 min. 15. Completely remove remaining PEG/NaCL and resuspend pellet in 500 μL phage dilution buffer. 16. Optional: Remaining cell debris might be removed by an additional centrifugation step (2 min, 4 C, 16,000 Â g). 17. Titer phage solution. Phage can be stored at 4 C.

3.5 Titration 1. Prepare an 5 mL overnight culture of E. coli XL1-Blue MRF’ in 2ÂYT-T medium by shaking at 250 rpm and 37 C. 2. Inoculate 50 mL fresh 2ÂYT-T medium with the overnight culture to an OD600 of about 0.05. Grow culture at 250 rpm and 37 CuptoOD600 ~ 0.5 (see Note 3). 3. Make a serial dilution of the phage solution in PBS. The number of eluted phage during panning depends on several parameters such as antigen, library, panning round, washing stringency, etc. The phage titer can vary from 103 to 107 cfu. The phage preparation after re-ampliﬁcation of the eluted phage has a titer of about 1012–1013 cfu/mL. Antibody Afﬁnity and Stability Maturation by Error-Prone PCR 401

4. Infect 50 μL bacteria with 10 μL of each phage dilution and incubate for 30 min at 37 C without shaking. 5. Plate the 60 μL infected bacteria on 2ÂYT-GA agar plates and incubate plates overnight at 37 C. 6. Count the colonies and calculate the colony-forming units (cfu) titer according to the respective dilution.

3.6 Selection by 1. Coat antigen overnight in different amounts of 1–50 ng/well Panning in a Maxisorp stripe (Nunc) using PBS Buffer (4 C). 2. Block microtiter plate stripe with 350 μL MPBS-T for 1 h at room temperature. 3. Discard blocking solution, wash three times with PBS-T (see Note 4). 4. Incubate x cfu of antibody phage from the library in 150 μL MPBS-T and add this solution to the wells. The appropriate amount of phage depends on the total size of the library as determined in Subheading 3.2, step 11. The amount of phage should exceed library size by 100Â. If you library size is, e.g., 108, use 1010 cfu phage particles. 5. Incubate for 1 h at room temperature. 6. Wash three times with PBS-T (see Note 4). 7. Put the stripe in 2 L sterile PBS, incubate under soft stirring at 4 C for 1 week. 8. It is recommended to repeat steps 6 + 7 one time weekly (3–4 weeks total). 9. Elute with 200 μL trypsin solution for 30 min at 37 C(see Note 5). 10. Use 10 μL of phage solution for titration as described in Subheading 3.5. 11. Inoculate 50 mL 2YT-T with an overnight culture of E. coli XL1-Blue MRF’ in 100 mL Erlenmeyer ﬂask and grow at 250 rpm and 37 C until the culture reaches OD600 ~ 0.5. 12. Infect the culture with the remaining 190 μL trypsin-phage solution and incubate for 30 min without and 30 min with shaking at 37 C. 13. Harvest the infected bacteria by centrifugation for 10 min at 3200 Â g in 50 mL polypropylene tubes. Resolve the pellet in 250 μL2ÂYT-GA and plate the bacterial on 2ÂYT agar plates (15 cm petri dish). Grow overnight at 37 C. 14. Pick single colonies from these plates to produce soluble monoclonal antibody-fragments (Subheading 3.8). 402 Tobias Unkauf et al.

3.7 Selection 1. For all selection processes “in solution” the antigen must be in Solution available unlabeled and biotinylated. 2. The selection process can be either performed by systematically reducing the antigen concentration (A) or by competition with free antigen or soluble antibody-fragments (B). 3. Block 50–100 μL streptavidin beads M-280 (3.25–6.5 Â 108 beads) with 1.5 mL panning block for 1 h at room temperature. Use magnet particle concentrator (Dynal) for bead separation. 4. Use 50 μL blocked streptavidin beads for library preincubation for 30 min at room temperature to remove unwanted streptavidin binders from the library. Discard beads afterward. 5. Use the preincubated library for the selection process. 6. (A) Mix 1 Â 1011 phage from your preincubated library with 50 nM biotinylated antigen in panning block solution. Incu- bate for 1 h at room temperature, gently mixing using an overhead shaker (see Note 6). 7. (A) Add 100 μL blocked streptavidin beads. Perform capturing by incubating 15 min at room temperature and gently mixing in an overhead shaker. 8. (A) Separate the streptavidin beads carrying the captured antibody phage from the solution in the magnetic rack. 9. (A) Discard the supernatant and wash the streptavidin beads with fresh washing buffer for 1–2 min using the overhead shaker. Collect streptavidin beads in the magnetic particle concentrator. Following this procedure wash three times with PBS-T, two times with MPBS-T, two times with PBS, one time with MPBS, and ﬁnally two times with PBS. 10. (A) Elute captured phage by adding 500 μL trypsin, incubate for 30 min at 37 C. Separate and discard the streptavidin beads, use the supernatant for titration and the production of new phage (start at Subheading 3.4, step 4 by adding the whole eluted phage to the cells (see Note 5)). 11. (A) Start the next round by repeating the complete procedure with the newly produced phage particles. Use reduced amount of biotinylated antigen (e.g., 5 nM). The amount of streptavidin beads can be also reduced to 50 μL (3.25 Â 108 beads). 12. (A) Perform up to four selection round, while constantly reducing the antigen amount in subsequent rounds (see Note 7). 13. (A) Pick colonies from the plates that were used for titration to produce soluble antibody fragments. 14. (B) Perform library and bead blocking as described in steps 3–5. Antibody Afﬁnity and Stability Maturation by Error-Prone PCR 403

15. (B) Mix 1 Â 1012 phage particles from the library with 1 nM biotinylated antigen in 1.5 mL panning block. 16. (B) Incubate for 1 h at room temperature, gently mixing in an overhead shaker. 17. (B) Add up to 1 mM (max. factor 1000) of the unbiotinylated antigen or the parental antibody. 18. (B) Incubate 1 week at 4 C, gently mixing in an overhead shaker to equilibrate the system. 19. (B) Wash the beads three times with panning block solution and four times with PBS-T (increase washing steps if necessary). 20. (B) Elute phage as described in step 10. 21. (B) Titer eluted phage and pick monoclonals for the production of soluble monoclonal antibody fragments.

3.8 Production 1. Fill each well of a 96-well U-bottom polypropylene plate with of Soluble Monoclonal 150 μL2ÂYT-GA medium. Antibody Fragments 2. Pick 92 clones with sterile tips from the desired panning round and inoculate each well. Use the wells H3, H6, H9, and H12 for controls. H3 and H6 are negative controls—these wells will not be inoculated. Inoculate the wells H9 and H12 with the clone containing the phagemid encoding for the “parental” antibody fragment. Seal the plate with breathable sealing ﬁlm. 3. Incubate overnight in a microtiter plate shaker at 37 C and 1200 rpm. 4. (A) Fill a new 96-well polypropylene microtiter plate with 150 μL2ÂYT-GA/well and add 10 μL of the overnight cultures to each well. Incubate at 37 C and 1200 rpm for 2 h. (B) Add 30 mL glycerol solution to the remaining 140 mL overnight cultures. Mix by pipetting and store this master plate at À80 C. 5. Pellet the bacteria in the microtiter plates by centrifugation at 3200 Â g and 4 C for 10 min. Carefully remove the glucose- containing medium above the pellets without disturbing the pellet. 6. Add 180 μL2ÂYT-A supplemented with 50 μM IPTG, seal the plate with a breathable sealing ﬁlm. Incubate overnight at 30 C and 1200 rpm (see Note 8). 7. Pellet the bacteria by centrifugation at 3200 Â g for 10 min. Transfer the supernatant containing the soluble antibody fragments carefully to a new 96-well plate without disturbing the bacteria pellet. Supernatant can be stored at 4 C for a short time or can be used directly in ELISA (see Note 9). 404 Tobias Unkauf et al.

3.9 ELISA of Soluble 1. To analyze the antigen speciﬁcity of the soluble monoclonal Monoclonal Antibody antibody fragments, coat 100 ng antigen per well in PBS for Fragments 1 h at room temperature or overnight at 4 C in a maxisorb microtiter plate (Nunc). 2. Wash the coated microtiter plate wells three times with PBST. The washing should be performed with an ELISA washer (e.g., TECAN Columbus Plus) for reproducible washing results. If no ELISA washer is available, wash manually three times with PBST. 3. Block the antigen-coated wells with MPBS for 1 h at room temperature. 4. Wash three times with PBS-T. 5. Fill each well with 10 μL antibody solution and 90 μL MPBS-T. Incubate for 1 h at room temperature. 6. Wash three times with PBS-T. 7. Incubate wells with 100 μL α-tag antibody solution for 1 h at room temperature. Dilute the antibody according to the manufacturer’s instruction in MPBS-T. 8. Wash three times with PBS-T. 9. Incubate with 100 μL of appropriate HRP conjugate (e.g., goat α-mouse HRP conjugate if α-tag antibody is of murine origin). Dilute according to the manufacturer’s instruction in MPBS-T. 10. Wash three times with PBS-T. 11. Shortly before use, mix 10 parts TMB solution A with 0.5 parts TMB solution B. Add 100 μL of the prepared TMB solution into each well and incubate for 1–15 min until a bright blue color is developed. 12. Stop the reaction by adding 100 μL 0.5 M sulfuric acid to each well. The color turns from blue to yellow. 13. Measure the extinction at 450 nm in an ELISA Reader to identify positive candidates. 14. Sequence candidates to eliminate identical clones.

3.10 Stability 1. To screen soluble monoclonal antibody fragments for increased Screening of Soluble thermal stability, coat 100 ng antigen per well in PBS for 1 h at Monoclonal Antibody room temperature or overnight at 4 C in a maxisorb microtiter Fragments plate (Nunc). 2. In a new polypropylene 96-well plate mix 10 μL of each antibody solution with 90 μL PBS. Incubate the plate at elevated temperature for 1 h. The exact temperature depends on the stability of the parental antibody. If no prior knowledge is available prepare multiple plates and incubate at 48 C, 50 C, 52 and 54 C respectively. Antibody Afﬁnity and Stability Maturation by Error-Prone PCR 405

3. Wash the coated microtiter plate wells three times with PBST. The washing should be performed with an ELISA washer (e.g., TECAN Columbus Plus) for reproducible washing results. If no ELISA washer is available, wash manually three times with PBST. 4. Block the antigen-coated wells with MPBS for 1 h at room temperature. 5. Wash three times with PBS-T. 6. Add the heat-treated antibody solutions to the antigen-coated wells and incubate for 1 h at room temperature. 7. Follow the standard screening procedure as described in Sub- heading 3.9 starting with step 6.

4 Notes

1. A total of 2–3 rounds of error-prone PCR can be performed by using the same primer set. After more rounds, the PCR product yield and quality will decrease rapidly. Switch to the next, inner primer set if more rounds of error-prone PCR are required. 2. Vectors containing only small inserts <100 bp between the cloning sites do not require purification by agarose gel electrophoresis. Instead, the vector can be digested, dephosphorylated, and directly purified using a PCR purification kit. Avoiding agarose gel electrophoresis can tremendously improve following cloning steps.

3. If the bacteria have reached OD600¼0.5 before they are needed, the culture can be kept on ice for max. 60 min to maintain the F pili on the E. coli cells. 4. The washing should be performed with an ELISA washer for reproducible washing results. If no washer is available wash manually three times with PBS-T. 5. Phagemids like pSEX81 [24], pHAL14 [25], or pHAL30 [23] have coding sequences for a trypsin-speciﬁc cleavage site between the antibody fragment gene and the gIII. Trypsin also cleaves within antibody fragments whereas the phage is very resistant to this procedure. As a result, also vectors without a trypsin cleavage site might be used successfully. 6. The initial antigen concentration depends on the equilibrium dissociation constant (KD) of your parent clone and the KD you try to achieve by these methods. Variation in the concentration should lead to an optimal result. 7. The ﬁnal antigen concentration should be a maximum of ten times less than the desired KD for improved binders. E.g., use 406 Tobias Unkauf et al.

0.5 nM antigen, when binders with KD around 5 nM are desired. 8. The appropriate IPTG concentration for induction depends on the vector design. A concentration of 50 μM is well suited for vectors with a Lac promoter like pSEX81 [24], pIT2 [26], pHENIX [27], and pHAL14/pHAL30 [23]. 9. The method for the production of soluble antibodies works with vectors with (e.g., pHAL30) and without (e.g., pSEX81) an amber stop codon between antibody fragment and gIII. If the vector has no amber stop codon, the antibody::pIII fusion protein will be produced and may be used instead successfully [28].

References

1. Ecker DM, Jones SD, Levine HL (2015) The 9. Douthwaite JA et al (2015) Affinity maturation therapeutic monoclonal antibody market. of a novel antagonistic human monoclonal mAbs 7(1):9–14 antibody with a long VH CDR3 targeting the 2. Finlay WJ, Bloom L, Cunningham O (2011) Class a GPCR formyl-peptide receptor 1. mAbs Phage display: a powerful technology for the 7(1):152–166 generation of high specificity affinity reagents 10. Rajpal A et al (2005) A general method for from alternative immune sources. Methods greatly improving the affinity of antibodies by Mol Biol 681:87–101 using combinatorial libraries. Proc Natl Acad 3. Tiller T et al (2013) A fully synthetic human Sci U S A 102(24):8466–8471 Fab antibody library based on fixed VH/VL 11. Liu JL et al (2012) Attainment of 15-fold framework pairings with favorable biophysical higher affinity of a fusarium-specific single- properties. mAbs 5(3):445–470 chain antibody by directed molecular evolution 4. Pantazes RJ, Maranas CD (2013) MAPs: a coupled to phage display. Mol Biotechnol 52 database of modular antibody parts for predict- (2):111–122 ing tertiary structures and designing affinity 12. Low NM, Holliger P, Winter G (1996) Mim- matured antibodies. BMC Bioinformatics icking somatic hypermutation: affinity matura- 14:168 tion of antibodies displayed on bacteriophage 5. Tomszak F et al (2016) Selection of recombi- using a bacterial mutator strain. J Mol Biol 260 nant human antibodies, in protein targeting (3):359–368 compounds. In: Prediction, Selection and 13. Chowdhury PS (2002) Targeting random Activity of Specific Inhibitors. Springer Inter- mutations to hotspots in antibody variable national Publishing, New York, pp 23–54 domains for affinity improvement. Methods 6. McCafferty J (1996) Phage display: factors Mol Biol 178:269–285 affecting panning efficiency, in phage display 14. Laffly E et al (2008) Improvement of an anti- of peptides and proteins. Academic Press, Bur- body neutralizing the anthrax toxin by simulta- lington, pp 261–276 neous mutagenesis of its six Hypervariable 7. Lamdan H et al (2013) Affinity maturation and loops. J Mol Biol 378(5):1094–1103 fine functional mapping of an antibody frag- 15. Renaut L et al (2012) Affinity maturation of ment against a novel neutralizing epitope on antibodies: optimized methods to generate human vascular endothelial growth factor. Mol high-quality ScFv libraries and isolate IgG can- BioSyst 9(8):2097–2106 didates by high-throughput screening. Meth- 8. Li B et al (2014) In vitro affinity maturation of ods Mol Biol 907:451–461 a natural human antibody overcomes a barrier 16. Hust M et al (2014) Selection of recombinant to in vivo affinity maturation. mAbs 6 antibodies from antibody gene libraries. Meth- (2):437–445 ods Mol Biol 1101:305–320 Antibody Affinity and Stability Maturation by Error-Prone PCR 407

17. Schier R et al (1996) Isolation of high-affinity Manual. Cold Spring Harbor Laboratory monomeric human anti-c-erbB-2 single chain Press, Cold Spring Harbor, New York, p 736 Fv using affinity-driven selection. J Mol Biol 23. Kugler€ J et al (2015) Generation and analysis of 255(1):28–43 the improved human HAL9/10 antibody 18. Thie H et al (2011) Rise and fall of an anti- phage display libraries. BMC Biotechnol 15:10 MUC1 specific antibody. PLoS One 6(1): 24. Welschof M et al (1997) The antigen-binding e15921 domain of a human IgG-anti-F(ab’)2 autoanti- 19. Friguet B et al (1985) Measurements of the body. Proc Natl Acad Sci U S A 94 true affinity constant in solution of antigen- (5):1902–1907 antibody complexes by enzyme-linked immu- 25. Hust M et al (2007) Handbook of therapeutic nosorbent assay. J Immunol Methods 77 antibodies. Wiley-VCH Verlag GmbH & (2):305–319 Co. KGaA, Weinheim 20. Della Ducata D et al (2015) Solution equilib- 26. Goletz S et al (2002) Selection of large diver- rium titration for high-throughput affinity esti- sities of antiidiotypic antibody fragments by mation of unpurified antibodies and antibody phage display. J Mol Biol 315(5):1087–1097 fragments. J Biomol Screen 20 27. Finnern R et al (1997) Human autoimmune (10):1256–1267 anti-proteinase 3 scFv from a phage display 21. Vernet T et al (2015) Spot peptide arrays and library. Clin Exp Immunol 107(2):269–281 SPR measurements: throughput and quantifi- 28. Mersmann M et al (1998) Monitoring of scFv cation in antibody selectivity studies. J Mol selected by phage display using detection of Recognit 28(10):635–644 scFv- pIII fusion proteins in a microtiter scale 22. Barbas CF III, Burton DR, Scott JK, Silverman assay. J Immunol Methods 220(1–2):51–58 GJ (2001) Phage Display: A Laboratory Part III

Complementary Approaches for Antibody Phage Display Selections Chapter 23

Upgrading Afﬁnity Screening Experiments by Analysis of Next-Generation Sequencing Data

Christian Grohmann and Michael Blank

Abstract

Computational analysis of next-generation sequencing data (NGS; also termed deep sequencing) enables the analysis of affinity screening procedures (or biopanning experiments) in an unprecedented depth and therewith improves the identification of relevant peptide or antibody ligands with desired binding or functional properties. Virtually any selection methodology employing the direct physical linkage of geno- and phenotype to select for desired properties can be leveraged by computational analysis. This article describes a concept how relevant ligands can be identified by harnessing NGS data. Thereby, the focus lays on improved ligand identification and describes how NGS data can be structured for single-round analysis as well as for comparative analysis of multiple selection rounds. Especially, the comparative analysis opens new avenues in the field of ligand identification. The concept of computational analysis is described at the example of the software tool “AptaAnalyzerTM.” This intuitive tool was developed for scientists without special computer skills and makes the computational approach accessible to a broad user range.

Key words Next-generation sequencing, NGS, Bioinformatics, In silico analysis, In vitro selection, Biopanning

1 Introduction

NGS was originally developed for the purpose of whole genome sequencing and re-sequencing. However, the large amount of sequencing data delivered by these platforms is also ideally suited for extensive analysis of complex collections of diversified gene segments such as antibody or peptide libraries. Scientists from other areas of basic, applied, and medical research realized the potential of the ground-breaking NGS technology and started its transfer to their own research fields [1]. The transfer into the field of display technologies already began in 2008 when NGS became first time affordable [2]. Virtually all affinity selection or “biopanning” procedures (like phage display, ribosome display, mRNA, display, yeast display) can be leveraged by NGS and subsequent in silico analysis [3–5]. The products (target-addressing ligands)

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_23, © Springer Science+Business Media LLC 2018 411 412 Christian Grohmann and Michael Blank

range from linear to (multi)cyclic peptides [6, 7] and more complex protein scaffolds like antibodies (Ab) [8], scFvs [9], or protein domains [10]. The present chapter describes biopanning experiments using the example of cysteine-knot miniproteins also called microbodies [11–13]. As microbodies bear several variable peptide loops that are spanned into a defined constant scaffold, they are well suited to explain the procedure model. Therefore, the displayed gene encoding the protein is PCR amplified and therewith made accessible to NGS. In doing so, naive libraries as well as enriched libraries of the affinity selection experiments can be analyzed in an unprecedented depth. Nowadays, NGS gains more and more popularity also in the field of affinity screening. One of the main reasons is that the traditional approach on the basis of multiple successive rounds of affinity selection and Sanger sequencing gives access only to a very limited number of clones (representing only <0.01% of the complete sequence space of the library) [2, 14]. Picking and cloning after repeated rounds typically enables no representative random sampling of the enriched libraries. Not least because of the small sample space is further decreased by dominant false positive hits that emerge for two important reasons: binding to non-target- related materials used during selection (e.g., plastics, albumin) [15] and due to propagation advantages [16, 17] of good replicat- ing sequences. In this context, the final outcome is the isolation of clones with no actual specificity for the original target. The complementation of the traditional screening procedure by NGS and computational data analysis aids to take a much more extensive picture of the entire sequence space, and thereby enables identifying the relevant clones with higher probability. It provides the possibility of characterizing millions of sequencing reads in a single NGS run and therewith the chance to identify enriched clones in very early rounds of biopanning (round one or two) [6]. The reduction of the number of selection rounds and therewith the reduction of the amplification bias thereby decreases the chance of isolating nonspecific false positive clones and accelerates the process of target-specific ligand identification. A systematic method how NGS can be used to upgrade display methodologies by NGS and subsequent data analysis is described in this article. It is dedicated to scientists who are not familiar with computational analyses of screening experiments and therefore focuses on the intuitive AptaAnalyzerTM software that enables them to extract most relevant information from screening experiments. Computational Analysis of Biopanning Experiments 413

2 Analysis of Deep Sequencing Data Derived from Biopanning Experiments

Computational analysis typically starts with the digitalization of deﬁned stages (selection rounds) of the biopanning experiment. Thereby, millions of sequences can be made accessible via NGS (Fig. 1a) [18]. Subsequently, the raw NGS datasets (each representing one round) are submitted to computational analysis (Fig. 1b). Data are typically structured into projects and subprojects and thereby ﬁled by selection rounds. In the next step datasets of respective rounds can be chosen for single or multiple dataset analysis.

2.1 Parsing The process that structures and organizes the millions of sequences and Archiving of Raw derived from naive as well as from enriched panning rounds is NGS Data termed parsing process. Here, typically “raw” sequences from FASTQ files how they are typically generated by various NGS procedures [19] are checked for their usability to be further analyzed. Sequences quality information (Phred quality scores) can be taken as a criterion to decide if a sequence is accepted or rejected for further analysis. Also for paired-end sequencing this quality information is used. Sequencing in paired end-read mode is necessary if the ligands’ variable region exceeds the possible maximum length of a read in single-read mode. Thereby, sequencing is performed from both the ends of a sequence. Sequences from both the ends are subsequently merged in the region of the common overlapping region. Thereby, nucleotides with the better quality scores are finally used. The AptaAnalyzerTM software files exploitable sequences in a way that individual libraries (naive as well as enriched libraries) are saved as individual data sets. Raw sequence data derived from

Fig. 1 Workﬂow of biopanning in combination with NGS and computational analysis. (a) NGS-amplicon generation from standard biopanning experiments and digitalization of individual selection rounds via NGS results into the datasets r1-r3. (b) Archiving of sequences in a database by preprocessing and parsing of NGS datasets as well as generation of results (tables and graphics) 414 Christian Grohmann and Michael Blank

Fig. 2 (a) Example of a database. Experiments are structured into projects and subprojects. Each experiment can contain several rounds whereby one round corresponds to one dataset. Datasets can be analyzed individually but also be compared against each other. (b) The “Sequence Parts Editor” shows the deﬁnition of the combinatorial library molecules: constant regions (1, 3, 5 and 7) and variable regions (2, 4 and 6). Here exemplarily shown for (c) a Microbody®-scaffold containing 3 variable loops

biopanning experiments are parsed round by round into a database, whereby individual rounds can be filed into projects and subprojects (Fig. 2a). A parsing process typically starts with the definition of combinatorial libraries molecules: position and sequence of constant motifs can be described and therewith the number, position, and length of variable sequence parts (Fig. 2b). This database finally is the source for subsequent computational analysis. Datasets (thus selection rounds) can be individually (single-dataset analysis, see Subheading 2.2) as well as comparably (multiple dataset analysis, see Subheading 2.3) analyzed.

2.2 Single-Dataset Typically, some statistic data are necessary to assess the significance Analysis of a dataset (number of total, accepted, rejected sequences, as well as Phred quality score). The percentage of unique sequences can serve as a first indicator if and to what extent enrichment took place (the less the number of unique sequences the more enriched sequences are present in the respective selection round). Further results can be given on the level of full sequences (complete clones) bearing several constant and variable parts, as well as on the level of the respective shorter variable regions. ® Figure 3 shows a tabular list of complete clones of Microbody sequences subdivided into constant and variable regions (also see Fig. 2c) including relative and absolute counts. Fig. 3 AptaAnalyzerTM table view, showing results on the level of clones (full sequences). The list of enriched sequences can further be expanded by user-defined columns that can contain clone-IDs, experimental data like affinity values (KD) or further information like specificity (e.g., if the ligand is a plastic binder). Table filters enable us to filter the data in a way that, e.g., matrix binders are filtered out and only clones with adjusted affinity values and/or specificity criteria are given 416 Christian Grohmann and Michael Blank

Sequences of enriched clones can be identified and therewith relevant variable regions (here loop 1–3). In the next step the analysis of the individual, variable loops of interest can be executed. However, for such a detailed analysis the big amount of data derived from NGS has to be systematically de-convoluted in order to enable the necessary highly performant computation of multiple datasets each comprising millions of sequences. The identification of clones of interest is achieved by clustering its variable sequence regions (here loop 1–3) in accordance with similarity criteria into sequence families [20]. Subsequent computational analysis can then be performed by working with so-called leader sequences, whereby each leader represents a defined family. In case a defined family’s leader sequence shows the expected behavior the analysis type of interest can be “replayed” on the level of all member sequences of the family. AptaAnalyzerTM uses the concept of data de-convolution for table views as well as for graphs. The different table views in Fig. 4a–c give information on different levels of detail, whereby family’s leader sequences (Fig. 4a) and its family members (Fig. 4b, c) are interlinked. Primary graphical analysis of combinatorial libraries for example gives the distribution of full sequence or sequence regions frequencies, visualizes the connectivity of different variable regions, or gives distribution of lengths of variable regions in this dataset. Naive libraries can be quality checked for equal distribution of clones to ensure that the starting repertoire of combinatorial molecules is diverse enough and not limited by repression of molecules with a higher copy count. For libraries of advanced rounds, the enriched clones can easily be identified by visualizing the sequence frequencies in a diagram. Each bar in Fig. 5a represents an entire sequence family, and thus visualizes the sum of all clones that have been clustered into this family. Depending on the amount of enrichment and the applied sequencing depth each family may be built up by tens (early rounds) to several hundreds (later rounds) of closely related sequences. Independent of the applied clustering algorithm and independent of the size of each family, one family is represented by one leader sequence. In the first instance this clone can be tested in binding- or functional assays. The de-convolution approach gives access to relevant clones, even if they are underrepresented in the dataset (may be due to poor amplification behavior): underrepresented but relevant clones are likely grouped into one of the top ranked families. As soon as wet lab experiments confirmed a top ranked family to be relevant by testing of the leader sequence, also the quite rare member sequences, that build up this family can be examined and tested more closely. Without de-convolution via clustering such clones would be invisible and thus inaccessible. Fig. 4 Table views showing enriched variable regions of “Loop 2” of Microbodies. Table view (a) gives representative leader sequences of all families identified in selected dataset (Round_3_beads). (b) The family members of the selected leader sequence (row 6, YRSRIG) selected in table view “a” are listed in the context of other constant and variable sequence regions. (c) Due to the degeneracy of codons the same clone can be specified by different nucleic acids. All different versions of nucleic acids of the clone selected in table view “b” (clone of row 1) are listed 418 Christian Grohmann and Michael Blank a Ranking of Families b Ranking of Members of Family 8 80 “YRSRIG”

6 60

4 40 Frequency [%] 2 Frequency [%] 20

0 0 GQTHRGSRHDPGYHT STDEYPSHW SYLGGE IPYRKR YRSRIG QTLKKSEQY GSRKQQ QSNYTR QSNLNKNQDTKA YRSRIG YRSRTG HRSRIG YPSRIG YPSCIG YRTRIG YCSRIG YRSRIR YSSRIG YRSHIG QRSRIG YRSYIG YRKRIG

c Position Analysis of Family “YRSRIG” 1

0.8

0.6 YRSRIG

0.4 Frequency [%] 0.2

T 0 12 3 45 6 Position in Loop 2

Fig. 5 Analysis of “loop 2 region” (a) Frequency distribution of the top 10 ranked loop 2 families. Each bar represents one family of sequences. (b) It shows the frequency distribution of all family members of family “YRSRIG” (marked by an arrow in “a”). The loop 2 sequence “YRSRIG” makes up 78% of the family and about 1.7% of the entire dataset. The second most frequent family member “YRSRTG” (marked by an asterisk) makes up 6.4% of the family but only 0.05% of the entire dataset. In the conventional approach clones with loop “YRSRTG” are unlikely cloned and picked by accident. The analysis of individual positions of family YRSRIG is given in (c). The stacked bar chart diagram conﬁrms that the visualized family is dominated by the clone YRSRIG. E.g., position ﬁve of the family is dominated by “I” (Ileu), however to about 1% of position “5” can be occupied with a “T” (Thr)

Clustering can be performed by applying different methods: the simplest is counting and ranking of the most frequent identical sequences followed by fuzzy searching for other sequences that differ by a deﬁned number of amino acid building blocks. The COMPAS tool applies other criteria like the so-called co-occurrence approach [20]. In this approach regions are rated to be similar, if they bear two motifs that are at the same time highly overrepresented in the enriched dataset. As ﬁnally motifs mediate Computational Analysis of Biopanning Experiments 419

binding to the target molecule this method most likely clusters sequences addressing the same sites of the target molecule. Another method clusters sequence families by the criteria of Shannon’s information entropy [20]. The affiliation of sequences to families is successively tested while monitoring their influence on the family’s relative information entropy—a family is defined via the first minimum. Other alignment tools enable user-defined blosum matrices to cluster and align sequences [21]. However, such CPU intensive alignment algorithms are not always compatible with big datasets, and thus have to be applied in a second clustering step. This second clustering step (or de-convolution step two) has therefore to be performed with the family’s leader sequences to build up so-called super families or family clans [20]. Another powerful de-convolution step two criterion however is the tracing of the relative frequencies of leader sequences over different selection cycles. Thereby, different enriched libraries are compared on the level of individual clones. Thus, analysis is performed via comparison of multiple rounds of the same or of different biopanning experiments.

2.3 Multiple Dataset One variant for data de-convolution is the tracing of the frequency Analysis of a representative sequence (one for each sequence family) over other biopanning rounds. This strategy opens up new avenues in the field of ligand identification as it enables comparing biopanning rounds at very high resolution—on the level of monoclonal sequences. This approach is of special value if biopanning experiments are designed in a way that additional control experiments are performed in parallel to affinity screening against the actual target of interest. Such an experimental design allows monitoring if (or to what extent) defined monoclonal sequences are present or absent in the control experiments. Control experiments can be (1) pannings against closely related targets, or (2) pannings performed in the presence of a specific competitor or (3) pannings against cell lines expressing the target on the surface instead of being, e.g., a recombinant target protein. Depending on the type and nature of the control experiments the scientist can hypothesize if antibody- or peptide ligands of interest can be expected to be enriched or decreased in respective control experiments. Blank et al. applied this strategy in an affinity screening experiment termed SELEX (Systematic Evolution of Ligands by Exponential enrichment) [18]. The goal was the identification of ligands addressing the target protein gp120 in a way that gp120—CD4 interaction is inhibited. Screening experiments have been performed against the recombinant target protein gp120wt (wild type). But in parallel other control experiments have been performed against the variant gp120mt (mutant) that was mutated at the binding site of interest (at its interaction site to the CD4 receptor). The hypothesis is that ligands of interest get exclusively enriched against the gp120wt 420 Christian Grohmann and Michael Blank

protein, whereas irrelevant ligands get enriched against gp120wt as well as gp120mt. To identify gp120-CD4 interaction inhibiting clones the relative frequency of clones enriched in the “wild type experiment” was traced over NGS datasets derived from the “mutant experiment” that was only performed as a control for in silico comparison. A comparison revealed a very poorly but exclusively enriched clone (making up only 0.5% of the finally enriched pool). This clone showed the hypothesized discriminating enrichment. This rare clone could not be identified by the conventional screening approach of cloning and picking. The identification of sequences from enriched panning rounds via NGS (millions of sequences) in contrast to the identification via Sanger sequencing (hundreds to thousands) provides the possibility of comparing screening rounds on a monoclonal level at very high resolution, and thus allows the comparison of datasets derived from affinity screening experiments performed against positive and negative targets. However, sole counting of identical sequences and subsequent comparison of their relative frequencies in multiple datasets would be quite computing intensive as enriched libraries typically still contain tens of thousands of different clones. To finally identify the needle in the haystack, computational analysis must provide both de-convolution and filter mechanisms that enable the extraction of sequences and sequence families with the desired differential abundance in defined datasets. In analogy to single-dataset analysis, in the first instance the comparison is again performed with leader sequences that represent all family members. In the second step this type of comparative analysis can be performed with the family members of those sequences that show expected discriminatory enrichment (¼ discriminatory binding). The concept is illustrated in Fig. 6: Overview results are generated at the example of leader sequences. Bars (Fig. 6a), dots (Fig. 6b), or segments (Fig. 6c) represent the sum of family members of sequences. Mouse Click into the region of interest finally gives detailed results on the level of all individual monoclonal family members (Fig. 6 in boxes).

3 Outlook

NGS and computational analysis support biopanning experiments at different stages: (1) the quality of naive libraries can be checked to identify batches with poor quality to exclude them from laborious screening experiments. (2) The design of an experiment in combination with the suitable in silico strategy enables the dissection of enriched pools at very high resolution. Rare (poor replica- tors) but relevant clones as well as frequent but irrelevant clones (separation matrix binders or ampliﬁcation parasites) can be identi- ﬁed and selected or rejected for further analysis. Clones can be Computational Analysis of Biopanning Experiments 421 abTracing of Comparison of “Leader “Leader Sequences” Sequences” 24 max 8 YRSRIG aa. GQTHRGSRH 22 aa ab ab. DPGYHT ac. YRSRIG 20 6 ad. QSNLNKNQDTKA 2-2

-4 4 2 min Data density aa ab Frequency Round 3 2-6 Frequency ac 2 Tracing of Members of Family 2-8 Comparison of Members 1 0 aa “YRSRIG” 2 of Family “YRSRIG”

-10 -2 ab ac 0,8 2 2 aa aa. YRSRIG -4 ab. YRSRTG -12 2 0 0,6 ac. HRSRIG 2 1234 1234 1234 aa ad. YPSRIG 2-14 2-12 2-10 2-8 2-6 2-4 2-6 0,4 ae. YRSCIG 1: Round 1 Frequency Round 2 2-8

0,2 -10 2: Round 2 Frequency Round 3 2 Relative Frequency

aa ab 3: Round 3 0 ac ad ae 2-12 4: Round 3v 1234 1234 1234 1234 1234 2-14 2-13 2-12 2-11 2-10 2-9 2-8 2-7 2-6 2-5 2-4 2-3 1: Round1; 2: Round2 Frequency Round 2 3: Round3; 4: Round3v

c Distribution of “Leader Sequences” A

B C Distribution of Members of Family “YRSRIG”

A: Round 2 B: Round 3 B C C: Round 3v

A: Round2; B: Round3; C: Round3v

Fig. 6 Comparative dataset analysis of “Loop 2” regions. (a) Tracing the frequency of families at the example of family’s respective leader sequences. 1: round 1; 2: round 2, 3: round 3; 4: round 3v (variant of round 3). E. g., family “aa” (GQTHRGSRH) could be identified in round 3 as well as in round 3v. Mouse click on a defined barchart series gives detailed results on the level of its family members. Results for barchart series “ac” (YRSRIG) show frequency distribution of respective family members. (b) Comparison of two datasets: Family’s frequencies are visualized at the example of family’s respective leader sequences. Comparison of selection round 2 and 3 enables graphical identification of sequences, e.g., with high amplification fold value. E.g., mouse click on the dot of family “YRSRIG” gives results on the level of its family members. (c) Distribution of families in datasets A: round 2; B: round 3 and; C: round 3v (variant of round 3) at the example of family’s leader sequences. Click on respective segments, e.g., A “AND” C lists leader sequences that have been selected in round 3 as well as in its variant (round 3v). Selection of a defined family visualizes the distribution of all its family members 422 Christian Grohmann and Michael Blank

identified in very early selection rounds and ligands addressing defined target epitopes can be already identified in silico. (3) Clus- tering and comparison of related sequences enables identifying conserved thus relevant sequence positions—valuable information for lead optimization. Screening campaigns can finally be accelerated and the chance to identify ligands with desired binding properties can be increased. Moreover, computational analysis offers dynamic insight into selection processes. It can be expected that NGS and in silico analysis will contribute more and more to a profound understanding of the influence of applied selection pressures (like addition of competi- tors, amount of washing, or type of target immobilization). There- with the usage of NGS and in silico analysis enables further streamlining of in vitro evolution processes on different platforms. Nowadays, the digitalization of screening rounds (NGS amplicon generation as well as NGS) is standardized [22]. There are algorithms and programs that align sequences, perform epitope- mapping, or identify motifs like MatLab’s bioinformatics toolbox [22], PHASTpep [23], MIMOP [24], SiteLight [25], PEPTIDE [26], MEME [27], DNAStar [28], SliMFinder [29], or MAFFT [30]. Anyway, the actual bottleneck is the computational analysis. To efficiently contribute to selection procedures, the computational analysis has to meet a variety of requirements. Of importance is a convenient usability of the software: for the parsing process as well as for the actual analysis of screening rounds. The software should also allow the archiving of selection experiments (sequence information but also findings out of the laboratory like binding affinity, specificity, etc.) to enable a continuous understanding and optimization of biopanning procedures. The used algorithms have also to be compatible with big data. Often tens of millions sequences have to be computed in a reasonable time on standard computers. The used algorithms should also function without error-prone query for parameters. Results have to be in the form of comprehensible, interactive graphs and tables and the usage of filter settings should allow a systematic reduction of big data, down to a number of sequences for subsequent experimental testing. For a widespread use a kind of standard software that considers these requirements is necessary. The AptaAnalyzer software tool was designed for scientists without special computer skills and provides all essential functionalities to archive and analyze biopanning experiments. AptaAnalyzerTM will successively contribute with every new analyzed experiment to bring more light into the black box of affinity screening experiments. Computational Analysis of Biopanning Experiments 423

References

1. Kahvejian A, Quackenbush J, Thompson JF 11. Getz JA, Rice JJ, Daugherty PS (2011) (2008) What would you do if you could Protease-resistant peptide ligands from a knot- sequence everything? Nat Biotechnol tin scaffold library. ACS Chem Biol 6:837–844 26:1125–1133 12. Zoller F, Markert A, Barthe P, Hebling U, 2. Mannocci L, Zhang Y, Scheuermann J, Altmann A, Lindner T, Mier W, Haberkorn U Leimbacher M, De Bellis G, Rizzi E et al (2013) A disulfide-constrained miniprotein (2008) High-throughput sequencing allows with striking tumor-binding specificity devel- the identification of binding molecules isolated oped by ribosome display. Angew Chem Int from DNA-encoded chemical libraries. Proc Ed Engl 52:11760–11764 Natl Acad Sci U S A 105:17670–17675 13. Zoller F, Haberkorn U, Mier W (2011) Mini- 3. Dower WJ, Mattheakis LC (2002) In vitro proteins as phage display-scaffolds for clinical selection as a powerful tool for the applied applications. Molecules 16:2467–2485 evolution of proteins and peptides. Curr Opin 14. Liu GW, Livesay BR, Kacherovsky NA, Chem Biol 6:390–398 Cieslewicz M, Lutz E, Waalkes A, Jensen MC, 4. Van Blarcom T, Rossi A, Foletti D, Sundar P, Salipante SJ, Pun SH (2015) Efficient identifi- Pitts S, Bee C, Melton Witt J, Melton Z, Hasa- cation of murine M2 macrophage peptide tar- Moreno A, Shaughnessy L, Telman D, Zhao L, geting ligands by phage display and next- Cheung WL, Berka J, Zhai W, Strop P, generation sequencing. Bioconjug Chem Chaparro-Riggers J, Shelton DL, Pons J, Raj- 26:1811–1817 pal A (2015) Precise and efficient antibody 15. Menendez A, Scott JK (2005) The nature of epitope determination through library design, target-unrelated peptides recovered in the yeast display and next-generation sequencing. J screening of phage-displayed random peptide Mol Biol 427:1513–1153 libraries with antibodies. Anal Biochem 5. Olson CA, Nie J, Diep J, Al-Shyoukh I, Taka- 336:145–157 hashi TT, Al-Mawsawi LQ, Bolin JM, Elwell 16. Matochko WL, Cory Li S, Tang SK, Derda R AL, Swanson S, Stewart R, Thomson JA, Soh (2014) Prospective identification of parasitic HT, Roberts RW, Sun R (2012) Single-round, sequences in phage display screens. Nucleic multiplexed antibody mimetic design through Acids Res 42:1784–1798 mRNA display. Angew Chem Int Ed Engl 17. Vodnik M, Zager U, Strukelj B, Lunder M 51:12449–12453 (2011) Phage display: selecting straws instead 6. t’Hoen PA, Jirka SM, Ten Broeke BR, Schultes of a needle from a haystack. Molecules EA, Aguilera B, Pang KH et al (2012) Phage 16:790–817 display screening without repetitious selection 18. Blind M, Blank M (2015) Aptamer selection rounds. Anal Biochem 421:622–631 technology and recent advances. Mol Ther 7. Dias-Neto E, Nunes DN, Giordano RJ, Sun J, Nucleic Acids 4:e223 Botz GH, Yang K et al (2009) Next-generation 19. Pareek CS, Smoczynski R, Trtyn A (2011) phage display: integrating and comparing avail- Sequencing technologies and genome able molecular tools to enable cost-effective sequencing. J Appl Genet 52:413–435 high-throughput analysis. PLoS One 4:e8338 20. Blank M (2016) Next-generation analysis of 8. Ravn U, Gueneau F, Baerlocher L, Osteras M, deep sequencing data: bringing light into the Desmurs M, Malinge P et al (2010) By-passing black box of SELEX experiments. Methods in vitro screening--next generation sequencing Mol Biol 1380:85–95 technologies applied to antibody display and in silico candidate selection. Nucleic Acids Res 21. Lal D, Verma M (2017) Large-scale sequence 38:e193 comparison. Methods Mol Biol 1525:191–224 9. Xuelian B, Jihye K, Seungmin K, Wankyu K, 22. Matochko WL, Derda R (2015) Next- Hyunbo S (2015) A novel human scFv library generation sequencing of phage-displayed pep- with non-combinatorial synthetic CDR diver- tide libraries. Methods Mol Biol sity. PLoS One 10:e0141045 1248:249–266 10. Ernst A, Gfeller D, Kan Z, Seshagiri S, Kim 23. Brinton LT, Bauknight DK, Dasa SSK, Kelly PM, Bader GD, Sidhu SS (2010) Coevolution KA (2016) PHASTpep: analysis software for of PDZ domain-ligand interactions analyzed discovery of cell-selective peptides via phage by high-throughput phage display and deep display and next-generation sequencing. PLoS sequencing. Mol BioSyst 6:1782–1790 One 11:e0155244 424 Christian Grohmann and Michael Blank

24. Moreau V, Granier C, Villard S, Laune D, 28. Burland TG (2000) DNASTAR’s Lasergene Molina F (2006) Discontinuous epitope pre- sequence analysis software. Methods Mol Biol diction based on mimotope analysis. Bioinfor- 132:71–91 matics 22:1088–1095 29. Edwards RJ, Davey NE, Shields DC (2007) 25. Halperin I, Wolfson H, Nussinov R (2003) SLiMFinder: a probabilistic method for identi- SiteLight: binding-site prediction using phage fying over-represented, convergently evolved, display libraries. Protein Sci 12:1344–1359 short linear motifs in proteins. PLoS One 2: 26. Pizzi E, Cortese R, Tramontano A (1995) e967 Mapping epitopes on protein surfaces. Biopo- 30. Katoh K, Standley DM (2013) MAFFT multi- lymers 36:675–680 ple sequence alignment software version 7: 27. Bailey TL, Johnson J, Grant CE, Noble WS improvements in performance and usability. (2015) The MEME suite. Nucleic Acids Res Mol Biol Evol 30:772–780 43:39–49 Chapter 24

Next-Generation DNA Sequencing of VH/VL Repertoires: A Primer and Guide to Applications in Single-Domain Antibody Discovery

Kevin A. Henry

Abstract

Immunogenetic analyses of expressed antibody repertoires are becoming increasingly common experimental investigations and are critical to furthering our understanding of autoimmunity, infectious disease, and cancer. Next-generation DNA sequencing (NGS) technologies have now made it possible to interrogate antibody repertoires to unprecedented depths, typically by sequencing of cDNAs encoding immunoglobulin variable domains. In this chapter, we describe simple, fast, and reliable methods for producing and sequencing multiplex PCR amplicons derived from the variable regions (VH,VHHorVL) of rearranged immunoglobulin heavy and light chain genes using the Illumina MiSeq platform. We include complete protocols and primer sets for amplicon sequencing of VH/VHH/VL repertoires directly from human, mouse, and llama lymphocytes as well as from phage-displayed VH/VHH/VL libraries; these can be easily be adapted to other types of amplicons with little modiﬁcation. The resulting amplicons are diverse and representative, even using as few as 103 input B cells, and their generation is relatively inexpensive, requiring no special equipment and only a limited set of primers. In the absence of heavy-light chain pairing, single- domain antibodies are uniquely amenable to NGS analyses. We present a number of applications of NGS technology useful in discovery of single-domain antibodies from phage display libraries, including: (i) assessment of library functionality; (ii) conﬁrmation of desired library randomization; (iii) estimation of library diversity; and (iv) monitoring the progress of panning experiments. While the case studies presented here are of phage-displayed single-domain antibody libraries, the principles extend to other types of in vitro display libraries.

Key words Next-generation DNA sequencing, Single-domain antibody, Phage display library

1 Introduction

Immunoglobulin molecules (antibodies and their cell-surface counterparts, the B-cell receptors) are glycoproteins made up of heavy and single light chains arranged as “dimers of heterodimers,” with the variable domain of the heavy chain (VH) typically

This is National Research Council Canada Publication Number: 53333.

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_24, © Springer Science+Business Media LLC 2018 425 426 Kevin A. Henry

contributing more to antigen binding than that of the light chain (VL)[1]. In addition to conventional antibodies, camelid ungulates and cartilaginous fish naturally produce heavy-chain-only antibodies in which the heavy chain variable domain (VHH or VNAR) is not paired with a cognate VL domain [2, 3]. Rare autonomous VH and VL domains have also been isolated or engineered from species that do not produce such molecules naturally (e.g., human [4], mouse [5]). In vitro display libraries, including phage-display libraries, are now routinely constructed from large repertoires of VH and/or VL domains, either as single domains (for VHH, VNAR and autonomous VH/VL domains) or as pairs of VH:VL domains in Fab and scFv libraries [6]. During B-cell development, the exon encoding each variable domain is generated by somatic recombination of germline gene segments; joining of these segments occurs “imprecisely” through exonuclease processing and incorporation of palindromic and non-templated nucleotides at inter-segment junctions [7]. After B-cell maturation, variable domains are further diversified upon encounter with antigen through somatic hypermutation and affinity maturation [8]. The genetic characteristics of immunoglobulin variable domain repertoires (i.e., sequence diversity among homologous germline gene segments; high clonal diversity of B-cell populations; potential for somatic mutation) make them challenging templates for next-generation DNA sequencing (NGS). The major challenges involved include: (i) sequence variation between germline genes at the 50 end (leader sequences and FR1), necessi- tating either multiplex PCR or 50 RACE approaches [9], both of which may bias repertoire diversity; (ii) ontogeny of immunoglobulin variable domains through somatic recombination of gene segments, making regions of each B-cell receptor unique at the nucleotide level with no genomic comparison possible; and (iii) difficulties in distinguishing rare somatic variants from sequencing errors, which can be mitigated by barcoding of individual cDNA molecules [10]. Nevertheless, several approaches have been developed for interrogating immunoglobulin VH and VL repertoires, most of which rely on the availability of large amounts of input material (e.g., whole peripheral blood lymphocytes [9]; high- frequency B-cell subsets obtained from large blood samples [11, 12]; bulk antigen-specific B cells [13]; phage-displayed libraries [14]) or are limited in throughput to thousands of sequences (e.g., linked VH and VL amplicons generated from single B cells [15, 16]). The development of these methods has fueled rapid progress in many areas of immunological research. In this chapter, we describe a basic set of techniques for NGS of immunoglobulin VH and VL domains using the Illumina MiSeq platform, both directly from lymphocytes and from phage- displayed single-domain antibody (sdAb) libraries (VHH, VH or VL). The methods outlined here are simple and inexpensive, Next-Generation DNA Sequencing of VH/VL sdAb Repertoires 427

requiring only standard molecular biology skills and equipment available in most labs, and are meant to serve both as a starting point for investigators wishing to undertake NGS analyses of antibody repertoires as well as a guide to useful tips and tricks for applying NGS to in vitro display libraries. In particular, we illustrate a number of examples in which NGS is used to probe the diversity and functionality of sdAb libraries, as well as to observe the behavior of individual library variants under selection; this information is highly useful in the design of novel in vitro display libraries and selection of binders from such libraries. While the case studies presented in this chapter are of phage-displayed sdAb libraries, the principles underlying them apply generally to all in vitro display libraries.

2 Materials

Prepare all solutions using ultrapure water (prepared by purifying deionized water to attain a sensitivity of 18 M Ω cm at 25 C) and analytical grade reagents. Prepare and store all reagents at room temperature unless indicated otherwise. Carefully follow all waste disposal regulations when disposing of hazardous waste materials. Institutional and other relevant research ethics approvals must be obtained prior to starting work with animal and/or human samples. All oligonucleotides listed in this protocol were purchased commercially, desalted but with no additional puriﬁcation unless otherwise indicated.

2.1 Preparation 1. Freshly obtained whole blood sample (from human, mouse or of Peripheral Blood llama; see Note 1). Mononuclear Cells 2. BD Vacutainer™ blood collection tubes with sodium citrate (PBMCs) (BD Biosciences, San Jose, USA; see Note 2). 3. Sterile hypodermic needles of appropriate length and gauge (see Note 3). 4. 0.5 M ethylenediaminetetraacetic acid (EDTA) stock solution (per L): 186.1 g EDTA-Na2·2H2O in ultrapure H2O. Adjust pH to 8.0 with NaOH and store at room temperature. 5. Phosphate-buffered saline (PBS) supplemented with 2 mM EDTA (per L): 8 g (137 mM) NaCl, 0.2 g (2.37 mM) KCl, 1.44 g (10 mM) Na2HPO4, 0.24 g (1.8 mM) KH2PO4,4mL 0.5 M EDTA stock solution dissolved in ultrapure H2O. Adjust pH to 7.4 with HCl, sterilize by autoclaving, and store at room temperature. ® 6. Histopaque -1077 (Sigma-Aldrich, St. Louis, USA) or Ficoll- ® Paque PLUS (GE Healthcare, Chicago, USA) density separation media, stored at 4 C. 428 Kevin A. Henry

® 7. Lympholyte -M (Cedarlane Labs, Burlington, Canada) density separation medium. 8. 50 mL Falcon tubes. 9. Sorvall™ Legend™ refrigerated tabletop centrifuge with swinging bucket rotor (Thermo-Fisher Scientiﬁc, Waltham, USA), or similar instrument. 10. Pasteur pipettes. 11. TC20™ automated cell counter (Bio-Rad Laboratories, Her- cules, USA), or similar instrument, with counting slides and 0.4% (w/v) Trypan blue solution. 12. Mr. Frosty™ freezing container (Thermo-Fisher Scientiﬁc), or similar instrument. 13. Isopropanol. 14. À80 C freezer. 15. Fetal bovine serum. 16. Tissue-culture grade dimethyl sulfoxide (DMSO). 17. Nunc™ cryobank vials (Sigma-Aldrich). 18. Cryogenic storage dewar containing an appropriate volume of liquid nitrogen. 19. 37 C water bath. 20. RPMI media supplemented with 10% (v/v) fetal bovine serum, 100 U/mL penicillin and 100 μg/mL streptomycin.

® 2.2 RNA Extraction 1. RNeasy Mini Kit (Qiagen, Hilden, Germany), or similar and cDNA Synthesis product. 2. 0.5 M ethylenediaminetetraacetic acid (EDTA) stock solution (per L): 186.1 g EDTA-Na2·2H2O in ultrapure H2O. Adjust pH to 8.0 with NaOH and store at room temperature. 3. RNase-free TE buffer (per L): 1.21 g (10 mM) Tris, 2 mL 0.5 M EDTA stock solution in ultrapure H2O. Adjust pH to 7.5 with HCl and store at room temperature. ® 4. NanoDrop ND-1000 spectrophotometer (Thermo-Fisher Scientiﬁc) or similar instrument. 5. À80 C freezer. 6. qScript™ cDNA supermix (Quanta Biosciences, Beverley, USA), or similar product. 7. 0.2 mL PCR tubes, strips or plates. ® 8. GeneAmp PCR System 9700 thermal cycler (Thermo-Fisher Scientiﬁc) or similar instrument. 9. À20 C freezer. Next-Generation DNA Sequencing of VH/VL sdAb Repertoires 429

2.3 Construction See Subheading 3.3 and references therein. and Panning of Phage- Displayed Single- Domain Antibody Libraries

2.4 Next-Generation 1. Gene-specific 1st round PCR primers as in Table 1 bearing DNA Sequencing universal tags. The gene-specific regions of these primers are adapted from those found in the following sources: human lymphocytes [17, 18]; mouse lymphocytes [19]; llama lymphocytes [20]. 2. Barcoded 2nd round tagging PCR primers as in Table 2. All primers in this table should be purified either by HPLC or by polyacrylamide gel electrophoresis. ® 3. AmpliTaq Gold DNA polymerase with buffer II (Thermo- Fisher Scientific). 4. dNTP mix, 10 mM each. ® 5. GeneAmp PCR System 9700 thermal cycler or similar instrument. 6. 0.2 mL PCR tubes, strips or plates. 7. Agarose gel electrophoresis equipment and power supply. 8. 50Â TAE buffer (per L): 242 g (2 M) Tris base, 57.1 mL glacial acetic acid, 100 mL 0.5 M EDTA solution in ultrapure H2O. Store at room temperature. Dilute 1:50 in H2Oto prepare 1Â TAE buffer. 9. 1% (w/v) agarose gel, prepared in 1Â TAE buffer. ® 10. PureLink PCR purification kit (Life Technologies, Carlsbad, USA). 11. 1.5 mL Eppendorf tubes. ® 12. Phusion high-fidelity DNA polymerase (Thermo-Fisher Scientific). ® 13. QIAquick gel extraction kit (Qiagen). ® 14. Dark Reader transilluminator (Clare Chemical, Dolores, USA), or similar instrument. 15. GelGreen™ nucleic acid gel stain (Biotium, Fremont, USA). ® ® 16. Agencourt AMPure XP beads (Beckman Coulter, Brea, USA). 17. 70% (v/v) ethanol.

18. Sterile ultrapure Milli-Q H2O. ® 19. NanoDrop ND-1000 spectrophotometer, or similar instrument. Table 1 First round gene-specific PCR primers used to amplify rearranged VH,VL or VHH genes from lymphocytes or phage-displayed libraries

Template Locus Primer name Sequence (50-30)a,b

Human lymphocytes Heavy chain (igh) seqF-VH1 CGCTCTTCCGATCTCTG(N4–6 )GGCCTCAGTGAAGGTCTCCTGCAAG seqF-VH2 CGCTCTTCCGATCTCTG(N4–6 )GTCTGGTCCTACGCTGGTGAAACCC seqF-VH3 CGCTCTTCCGATCTCTG(N4–6 )CTGGGGGGTCCCTGAGACTCTCCTG seqF-VH4 CGCTCTTCCGATCTCTG(N4–6 )CTTCGGAGACCCTGTCCCTCACCTG seqF-VH5 CGCTCTTCCGATCTCTG(N4–6 )CGGGGAGTCTCTGAAGATCTCCTGT seqF-VH6 CGCTCTTCCGATCTCTG(N4–6 )TCGCAGACCCTCTCACTCACCTGTG seqR-JH TGCTCTTCCGATCTGAC(N4–6 )CTTACCTGAGGAGACGGTGACC seqR-IgM TGCTCTTCCGATCTGAC(N4–6 )GGTTGGGGCGGATGCACTCC seqR-IgG TGCTCTTCCGATCTGAC(N4–6 )SGATGGGCCCTTGGTGGARGC seqR-IgA TGCTCTTCCGATCTGAC(N4–6 )CTTGGGGCTGGTCGGGGATG Kappa chain (igk) seqF-VK1 CGCTCTTCCGATCTCTG(N4–6 )ATGAGGSTCCCYGCTCAGCTCCTGGG seqF-VK2 CGCTCTTCCGATCTCTG(N4–6 )CTCTTCCTCCTGCTACTCTGGCTCCCAG seqF-VK3 CGCTCTTCCGATCTCTG(N4–6 )ATTTCTCTGTTGCTCTGGATCTCTG seqR-CK TGCTCTTCCGATCTGAC(N4–6 )CAGCAGGCACACAACAGAGGCAGTTCC Lambda chain (igl) seqF-VL1 CGCTCTTCCGATCTCTG(N4–6 )GCACAGGGTCCTGGGCCCAGTCTG seqF-VL2 CGCTCTTCCGATCTCTG(N4–6 )GCTCTGTGACCTCCTATGAGCTG seqF-VL3 CGCTCTTCCGATCTCTG(N4–6 )GGTCTCTCTCSCAGCYTGTGCTG seqF-VL4 CGCTCTTCCGATCTCTG(N4–6 )GTTCTTGGGCCAATTTTATGCTG seqF-VL5 CGCTCTTCCGATCTCTG(N4–6 )GAGTGGATTCTCAGACTGTGGTG seqF-VL6 CGCTCTTCCGATCTCTG(N4–6 )GCTCACTGCACAGGGTCCTGGGCC seqF-VL7 CGCTCTTCCGATCTCTG(N4–6 )GCTTACTGCACAGGATCCGTGGCC seqF-VL8 CGCTCTTCCGATCTCTG(N4–6 )ACTCTTTGCATAGGTTCTGTGGTT seqF-VL9 CGCTCTTCCGATCTCTG(N4–6 )TCTCACTGCACAGGCTCTGTGACC seqF-VL10 CGCTCTTCCGATCTCTG(N4–6 )ACTTGCTGCCCAGGGTCCAATTC seqR-CL TGCTCTTCCGATCTGAC(N4–6 )CACCAGTGTGGCCTTGTTGGCTTG

Mouse lymphocytes Heavy chain (igh) seqF-VH1 CGCTCTTCCGATCTCTG(N4–6 )AGRTYCAGCTGCARCAGTCT seqF-VH2 CGCTCTTCCGATCTCTG(N4–6 )AGGTCCAACTGCAGCAGCC seqF-VH3 CGCTCTTCCGATCTCTG(N4–6 )TCTGCCTGGTGACWTTCCCA seqF-VH4 CGCTCTTCCGATCTCTG(N4–6 )GTGCAGCTTCAGGAGTCAG seqF-VH5 CGCTCTTCCGATCTCTG(N4–6 )GAGGTGAAGCTTCTCGAGTC seqF-VH6 CGCTCTTCCGATCTCTG(N4–6 )GAAGTGAAGCTGGTGGAGTC seqF-VH7 CGCTCTTCCGATCTCTG(N4–6 )ATGKACTTGGGACTGARCTGT seqF-VH8 CGCTCTTCCGATCTCTG(N4–6 )CAGTGTGAGGTGAAGCTGGT seqF-VH9 CGCTCTTCCGATCTCTG(N4–6 )CCAGGTTACTCTGAAAGAGTC seqF-VH10 CGCTCTTCCGATCTCTG(N4–6 )TGTGGACCTTGCTATTCCTGA seqF-VH11 CGCTCTTCCGATCTCTG(N4–6 )TGTTGGGGCTGAAGTGGGTTT seqF-VH12 CGCTCTTCCGATCTCTG(N4–6 )ATGGAGTGGGAACTGAGCTTA seqF-VH13 CGCTCTTCCGATCTCTG(N4–6 )AGCTTCAGGAGTCAGGACC seqF-VH14 CGCTCTTCCGATCTCTG(N4–6 )CAGGTGCAGCTTGTAGAGAC seqF-VH15 CGCTCTTCCGATCTCTG(N4–6 )ATGCAGCTGGGTCATCTTCTT seqF-VH16 CGCTCTTCCGATCTCTG(N4–6 )GACTGGATTTGGATCACKCTC seqF-VH17 CGCTCTTCCGATCTCTG(N4–6 )TGGAGTTTGGACTTAGTTGGG seqR-JH TGCTCTTCCGATCTGAC(N4–6 )CTYACCTGAGGAGACDGTGA seqR-IgM TGCTCTTCCGATCTGAC(N4–6 )CATGGCCACCAGATTCTTATC seqR-IgG1 TGCTCTTCCGATCTGAC(N4–6 )AGGGAAATARCCCTTGACCAG seqR-IgG2 TGCTCTTCCGATCTGAC(N4–6 )AGGGAAGTAGCCTTTGACAAG seqR-IgA TGCTCTTCCGATCTGAC(N4–6 )GAATCAGGCAGCCGATTATCAC Kappa chain (igk) seqF-VK1 CGCTCTTCCGATCTCTG(N4–6 )TGATGACCCARACTCCACT seqF-VK2 CGCTCTTCCGATCTCTG(N4–6 )GCTTGTGCTCTGGATCCC seqF-VK3 CGCTCTTCCGATCTCTG(N4–6 )CTGCTGCTCTGGGTTCC seqF-VK4 CGCTCTTCCGATCTCTG(N4–6 )CAGCTTCCTGCTAATCAGTG seqF-VK5 CGCTCTTCCGATCTCTG(N4–6 )CTCAGATCCTTGGACTTHTG seqF-VK6 CGCTCTTCCGATCTCTG(N4–6 )TGGAGTCACAGACYCAGG seqF-VK7 CGCTCTTCCGATCTCTG(N4–6 )TGGAGTTTCAGACCCAGG seqF-VK8 CGCTCTTCCGATCTCTG(N4–6 )CTGCTMTGGGTATCTGGT seqF-VK9 CGCTCTTCCGATCTCTG(N4–6 )CWTCTTGTTGCTCTGGTTTC seqF-VK10 CGCTCTTCCGATCTCTG(N4–6 )GATGTCCTCTGCTCAGTTC seqF-VK11 CGCTCTTCCGATCTCTG(N4–6 )CCTGCTGAGTTCCTTGGG seqF-VK12 CGCTCTTCCGATCTCTG(N4–6 )CTGCTGCTGTGGCTTACA seqF-VK13 CGCTCTTCCGATCTCTG(N4–6 )CCTTCTCAACTTCTGCTCT seqF-VK14 CGCTCTTCCGATCTCTG(N4–6 )AGGGCCCYTGCTCAGTTT seqF-VK15 CGCTCTTCCGATCTCTG(N4–6 )ATGAGGGTCCTTGCTGAG seqF-VK16 CGCTCTTCCGATCTCTG(N4–6 )GAGGTTCCAGGTTCAGGT seqF-VK17 CGCTCTTCCGATCTCTG(N4–6 )CCATGACCATGYTCTCACT seqF-VK18 CGCTCTTCCGATCTCTG(N4–6 )ATGGAAACTCCAGCTTCATTT seqF-VK19 CGCTCTTCCGATCTCTG(N4–6 )ATGAGACCGTCTATTCAGTT seqR-CK TGCTCTTCCGATCTGAC(N4–6 )GCACCTCCAGATGTTAACTG Lambda chain (igl) seqF-VL1 CGCTCTTCCGATCTCTG(N4–6 )GCCTGGAYTTCACTTATACTC seqF-VL2 CGCTCTTCCGATCTCTG(N4–6 )TGGCCTGGACTCCTCTCTT seqF-VL3 CGCTCTTCCGATCTCTG(N4–6 )ACTCAGCCAAGCTCTGTG seqR-CL TGCTCTTCCGATCTGAC(N4–6 )AGCTCCTCAGRGGAAGGTG

Llama lymphocytes Heavy-chain only seqF-MJ1 CGCTCTTCCGATCTCTG(N4–6 )SMKGTGCAGCTGGTGGAKTCTGGGGGA (igh) seqF-MJ2 CGCTCTTCCGATCTCTG(N4–6 )CAGGTAAAGCGGAGGAGTCTGGGGGA seqF-MJ3 CGCTCTTCCGATCTCTG(N4–6 )CAGGCTCAGGTACAGCTGGTGGAGTCT seqR-CH2 TGCTCTTCCGATCTGAC(N4–6 )CGCCATCAAGGTACCAGTTGGA seqR-CH2b3 TGCTCTTCCGATCTGAC(N4–6 )GGGGTACCTGTCATCCACGGACCAGCTGA c Phage-displayed Any seqF-FdT CGCTCTTCCGATCTCTG(N4-6 )GCAATTCCTTTAGTTGTTCCTTTCTATTCTCAC c VHH/VH/VL library seqR-FdT TGCTCTTCCGATCTGAC(N4-6 )GAGGTTTTGCTAAACAACTTTCAACAGTTTC aUniversal tag sequences are indicated in underlined italics, random nucleotides in bold and gene-specific sequences annealing to target template in ordinary type b N4–6 indicates a stretch of 4–6 N-nucleotides. Data quality is marginally improved by substituting an equimolar mixture of three primers in the place of primers listed in this table, each bearing a different number of N-nucleotides (4, 5 or 6) c These primers are designed to amplify from sdAb (VHH, VH,orVL) libraries in fd-tet vectors. For other vectors, substitute vector-specific primers immediately flanking genes encoding antibody variable regions Table 2

Universal second-round barcoded PCR primers used to tag first-round amplicons with MiSeq indexes and adapter sequences Henry A. Kevin 432

Primer name Sequence (50-30)a

P5-seqF AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCTCTG

P7-index1-seqR CAAGCAGAAGACGGCATACGAGATCGTGATGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index2-seqR CAAGCAGAAGACGGCATACGAGATACATCGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index3-seqR CAAGCAGAAGACGGCATACGAGATGCCTAAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index4-seqR CAAGCAGAAGACGGCATACGAGATGCCTAAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index5-seqR CAAGCAGAAGACGGCATACGAGATCACTGTGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index6-seqR CAAGCAGAAGACGGCATACGAGATATTGGCGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index7-seqR CAAGCAGAAGACGGCATACGAGATGATCTGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index8-seqR CAAGCAGAAGACGGCATACGAGATTCAAGTGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index9-seqR CAAGCAGAAGACGGCATACGAGATCTGATCGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index10-seqR CAAGCAGAAGACGGCATACGAGATAAGCTAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index11-seqR CAAGCAGAAGACGGCATACGAGATGTAGCCGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index12-seqR CAAGCAGAAGACGGCATACGAGATTACAAGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index13-seqR CAAGCAGAAGACGGCATACGAGATTTGACTGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index14-seqR CAAGCAGAAGACGGCATACGAGATGGAACTGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index15-seqR CAAGCAGAAGACGGCATACGAGATTGACATGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index16-seqR CAAGCAGAAGACGGCATACGAGATGGACGGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index17-seqR CAAGCAGAAGACGGCATACGAGATCTCTACGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index18-seqR CAAGCAGAAGACGGCATACGAGATGCGGACGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index19-seqR CAAGCAGAAGACGGCATACGAGATTTTCACGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index20-seqR CAAGCAGAAGACGGCATACGAGATGGCCACGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index21-seqR CAAGCAGAAGACGGCATACGAGATCGAAACGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index22-seqR CAAGCAGAAGACGGCATACGAGATCGTACGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index23-seqR CAAGCAGAAGACGGCATACGAGATCCACTCGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC P7-index24-seqR CAAGCAGAAGACGGCATACGAGATGCTACCGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index25-seqR CAAGCAGAAGACGGCATACGAGATATCAGTGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index26-seqR CAAGCAGAAGACGGCATACGAGATGCTCATGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index27-seqR CAAGCAGAAGACGGCATACGAGATAGGAATGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index28-seqR CAAGCAGAAGACGGCATACGAGATCTTTTGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index29-seqR CAAGCAGAAGACGGCATACGAGATTAGTTGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index30-seqR CAAGCAGAAGACGGCATACGAGATCCGGTGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index31-seqR CAAGCAGAAGACGGCATACGAGATATCGTGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC etGnrto N eunigo V of Sequencing DNA Next-Generation P7-index32-seqR CAAGCAGAAGACGGCATACGAGATTGAGTGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index33-seqR CAAGCAGAAGACGGCATACGAGATCGCCTGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index34-seqR CAAGCAGAAGACGGCATACGAGATGCCATGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index35-seqR CAAGCAGAAGACGGCATACGAGATAAAATGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index36-seqR CAAGCAGAAGACGGCATACGAGATTGTTGGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index37-seqR CAAGCAGAAGACGGCATACGAGATATTCCGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index38-seqR CAAGCAGAAGACGGCATACGAGATAGCTAGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index39-seqR CAAGCAGAAGACGGCATACGAGATGTATAGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index40-seqR CAAGCAGAAGACGGCATACGAGATTCTGAGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index41-seqR CAAGCAGAAGACGGCATACGAGATGTCGTCGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index42-seqR CAAGCAGAAGACGGCATACGAGATCGATTAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC H /V

GCTGTA L

P7-index43-seqR CAAGCAGAAGACGGCATACGAGAT GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC 433 Repertoires sdAb

P7-index44-seqR CAAGCAGAAGACGGCATACGAGATATTATAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index45-seqR CAAGCAGAAGACGGCATACGAGATGAATGAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index46-seqR CAAGCAGAAGACGGCATACGAGATTCGGGAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index47-seqR CAAGCAGAAGACGGCATACGAGATCTTCGAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC

P7-index48-seqR CAAGCAGAAGACGGCATACGAGATTGCCGAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC aMiSeq P5/P7 adapters are indicated in italics, index sequences in bold underline and universal tag sequences annealing to ﬁrst-round PCR products in regular type 434 Kevin A. Henry

20. Bioanalyzer 2100 (Agilent, Santa Clara, USA), or similar instrument. 21. MiSeq Sequencing System (Illumina, San Diego, USA). 22. 500-cycle Reagent Kit V2 or 600-cycle Reagent Kit V3 (Illu- mina, San Diego, USA).

2.5 Data Analysis 1. FastQC version quality control tool [21], version 0.11.5. Avail- able online at http://www.bioinformatics.babraham.ac.uk/pro jects/fastqc/. 2. FLASH paired-end read overlapping tool [22], version 1.2.11. Available online at https://ccb.jhu.edu/software/FLASH/. 3. FASTX toolkit, including FASTQ quality ﬁltering tool [23], version 0.0.14. Available online at http://hannonlab.cshl.edu/ fastx_toolkit/. 4. R version 3.3.2, including packages “seqinr” and “stringdist.” Available online at https://cran.r-project.org/.

3 Methods

In the protocols that follow, we describe workﬂows for interrogating repertoires of single-immunoglobulin variable domains (VH, VL, or VHH) using next-generation DNA sequencing, both directly from lymphocytes and from phage-displayed libraries. While the applications we highlight pertain to phage-displayed single-domain antibody libraries, we also include methods for sequencing directly from lymphocytes; these are valuable both for investigations of in vivo immune responses and for evaluating the quality of phage- displayed libraries constructed from natural sources of diversity (Fig. 1).

3.1 Preparation 1. Collect whole blood samples directly into citrated tubes using a of Peripheral Blood sterile hypodermic needle of appropriate length and gauge (see Mononuclear Cells Note 4). (PBMCs) 2. Dilute blood samples with two volumes of ice-cold PBS containing 2 mM EDTA. 3. Gently layer 35 mL of diluted blood sample onto 15 mL ® ® Histopaque -1077 or Ficoll-Pacque PLUS (human and ® llama blood samples) or Lympholyte -M (mouse blood samples) in a 50 mL Falcon tube. Centrifuge at 400 Â g for 30 min at room temperature with no brake. 4. Using a Pasteur pipette, carefully aspirate and discard the top- most plasma layer. 5. Using a Pasteur pipette, carefully aspirate the PBMC layer (interphase) and transfer to a new 50 mL Falcon tube Next-Generation DNA Sequencing of VH/VL sdAb Repertoires 435

Fig. 1 Evaluation of the potential effects of bias on VH domains sequenced directly from human B cells. (a)A 4 + rearranged IGHV3-encoded VH domain-containing plasmid was doped into cDNA derived from 10 CD19 human B cells at the indicated ratios (assuming 300 copies of immunoglobulin heavy-chain mRNA per cell), then VHs were ampliﬁed and sequenced as described in this chapter. Recovery rates of the plasmid-encoded VH show no major deviation from the expected frequency. (b) IGHV gene family usage from three different subpopulations of CD19+ B cells from a single individual, sequenced as described in this chapter. Data are representative of approximately 50,000–200,000 sequences per B-cell pool. Each subpopulation has widely different gene usage, indicating that all of the multiplexed PCR primers are able to amplify their targets

containing 30 mL PBS containing 2 mM EDTA. Centrifuge at 200 Â g for 10 min at room temperature, then carefully remove the supernatant. Repeat this step twice. 6. Prior to ﬁnal wash, count cells to estimate density and viability (should be 95%; see Note 5). 7. Resuspend cells in sterile freeze medium (90% (v/v) fetal bovine serum, 10% (v/v) DMSO) equilibrated to room temperature at a density of 5–10 Â 106 cells/mL. Aliquot into cryovials. 8. Place cryovials inside a Mr. Frosty™ freezing container containing fresh isopropanol, and place in À80 C freezer overnight. 9. The next day, move the cells on dry ice to liquid nitrogen storage. 436 Kevin A. Henry

10. To thaw the cells, move cryovials as quickly as possible from liquid nitrogen storage into a 37 C water bath. Gently agitate until no visible ice crystals remain, then add to 15 mL warm RPMI media supplemented with 10% (v/v) fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin. Pellet the cells at 200 Â g for 5 min at room temperature, then resuspend in PBS or other isotonic buffer (see Note 6).

3.2 RNA Extraction 1. Extract total cellular RNA from 104–106 PBMCs or and cDNA Synthesis 103–105 puriﬁed B-cells and elute in 30 μL RNase-free TE buffer. Store RNA at À80 C if not using immediately for reverse transcription.

2. Measure RNA concentration and A260/A280 ratio using a spectrophotometer (see Note 7). 3. In a 0.2 mL PCR tube, combine 16 μL RNA with 4 μL qScript™ cDNA supermix. In a thermal cycler, incubate at 25 C for 5 min, 42 C for 1 h, then 85 C for 5 min. Store cDNA at À20 C.

3.3 Construction Detailed protocols for constructing, rescuing, and panning phage- and Panning of Phage- displayed camelid VHH libraries in a phagemid format [20, 24] and Displayed Single- human synthetic VH/VL libraries in a phage format [25, 26] have Domain Antibody been previously published. Libraries

3.4 Next-Generation High-throughput sequencing of phage-displayed sdAb libraries DNA Sequencing during all stages of their construction (at the DNA level, prior to transformation of Escherichia coli; from phage or phagemid DNA in E. coli cells; from phage or phagemid DNA encapsulated in the phage virion) as well as during their selection can yield highly valuable information. Both phage particles themselves, phage/phagemid single-stranded DNA and phage/phagemid replicative form DNA isolated from E. coli cells are appropriate templates for NGS. We caution that phage particles can inhibit PCR, and that sequences within the phage genome can generally only be successfully amplified within a relatively narrow window of ~105–107 particles per PCR reaction. We also caution that replicative form DNA isolated from infected E. coli cells (e.g., overnight cultures for the purpose of phage amplification) will reflect growth advantages conferred by library variants, which can significantly bias the composition of synthetic sdAb libraries.

1. Amplify genes encoding rearranged VH/VL/VHH domains in 25-μL PCR reactions containing 1Â ABI Buffer II, 1.5 mM MgCl2, 200 μM each dNTP, 5 pmol each primer or primer mixture from Table 1, 1 U of AmpliTaq Gold DNA polymerase and 1 μL of template. For ampliﬁcation directly from Next-Generation DNA Sequencing of VH/VL sdAb Repertoires 437

lymphocytes, equimolar mixtures of all forward primers annealing in FR1 for a given locus (e.g., human igh) should be used in a single reaction. Depending on the application, either a single reverse primer or a mixture of isotype-specific primers annealing in constant regions may be used. Template may be cDNA derived from lymphocytes, single-stranded or replicative form DNA (1–100 ng/μL) or phage particles (~106 particles/μL). Cycle the reactions as follows: 95 C for 7 min; 35 cycles of (94 C for 30 s, 55 C for 45 s, and 72 C for 2 min); 72 C for 10 min. 2. Electrophorese 5 μL aliquots of the PCRs in 1% (w/v) agarose gels in TAE buffer and confirm amplification of a ~400 bp band. ® 3. Purify the amplicons using a PureLink PCR purification kit (300 bp cutoff; see Note 8). 4. Conduct second round “tagging” PCRs 50 μL reaction volumes containing 1Â Phusion HF Buffer, 1.5 mM MgCl2, 200 μM each dNTP, 10 pmol of each primer pair (e.g., P5-seqF and P7-index1-seqR; see Note 9), 0.25 U Phusion High- Fidelity DNA polymerase, and 5 μL first-round PCR as a template. Cycle as follows: 98 C for 30 s; 20 cycles of (98 C for 10 s, 65 C for 30 s, and 72 C for 30 s); 72 C for 5 min. 5. Electrophorese 5 μL aliquots of the PCRs in 1% (w/v) agarose gels in TAE buffer and confirm amplification of a ~450–500 bp band. 6. Pool equal volumes of all second round amplicons (see Notes ® 10 and 11) and purify using a PureLink PCR purification kit (300 bp cutoff). Subsequently, gel purify the pooled library ® from a 1% (w/v) agarose gel in TAE buffer using a QIAquick gel extraction kit and elute in 50 μL EB buffer. Perform final cleanup of the pooled NGS library using 90 μL of AMPure XP beads and elute in 20 μL ultrapure H2O. 7. Measure pooled amplicon purity and concentration using a High Sensitivity DNA Analysis kit on a BioAnalyzer 2100 instrument (Fig. 2a). 8. Sequence the pooled amplicons on an Illumina MiSeq Sequencing System using a 500-cycle Reagent Kit V2 or 600-cycle Reagent Kit V3 and a 5% PhiX genomic DNA spike. Diluting the amplicons to ~7–8 pM should yield a cluster density between 800 and 1000 K/mm2.

3.5 Data Analysis Construction and selection of phage-displayed antibody libraries has historically been a “black box,” with limited insight into the processes governing the generation of a library and the behavior of its members. NGS provides a window into these processes by allowing us to sample a signiﬁcant proportion of library diversity 438 Kevin A. Henry

Fig. 2 Quality control metrics for VH/VL/VHH amplicon sequencing using the Illumina MiSeq platform. (a) Representative Bioanalyzer 2100 trace of a VHH amplicon library produced from a phage-displayed library and purified as described in this chapter. The bulk of the DNA should be distributed around the expected library size (in this case, ~500–600 bp); the presence of smaller molecular weight bands suggests primer multimer contamination and means samples should be repurified. (b) Representative results of FastQC for amplicon sequencing of VH/VL/VHH domains on a MiSeq instrument as described in this chapter. Results for forward paired-end reads are shown but reverse reads will be similar; quality scores decline with read length but should not dip appreciably below Q25. If quality scores fall below Q20 at the 30 end of most reads, merging of forward and reverse reads will be inefficient and resequencing is recommended Next-Generation DNA Sequencing of VH/VL sdAb Repertoires 439 over the course of various type of manipulations. Template R scripts enabling the following analyses can be obtained from the authors by request. 1. Visualize per-base quality scores for forward and reverse reads using FastQC (Fig. 2b; see Note 12). In many cases, poor data quality cannot be compensated for in analysis and will artificially inflate diversity estimates. 2. Merge forward and reverse paired end reads using FLASH with default parameters. High-quality data with an appropriate degree of overlap (20 bp) should yield 90% merged sequences (see Note 13). 3. Quality filter the merged sequences using the FASTQ quality filtering tool within the FASTX toolkit. Accept only sequences with Q30 scores over 95% of bases in the read. Generally, this will result in discarding of ~30% of the lowest-quality reads. 4. Convert data from .fastq to .fasta format and read into R using the read.fasta function of package “seqinr.” Strip all primer- encoded non-antibody sequences. 5. Translate nucleotide sequences to protein using the translate function of package “seqinr.” 6. Assess library functionality by determining the proportion of nucleotide sequences that are in-frame (Fig. 3). All in vitro display libraries, regardless of their method of construction, will contain genetic imperfections at some frequency (e.g., single-nucleotide indels introduced by PCR primers). The frequency of out-of-frame library variants should remain static when phagemid library DNA is transformed into E. coli cells, assuming sufficient glucose is present to inhibit protein production, but can increase dramatically when cells are grown to high density and the library is rescued with helper phage. This phe- nomenon can be counteracted somewhat by rescue with M13K07ΔpIII hyper phage, which forces pIII-sdAb expression from phagemid DNA. Large increases in the proportion of out- of-frame sequences after rescue with helper phage are indicative of strong growth advantages for cells that do not produce a potentially toxic pIII-sdAb protein; loss of diversity through this mechanism is a major challenge for synthetic libraries. 7. Using protein sequences, determine the frequency of stop codons among library members, and as above (for phagemid libraries), assess whether the frequency of stop codons increases when rescued with helper phage. This can be accomplished by pattern matching the “*”character. Library members bearing stop codons are expected to be present to some degree, depending on the randomization strategy, but should represent a minor proportion of the library. 440 Kevin A. Henry

Fig. 3 Evaluation of phagemid sdAb library functionality using next-generation DNA sequencing. (a) A human VL domain was randomized synthetically in vitro and the resulting library transformed into E. coli TG1 cells, then library phage particles were rescued using M13K07 helper phage. The randomization design called for three CDR3 lengths (DNA size shown with black arrows). Out-of-frame variants are rare in the library but very common in the rescued phage, suggesting that E. coli cells producing in-frame library proteins are at a signiﬁcant growth disadvantage. (b) A dromedary VHH domain was randomized synthetically in vitro and the Next-Generation DNA Sequencing of VH/VL sdAb Repertoires 441

8. Using protein sequences, determine the frequency of each amino acid at each randomization position and assess the degree of deviation from planned randomization (Fig. 4a). Strong regression toward the parental residue suggests intoler- ance to substitution, and library redesign should be considered. 9. Determine the frequency of the most commonly observed library sequences. Immune VHH libraries are expected to have major variations in frequency, reflecting immunodomi- nance of particular B-cell clones, but synthetic libraries should have a relatively flat landscape, and the presence of high- frequency variants (especially the parental scaffold) signifies lost library “space” and lower overall diversity. 10. Determine the proportion of library members with a given number of randomized residues using the stringdist function of package “stringdist” (Fig. 4b). For synthetic antibody libraries with high theoretical diversity that are interrogated to limited depth, this analysis is more robust to the effects of sequencing error and undersampling than more traditional measures of diversity, such as the number of unique sequences observed. 11. Reduce the analysis to only regions of highest sequence diversity (for instance, CDR3 for VHH libraries, or randomization positions for synthetic libraries) by parsing these regions to new data objects. This is done to reduce the impact of sequencing errors in the longer sequences on subsequent analyses. 12. Measure enrichment of library variants over the course of one or more types of selection (Fig. 5). The observed fold- enrichments will depend both on the library and the selection, but minimally should be 10. Carefully defined analyses of such data are strongly predictive of the properties of soluble

recombinant sdAbs. ä

Fig. 3 (continued) resulting library transformed into E. coli TG1 cells, then library phage particles were rescued using either M13K07 helper phage or M13K07ΔpIII hyper phage. The randomization design called for a single CDR3 length (DNA size shown with black arrow). Out-of-frame VHHs make up the large majority of the helper phage-rescued library, but a minor proportion of the hyper phage-rescued library. (c) An immune VHH phagemid library was constructed from the lymphocytes of an immunized llama, transformed into E. coli TG1 cells and then rescued with M13K07 helper phage. Through all steps, changes in the relative frequencies of individual VHHs were measured. Frequencies are stable for most VHHs in the library (red parentheses: 95% of data; blue parentheses: 75% of all data), with slightly larger bias introduced at the library construction step compared to the phage rescue step (A)

(B)

30 NGS Data Theoretical 20 25 % of Library 51015 0

0 5 10 15

No. of Randomised Positions

Fig. 4 Evaluation of phage-displayed sdAb library diversity using next-generation DNA sequencing. (a)A dromedary VHH domain was randomized synthetically in vitro and the resulting library transformed into E. coli TG1 cells. The randomization design called for full randomization (20 amino acids) at six positions. A signiﬁcant bias toward the parental residue is observed at four positions (P2, P3, P4, P5). (b) A human VL domain was randomized synthetically in vitro and the resulting library transformed into E. coli TG1 cells. The design of the library called for full randomization of up to 16 residues and partial randomization of two additional residues. The proportion of library members bearing the indicated number of randomized positions (“NGS data”) is shown, along with the expected distribution for a library reaching the theoretical maximum diversity under this design (“Theoretical,” 2 Â 2 Â 2016 members) Next-Generation DNA Sequencing of VH/VL sdAb Repertoires 443

Fig. 5 Enrichment of antigen-specific VHHs over the course of a selection experiment. An immune VHH phagemid library was constructed from the lymphocytes of a llama immunized with three antigens, transformed into E. coli TG1 cells and then rescued with M13K07 helper phage. Individual VHHs making up the library are shown as rectangles, with size proportional their frequency in the library. Colored squares represent antigen-specific VHHs that were isolated and characterized and several rounds of selection. Many of these antibodies can be identified by their enrichment after a single round of selection, instead of the conventional strategy of waiting for them to rise to high frequency after multiple rounds of panning

4 Notes

1. While numbers of PBMCs and circulating peripheral B cells vary by species and by individual, a typical yield might be ~106 PBMCs/mL whole blood, of which ~10% is made up by B lymphocytes. 2. Other anticoagulants (EDTA, heparin) have been used successfully, although the potential for PCR inhibition makes these reagents less suitable. 3. 1- or 1.5-inch, 16–22-gauge needles are appropriate for the collection of human and llama blood samples, while 25-gauge or smaller needles can be used for mice. 4. Fresh blood samples with anticoagulant can be kept on ice or at 4 C for several hours, but should be processed to the point of cryopreservation of PBMCs as quickly as possible. It is possible to store fresh blood samples overnight at the same temperature, but signiﬁcant loss of B cell diversity should be expected. 444 Kevin A. Henry

5. If they are not being cryopreserved, purified PBMCs should be either used immediately for nucleic acid extraction or stored at 4 C in RNAlater solution or an equivalent product. 6. B cells and/or B-cell subsets can be further purified using magnetic bead selection and/or fluorescence-activated cell sorting, the parameters for which will vary depending on the application. 7. Expect yields of approximately 5 μg RNA per 106 cells. In a volume of 30 μL, this is equivalent to a concentration of 167 ng/μL; thus, below a threshold of 105 cells, spectropho- tometric measurements may be unreliable.

8. For amplification of VHH genes directly from llama lymphocytes, PCRs using the CH2b3 primer will produce two bands around ~600 bp and ~800 bp. The smaller band must be purified by gel extraction rather than PCR purification to exclude conventional VH/VL antibodies. 9. The forward primer (P5-seqF) is the same for all second-round PCRs. Ensure each sample is barcoded using a single reverse primer from Table 2 bearing a unique index sequence. 10. If very even coverage across samples is important, second- round PCRs can be purified individually and then pooled in equimolar amounts prior to final cleanup. As a general rule, we find this to be unnecessary, as sample volumes can be adjusted based on band intensity on analytical agarose gels. 11. While we include primers for multiplexing up to 48 different samples in Table 2, the degree of multiplexing depends on sequence depth required. Multiplexing 12 samples using a 500-cycle Reagent Kit V2 will yield about 106 reads per sample prior to merging and quality filtering. 12. In many circumstances, poor data quality can be improved simply by increasing the amount of PhiX DNA spiked into the run. Reducing clustering density may also improve data quality.

13. Merging of synthetically randomized phage-displayed VH/VL libraries of deﬁned CDR length will generally be more successful than merging of natural repertoires and libraries constructed from natural sources of diversity, from which variable domains with extremely long CDR3s may be lost at this step.

Acknowledgments

This work was supported by funding from the National Research Council of Canada and by a CIHR doctoral research award (KAH). Next-Generation DNA Sequencing of VH/VL sdAb Repertoires 445

References

1. Kabat EA, Wu TT (1991) Identical V region 13. Liao HX, Lynch R, Zhou T et al (2013) amino acid sequences and segments of Co-evolution of a broadly neutralizing HIV-1 sequences in antibodies of different specifici- antibody and founder virus. Nature ties. Relative contributions of VH and VL 496:469–476 genes, minigenes, and complementarity- 14. Chen W, Prabakaran P, Zhu Z et al (2012) determining regions to binding of antibody- Characterization of human IgG repertoires in combining sites. J Immunol 147:1709–1719 an acute HIV-1 infection. Exp Mol Pathol 2. Hamers-Casterman C, Atarhouch T, Muylder- 93:399–407 mans S et al (1993) Naturally occurring anti- 15. Busse CE, Czogiel I, Braun P et al (2014) bodies devoid of light chains. Nature Single-cell based high-throughput sequencing 363:446–448 of full-length immunoglobulin heavy and 3. Greenberg AS, Avila D, Hughes M et al (1995) light chain genes. Eur J Immunol 44:597–603 A new antigen receptor gene family that under- 16. DeKosky BJ, Ippolito GC, Deschner RP et al goes rearrangement and extensive somatic (2013) High-throughput sequencing of the diversification in sharks. Nature 374:168–173 paired human immunoglobulin heavy and light 4. Davies J, Riechmann L (1995) Antibody VH chain repertoire. Nat Biotechnol 31:166–169 domains as small recognition units. Biotech- 17. van Dongen JJ, Langerak AW, Bruggemann M nology (N Y) 13:475–479 et al (2003) Design and standardization of 5. Ward ES, Gussow D, Griffiths AD et al (1989) PCR primers and protocols for detection of Binding activities of a repertoire of single clonal immunoglobulin and T-cell receptor immunoglobulin variable domains secreted gene recombinations in suspect lymphoproli- from Escherichia coli. Nature 341:544–546 ferations: report of the BIOMED-2 Concerted 6. Bradbury AR, Sidhu S, Dubel S et al (2011) Action BMH4-CT98-3936. Leukemia Beyond natural antibodies: the power of 17:2257–2317 in vitro display technologies. Nat Biotechnol 18. He L, Sok D, Azadnia P et al (2014) Toward a 29:245–254 more accurate view of human B-cell repertoire 7. Jung D, Giallourakis C, Mostoslavsky R et al by next-generation sequencing, unbiased rep- (2006) Mechanism and control of V(D)J ertoire capture and single-molecule barcoding. recombination at the immunoglobulin heavy Sci Rep 4:6778 chain locus. Annu Rev Immunol 24:541–570 19. Rohatgi S, Ganju P, Sehgal D (2008) System- 8. Di Noia JM, Neuberger MS (2007) Molecular atic design and testing of nested (RT-)PCR mechanisms of antibody somatic hypermuta- primers for specific amplification of mouse rear- tion. Annu Rev Biochem 76:1–22 ranged/expressed immunoglobulin variable 9. Glanville J, Kuo TC, von Budingen HC et al region genes from small number of B cells. J (2011) Naive antibody gene-segment frequen- Immunol Methods 339:205–219 cies are heritable and unaltered by chronic lym- 20. Baral TN, MacKenzie R, Arbabi Ghahroudi phocyte ablation. Proc Natl Acad Sci U S A M (2013) Single-domain antibodies and 108:20066–20071 their utility. Curr Protoc Immunol 103:Unit 10. Khan TA, Friedensohn S, Gorter de Vries AR 2.17 et al (2016) Accurate and predictive antibody 21. Andrews S (2010) FastQC: a quality control repertoire profiling by molecular amplification tool for high throughput sequence data. fingerprinting. Sci Adv 2:e1501371 http://www.bioinformatics.babraham.ac.uk/ 11. Briney BS, Willis JR, Crowe JE Jr (2012) Loca- projects/fastqc tion and length distribution of somatic 22. Magoc T, Salzberg SL (2011) FLASH: fast hypermutation-associated DNA insertions and length adjustment of short reads to improve deletions reveals regions of antibody structural genome assemblies. Bioinformatics plasticity. Genes Immun 13:523–529 27:2957–2963 12. Reddy ST, Ge X, Miklos AE, Hughes RA et al 23. Schmieder R, Edwards R (2011) Quality con- (2010) Monoclonal antibodies isolated with- trol and preprocessing of metagenomic data- out screening by analyzing the variable-gene sets. Bioinformatics 27:863–864 repertoire of plasma cells. Nat Biotechnol 24. Hussack G, Arbabi-Ghahroudi M, van Faassen 28:965–969 H et al (2011) Neutralization of Clostridium 446 Kevin A. Henry

difﬁcile toxin A with single-domain antibodies non-aggregating binders. Protein Eng Des Sel targeting the cell receptor binding domain. J 25:313–318 Biol Chem 286:8961–8976 26. Arbabi-Ghahroudi M, To R, Gaudette N et al 25. Hussack G, Keklikian A, Alsughayyir J et al (2009) Aggregation-resistant VHs selected by (2012) A VL single-domain antibody library in vitro evolution tend to have disulﬁde- shows a high-propensity to yield bonded loops and acidic isoelectric points. Pro- tein Eng Des Sel 22:59–66 Chapter 25

High-Throughput IgG Reformatting and Expression

Chao-Guang Chen, Georgina Sansome, Michael J. Wilson, and Con Panousis

Abstract

We have recently described a one-step zero-background IgG reformatting method that enables the rapid reformatting of phage-displayed antibody fragments into a single-mammalian cell expression vector for full IgG expression (Chen et al. Nucleic Acids Res 42:e26, 2014). The strategy utilizes our unique positive selection method, referred to as insert-tagged (InTag) positive selection, where a positive selection marker (e.g. chloramphenicol-resistance gene) is cloned together with the antibody inserts into the expression vector. The recombinant clones containing the InTag adaptor are then positively selected without cloning background, thus bypassing the need to plate out cultures and screen colonies. This IgG reformatting method is rapid and can be automated and performed in a high-throughput (HTP) format. The use of InTag positive selection with the Dyax Fab-on-phage antibody library is demonstrated. We have further optimized the protocol for IgG reformatting since the initial publication of this method (Chen et al. Nucleic Acids Res 42:e26, 2014) and also updated the transient transfection protocol using Expi293F cells, which are described herein.

Key words Phage display, High-throughput (HTP), IgG reformatting, InTag positive selection, In- Fusion cloning, Ligation-independent cloning, Mammalian cell expression, Transient transfection

1 Introduction

IgG reformatting and expression are key steps in antibody discovery using phage-display technology when the ultimate clinical or diagnostic application format is an intact IgG. In a typical antibody discovery workflow, positive phage clones are first isolated by screening a phage-display antibody library. These clones are then reformatted and expressed in E. coli as soluble Fabs or scFvs for biophysical and functional analysis to narrow down the number of lead candidates. The selected few clones are finally reformatted into intact IgGs and expressed in mammalian cells for detailed characterization, before a lead antibody is selected for further development. The E. coli expression step is a lengthy and laborious procedure that is often hindered by problems, including low and variable

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_25, © Springer Science+Business Media LLC 2018 447 448 Chao-Guang Chen et al.

expression of antibody fragments and complex purification protocols. In addition, the antibody fragments produced in E. coli are often not suited for screening in biological assays requiring IgG-dependent effector functions and antibody avidity. Further- more, lipopolysaccharide (LPS) contamination can compromise in vitro and in vivo functional assays employed during the candidate screening process. While ideally biophysical and functional screening of all unique antigen-binding phage clones should be under- taken in an IgG format, the lack of simple and high-throughput (HTP) IgG reformatting methods necessitates the use of an E. coli expression step to narrow down the number of candidate antibody fragments before IgG reformatting and mammalian expression. The HTP reformatting of antibody fragments from a phage- display vector to intact IgGs in a mammalian expression vector presents some significant challenges. First, two genes expressing the light and heavy chains need to be cloned to express each IgG. Furthermore, the cloning requires the perfect fusion of coding regions between the variable antibody regions from the phage- display vector and the IgG constant regions, as well as signal peptides in the mammalian expression vector. A number of IgG reformatting methods have been reported where the heavy and light chains are generated in separate vectors and the IgG expressed by co-transfection in mammalian cells [1, 2], or sequentially cloned into a single-mammalian expression vector [3, 4]. However, these methods suffer from several key limitations, such as multiple and laborious cloning steps; use of restriction digestion for preparation of antibody inserts from the phage-display vectors, which can result in loss of clones containing internal restriction sites; and most importantly a requirement for single-colony analysis for the identification of the correct recombinant clone from cloning background which results from uncut and re-ligated vector. Very recently, a novel system (pMINERVA) has been used to convert scFvs from a phage-display vector directly into IgGs using recombination without any in vitro subcloning steps [5]. However, this strategy requires a phage-display library to be generated in a specially modified vector and hence is not suitable for other libraries. In addition, the IgGs generated using this method will have additional amino acid residues inserted at several regions of the antibody due to the use of particular recombination sites. To overcome these problems, we have recently developed a one-step zero-background IgG reformatting method with insert- tagged (InTag) positive selection [6]. Figure 1 illustrates the key features of this method. First, the light chain (LC) and variable heavy (VH) regions from a phage clone are amplified by a single- duplex PCR. Second, both fragments are cloned together with an InTag adaptor into a single-expression vector using In-Fusion cloning (Clontech), a type of ligation-independent cloning. No restriction enzyme digestion of the inserts is required for High-Throughput IgG Reformatting and Expression 449

Fig. 1 Schematic representation of zero-background IgG reformatting using InTag positive selection. (A) Phage-display vector. The LC and VH regions are ampliﬁed by duplex-PCR with primers P1 and P2, and P3 and P4 respectively. (B) CmR InTag adaptor vector. The InTag Adaptor consisting of BGHpA, CmR, pCMV and a mammalian secretion signal (S) was generated by SOE-PCR [7] and cloned into TOPO vector, where the AmpR was further removed by BspHI digestion and religation. The CmR InTag adaptor is isolated by AscI and MfeI digestion to be used for InTag cloning. (C) Mammalian expression vector. The vector contains a CMV promoter, a mammalian secretion signal, and human IgG4 constant region. LC and VH inserts are cloned together with the InTag adaptor into ApaLI and NheI sites of pRhG4 expression vector by either In-Fusion or cut-paste method. The resulting DNA is transformed into E. coli, and recombinant plasmids are positively selected in liquid media containing chloramphenicol. LC light chain, VH variable heavy region, GIII gene III, pLac lac promoter, S signal peptide, I intron, rbs ribosome binding site, BGHpA bovine growth hormone polyadenylation signal, pCMV CMV promoter, CmR chloramphenicol resistance marker, AmpR ampicillin resistance marker, KanR kanamycin resistance marker

In-Fusion cloning, thus eliminating the potential for clone loss due to the presence of internal cut sites. In-Fusion also requires fewer manipulations than traditional cut-paste cloning methods and is amenable to automation. Finally, recombinant clones expressing the new selection marker present in the InTag adaptor are directly 450 Chao-Guang Chen et al.

selected in liquid cultures following transformation. InTag positive selection makes plating out the cultures for colony screening unnecessary, thus dramatically simplifying the cloning procedure. This IgG reformatting method is rapid, efﬁcient and amenable to automation for HTP applications. In this chapter, we describe updated protocols for HTP IgG reformatting and mammalian expression by transient transfection where 0.5 mg of puriﬁed IgG is obtained from 5 mL cultures, facilitating extensive biophysical and functional screening of all antibody candidates resulting from a phage library screening campaign.

2 Materials

2.1 General 1. 1% Agarose gel: Dissolve 1 g of agarose in 100 mL of 1Â TAE Reagents buffer.

2. 2YT media: To prepare 1 L measure ~900 mL of ddH2O and dissolve 16 g Bacto Tryptone, 10 g Bacto yeast Extract and 5 g NaCl. Then adjust pH to 7.0 with 5 M NaOH. Adjust to 1 L with ddH2O and sterilize by autoclaving.

3. 50% Glucose in ddH2O. 4. 96-well PCR plate. 5. 96-well flat-bottom culture plate. 6. AccuPrime pfx DNA polymerase (Thermo Fisher Scientific, Waltham, United States). 7. AirPore plate seal (QIAGEN, Hilden, Germany). 8. Ampicillin (Amp) filter-sterilized stock solution (100 mg/mL in sterile ddH2O). Working concentration is 100 μg/mL. 9. Chloramphenicol (Cm) stock solution (34 mg/mL in ethanol). Working concentration is 34 μg/mL. 10. CutSmart Buffer (New England Biolabs, Ipswich, United States). 11. Ethidium Bromide. Working concentration is 10 μg/mL. 12. Gateway vector conversion system (Thermo Fisher Scientific). 13. Luria-Bertani (LB) Broth: To prepare 1 L dissolve 10 g Bacto Tryptone, 5 g Yeast Extract and 10 g NaCl, adjust pH to 7.5 and bring to 1 L with ddH2O. Sterilize by autoclaving. 14. pcDNA3.1 vector (Thermo Fisher Scientific). 15. PCR plate seals. 16. pCR4Blunt-TOPO vector (Thermo Fisher Scientific). 17. pShuttle vector (Dyax, Burlington, United States). 18. QIAquick Gel Extraction Kit (QIAGEN). 19. QIAquick PCR Purification Kit (QIAGEN). High-Throughput IgG Reformatting and Expression 451

20. QIAprep Spin Miniprep Kit (QIAGEN). 21. Restriction enzymes: ApaLI, AscI, MfeI-HF, NheI-HF and BspHI (New England Biolabs). 22. Super optimal broth (SOC) recovery media: 2% Bacto Tryp- tone, 0.5% Yeast Extract, 10 mM NaCl, 2.5 mM KCl, 10 mM MgCl2, 10 mM MgSO4, 20 mM glucose. Sterilize by ﬁltration through a 0.2 μm ﬁlter.

23. Sterile ddH2O. 24. T4 DNA Ligase and Buffer (Promega, Madison, United States). 25. Tris-acetate-EDTA (TAE) buffer:To prepare 1 L of 50Â TAE dissolve the following components in 600 mL of ddH2O: 242 g Tris base (FW ¼ 121), 57.1 mL glacial acetic acid and 100 mL 0.5 M EDTA (pH 8.0). Adjust the ﬁnal volume to 1 L with ddH2O. Prior to use dilute to a 1Â working solution (40 mM Tris (pH 7.6), 20 mM acetic acid, 1 mM EDTA).

2.2 Preparation 1. Dyax pRhG4 vector (Fig. 1)(see Note 1). of Linearized Vector 2. CmR InTag adaptor vector (Fig. 1)(see Subheading 2.6 and and Adaptor Note 2). for Cloning 3. Restriction enzymes: ApaLI, AscI, MfeI-HF, NheI-HF.

2.3 PCR 1. Primers used for LC and VH ampliﬁcation from Dyax phage- Ampliﬁcation of LC display library are listed in Table 1 (see Notes 3 and 4). and VH Regions 2. Overnight phagemid-containing bacterial culture.

2.4 In-Fusion 1. In-Fusion HD Cloning System CE (Clontech, Mountain View, Cloning United States). 2. Puriﬁed vectors and adaptors (from Subheading 2.2). 3. Cloning Enhancer (Clontech). 4. Stellar chemically competent cells (Clontech). 5. 24-well round-bottom deep well plates (Whatman, Little Chal- font, United States).

2.5 Isolation 1. QIAcube (optional) (QIAGEN). of Plasmid DNA 2. Light chain sequencing primer: and Sequencing 0 0 SqCMVpro1F 5 CCACTGCTTACTGGCACGTG3 Analysis 3. Heavy chain sequencing primer: SqCMVpro2F 50 CCACTGCTTACTGGCTTATC30

2.6 Transient 1. Miniprep plasmid DNA. Transfection 2. Expi293F Cells (Thermo Fisher Scientiﬁc). 452 Chao-Guang Chen et al.

Table 1 Primers for LC and VH amplification and reformatting

No Primer name Sequence 50-30

P1a DykappaFwd CTCCACAGGCGTGCACTCAGACATCCAGATGACCCAG P1b DyDivmtgFwd CTCCACAGGCGTGCACTCAGACATCGTCATGACTCAG P1c DyQsaltqFwd CTCCACAGGCGTGCACTCACAGAGCGCTTTGACTCAG P1d DyQsvltqFwd CTCCACAGGCGTGCACTCACAGAGCGTCTTGACTCAG P1e DyQseltqFwd CTCCACAGGCGTGCACTCACAGAGCGAATTGACTCAG P1f DyQyeltqFwd CTCCACAGGCGTGCACTCACAGTACGAATTGACTCAG P1g DySyeltqFwd CTCCACAGGCGTGCACTCAAGCTACGAATTGACTCAG P2a DyLCrevA CACAGTCGAGGCGCGCATTATGACTGTCTCCTTG P2b DyLCrevB CACAGTCGAGGCGCGCTCATTAAGACTGTTCCTTG P3 DyVHfwd CTCCGAAGTTCAATTGTTAGAGTCTGGTGGTGGCCT P4 DyVHrev GGAGCAGGGCGCTAGCGGGAAGACCGATG The primers were adapted from Jostock et al. 2004 [4] and their relative positions in the phage-display vector are illustrated in Fig. 1. All primers contain a 15 bp extension (underlined) required for In-Fusion cloning (see Note 3). P1 and P2 primers are used to amplify the LC and P3 and P4 primers are used to amplify the VH regions from the Fab phage clones. Two kappa P1 primers and 5 lambda P1 primers are available, from which the correct P1 primer will be selected for PCR based on the nucleotide sequence of the particular phage clone. Two P2 primers are synthesized to accommodate the sequence variation in the primer region. Both primers will be used in the PCR

3. ExpiFectamine 293 Transfection Kit. Contents include Expi- Fectamine 293 reagent; ExpiFectamine 293 Transfection Enhancer 1 and ExpiFectamine 293 Transfection Enhancer 2 (Thermo Fischer Scientific). 4. Expi293 Expression Medium (Thermo Fisher Scientific). 5. Gibco Antibiotic-Antimycotic (100Â) (Thermo Fisher Scientific). 6. Opti-MEM I Reduced-Serum Medium (Thermo Fisher Scientific).

7. Filter-sterilized LucraTone Lupin-20% in ddH2O (Merck Millipore, Billerica, United States). 8. Hemocytometer with trypan blue or cell counter. 9. 50 mL TubeSpin Bioreactor tubes.

2.7 Generation 1. We use SOE-PCR [7] to generate the CmR InTag adaptor of CmR InTag Adaptor (Fig. 1) consisting of BGH polyA signal (BGHpA), chloram- (See Note 2) phenicol resistance gene (CmR), CMV promoter (pCMV), and a mammalian secretion signal (S) using primers listed in Table 2. High-Throughput IgG Reformatting and Expression 453

Table 2 Primer sequences for CmR InTag adaptor generation

ID Name Sequence 50-30

A1 BGHpAfwd AAAGGCGCGCCTCGACTGTGCCTTCTAG A2 BGH_CmRrev TTTTACGTTTCTCGTTCAGCCATAGAGCCCACCGCATC A3 BGH_CmRfwd GATGCGGTGGGCTCTATGGCTGAACGAGAAACGTAAAA A4 CmR_CMVrev GTCAATAATCAATGTCAACTTACGCCCCGCCCTGCCA A5 CmR_CMVfwd TGGCAGGGCGGGGCGTAAGTTGACATTGATTATTGAC A6 CMV_SPrev CAGCTCCATCCCATGGTGGCGGCCCTATAGTGAGTCGTA A7 CMV_SPfwd TACGACTCACTATAGGGCCGCCACCATGGGATGGAGCTG A8 VHspRev TCACCGCTAGCGGCCGCAGA A2 and A3, A4 and A5, A6 and A7 primers are complementary to each other respectively to allow SOE-PCR

2. Amplify the BGHpA fragment from pcDNA3.1 by polymerase chain reaction (PCR) using A1 and A2 primers. 3. Amplify the CmR gene from the Gateway vector conversion system using A3 and A4 primers. 4. Amplify the CMV promoter region from pcDNA3.1 using A5 and A6 primers. 5. Amplify the signal peptide from the Dyax pShuttle vector [4] using primers A7 and A8. 6. Perform the PCR in a 50 μL reaction containing 10 ng of template DNA, 1Â AccuPrime Reaction Mix, 0.5 μM of each primer and 1 U of AccuPrime pfx DNA polymerase. PCR cycling conditions are as follows: 94 C for 3 min; 5 cycles of 93 C for 30 s, 50 C for 30 s and 68 C for 1 min; 25 cycles of 93 C for 20 s; 60 C for 30 s and 68 C for 1 min; and 68 C for 10 min. 7. Join the four DNA fragments by SOE-PCR [7] in a step-wise manner. First, join PCR products from steps 2 and 3 using primers A1 and A4, and steps 4 and 5 using primers A5 and A8, respectively. The two resultant fragments are then combined to form the ﬁnal adaptor using primers A1 and A8. Perform the PCRs as in step 6 using 2 μL each of the relevant PCR products as DNA templates. 8. Subclone the fragment into pCR4Blunt-TOPO vector according to the manufacturer’s instructions. 9. Digest the vector with BspHI and re-ligate to remove the AmpR gene, thus making this plasmid non-resistant to ampicillin. 454 Chao-Guang Chen et al.

3 Methods

3.1 Preparation 1. Set up a restriction enzyme digestion of pRhG4 vector by of Linearized Vector combining 5 μg of DNA with 50 units ApaL1 and 50 units and Adaptor Nhe1-HF, 20 μL of CutSmart buffer and ddH2O to a ﬁnal volume of 200 μL. Incubate at 37 ˚C for 3 h. 2. Run the digested vector on a 1% agarose/TAE gel containing 10 μg/mL ethidium bromide along with a DNA molecular weight marker. Excise the vector band and purify with QIA- quick gel extraction kit. Elute in 30 μL of Buffer EB and quantitate DNA using a Trinean Xpose or similar spectrophotometer. 3. To prepare the CmR InTag adaptor fragment, digest 5 μgof the adaptor plasmid with 50 units AscI and 50 units MfeI-HF, 20 μL of CutSmart buffer in a ﬁnal volume of 200 μL. Incubate at 37 ˚C for 3 h (see Note 5). 4. Run the digested Adaptor vector on a 1% agarose/TAE gel containing 10 μg/mL ethidium bromide in parallel with a DNA molecular weight marker. Excise the 1.8 kb fragment of the InTag adaptor and purify with QIAquick gel extraction kit. Elute in 30 μL and quantitate DNA using a Trinean Xpose or similar spectrophotometer.

3.2 PCR 1. Inoculate 120 μL of 2YT media containing 2% glucose and Amplification 100 μg/mL ampicillin in a 96-well flat-bottom culture plate of Antibody Light with 2 μL of bacterial glycerol stock containing phagemid. Chain and Variable 2. Seal with an AirPore Tape sheet and grow overnight at 37 C Heavy Chain shaking 255 rpm in an Infors Microtron Shaking Incubator Fragments (3 mm pitch). 3. Prepare PCR template by diluting 2 μL of the overnight culture with 198 μL of ddH2O in a 96-well plate. 4. Choose the correct P1 primer from the seven P1 primers available based on the phage LC sequence and mix with P2a, P2b, P3, and P4 primers in an equimolar ratio (4 μM each). 5. Prepare master stocks in multiples of 10 μL containing 2 μLof 10Â AccuPrime pfx Reaction Mix, 1 μL of the primer mix and 0.2 μL of AccuPrime pfx DNA polymerase. 6. Aliquot 10 μL of the mastermix into appropriate wells of a 96-well PCR plate. 7. Add 10 μL of diluted bacterial culture containing phagemid to corresponding wells (final volume ¼ 20 μL). 8. Seal the plate with a PCR plate seal. 9. Pulse-spin the plate to ensure that contents are all at the bottom of the well. High-Throughput IgG Reformatting and Expression 455

10. Transfer the plate to a PCR thermal cycler and start the following PCR program (see Note 6): 1 cycle of 94 C for 3 min; 5 cycles of 94 C for 30 s, 50 C for 30 s and 68 C for 1 min; 25 cycles of 94 C for 20 s; 60 C for 30 s and 68 C for 1 min; and 1 cycle of 68 C for 10 min. 11. Load 5 μL of PCR product on a 1% agarose/TAE gel containing 10 μg/mL ethidium bromide alongside a DNA molecular weight marker. Each successful PCR reaction should result in two products corresponding to the LC (~850 bp) and the VH (~450 bp) (see Note 7). 12. Proceed with cloning or store PCR products at À20 C.

3.3 In-Fusion 1. Take a fresh aliquot of Cloning Enhancer (see Note 9) from Cloning (See Note 8) À20 C and place immediately on ice. 2. Add 2 μL of cloning enhancer into the appropriate number of wells of a 96-well PCR plate. This is best done in a 96-well cooling block. 3. Transfer 5 μL of PCR products to the plate containing the Cloning Enhancer (ﬁnal volume ¼ 7 μL). 4. Cover with a suitable PCR plate seal. 5. Incubate in a PCR thermal cycler at 37 C for 15 min, then 80 C for a further 15 min. 6. Transfer the plate to ice or if not proceeding with cloning, store Cloning Enhancer-treated PCR products at À20 C. 7. For the In-Fusion cloning prepare a master mix in multiples of 9 μL containing 60 ng linearized pRhG4 vector, 50 ng CmR InTag adaptor, and 2 μL5Â In-Fusion HD enzyme Premix (see Note 10). 8. Dispense 9 μL per well of In-Fusion cloning mix into a 96-well PCR plate, then transfer 1 μL of Cloning Enhancer-treated PCR product to each well (ﬁnal volume ¼ 10 μL). 9. Incubate in a PCR thermal cycler at 50 C for 15 min. 10. Transfer the plate immediately to ice or store at À20 C if not proceeding with transformation.

3.4 Transformation 1. Thaw Stellar competent cells on ice just prior to use. 2. Aliquot 20 μL cells into a 96-well PCR plate on ice or a cooling block. 3. Add 2 μL In-Fusion reaction mix to 20 μL cells (see Note 11). 4. Heat-shock the cells for 30 s at 42 C using a thermal cycler. 5. Return plate to ice or cooling block for 2 min. 6. Add 80 μL pre-warmed SOC to each well. 456 Chao-Guang Chen et al.

7. Cover with an AirPore Tape sheet and incubate at 37 C for 1 h without shaking. 8. Transfer 100 μL transformation mix to 5 mL of LB containing 34 μg/mL Cm in a 24-well round bottom deep well plate (see Note 12). 9. Cover with an AirPore Tape sheet and incubate at 37 C shaking (220 rpm) for approximately 40 h (see Note 12).

3.5 IgG Reformatting 1. Scale up the PCR reaction described in Subheading 3.2 to Using Cut-and-Paste 50 μL scale. Method 2. Analyze 5 μL of the PCR products on 1% agarose/TAE gel as described in Subheading 3.2. 3. Purify the PCR products using QIAquick PCR Purification Kit and elute the DNA in 50 μL of Buffer EB. 4. Digest PCR products with ApaLI, AscI, MfeI-HF and NheI- HF. 5. Purify the DNA again using QIAquick PCR Purification Kit and elute the DNA in 30 μL of Buffer EB. 6. Set up a ligation reaction in a final volume of 10 μL containing 60 ng linearized pRhG4 vector, 50 ng CmR InTag adaptor, 50–100 ng of purified PCR products, 1 μL10Â Ligase Buffer, and 1 μL T4 DNA Ligase (1 unit). 7. Incubate the reaction at room temperature for 1–3 h or at 15 C for 4–18 h. 8. For transformation, add 3–5 μL ligation mix to 50 μLof chemically competent cells and incubate on ice for 15 min. 9. Heat-shock the cells for 60 s at 42 C and return the tube to ice for 2 min. 10. Add 500 μL pre-warmed SOC to each transformation and incubate at 37 C with shaking (220 rpm) for 1 h. 11. Transfer 500 μL to 4.5 mL of LB containing 34 μg/mL chloramphenicol in 24-well deep well plates. 12. Seal with an AirPore Tape sheet and incubate at 37 C shaking (220 rpm) for ~40 h.

3.6 Isolation 1. Transfer 2 mL of 2-day culture into 2 mL Eppendorf tubes and of Plasmid DNA pellet bacterial cells by centrifugation at 6800 Â g for 3 min at and Sequence room temperature. Conﬁrmation 2. Remove the supernatant and proceed with isolation of DNA using QIAprep Spin Miniprep Kit either manually or using QIAGEN QIAcube according to the manufacturer’s instructions. DNA is eluted in 100 μL Buffer EB. High-Throughput IgG Reformatting and Expression 457

3. DNA is sequenced with SqCMVpro1F and SqCMVpro2F primers using the ABI BigDye Terminator Cycle Sequencing Kit. 4. Reformatted IgG sequences are aligned and compared to the original phagemid sequences (see Note 13).

3.7 Transient 1. Culture Expi293F cells in Expi293F Expression media supple- Transfection mented with 10 mL/L Gibco Antibiotic-Antimycotic solution by seeding at 0.3–0.5 Â 106 cells/mL. Cells should not be allowed to reach densities above 5 Â 106 cells/mL. 2. On the day of transfection use a hemocytometer and trypan blue exclusion (or other methods of cell counting) to determine the cell count and viability. Only transfect if viability is greater than 95% and cell count between 3 and 5 Â 106 cells/ mL. 3. Pellet the required number of cells (1.25 Â 107 per 5 mL transfection) and resuspend with Expi293 Expression Medium to a final concentration of ~3 Â 106 cells/mL. 4. For each 5 mL transfection transfer 4.1 mL of the cells to each TubeSpin Bioreactor tube and shake at 250 rpm, 37 C, 8% CO2. 5. Aliquot 5 μg of plasmid DNA into a sterile 24-well plate and dilute with 0.25 mL of Opti-MEM I Reduced Serum Medium. For multiple transfections a multidispenser pipette can be used to add reagents. 6. In a separate tube dilute 13.5 μL ExpiFectamine in 0.25 mL of Opti-MEM per 5 mL transfection and incubate at room temperature for 5 min. 7. Add the diluted ExpiFectamine to the DNA to give a final volume of 0.5 mL per transfection and incubate at room temperature for 20 min to allow DNA and ExpiFectamine 293 Reagent complexes to form. 8. Add the 0.5 mL of complex to each tube containing 1.25 Â 107 cells to give a final volume of 4.6 mL. Incubate the cells at 37 C8%CO2, 250 rpm for 16–20 h. 9. Add 400 μL master mix containing 25 μL of Enhancer 1, 250 μL Enhancer 2 and 125 μL LucraTone Lupin to each TubeSpin Bioreactor tube. 10. Harvest the supernatants 5 days after transfection by centrifugation at 3000 Â g, for 20 min. 11. Filter the supernatant using a 0.22 μm polyethersulfone Steriflip-GP filter unit prior to analysis and purification as described by Schmidt et al. [8]. 458 Chao-Guang Chen et al.

4 Notes

1. The pRhG4 vector used in this study is from Dyax Corp, which was derived from the pCMV/myc/ER vector (Thermo Fisher Scientific) by incorporating a mammalian secretion signal and the human heavy chain constant region (CH1-hinge-CH2- CH3) of IgG4 as described previously [4]. Other mammalian expression vectors such as pcDNA3.1 can be readily adapted for use with IgG reformatting using InTag positive selection via the insertion of a heavy chain constant region (species and isotype of choice) and a mammalian secretion signal. 2. The InTag adaptor serves two key roles in this one-step zero- background IgG reformatting strategy. First, it provides the necessary regulatory elements (a polyadenylation site for the light chain, the CMV promoter and a mammalian signal peptide for the heavy chain) for expression in mammalian cells. Second, it provides a new antibiotic resistance marker (e.g., CmR) to facilitate the positive selection of recombinant clones. The IgG expression constructs we made contain two BGHpA sites and two CMV promoters. However, we find no instability issues with these constructs despite the presence of these duplications. 3. The primers described here are tailored for the Dyax “Fab-on- phage” antibody library and therefore need to be modified accordingly when different phage-display library and/or expression vectors are used. For the P1 primers that anneal to the N-terminal sequence of the light chains, we have designed two P1 primers for kappa and five P1 primers for the lambda light chains, respectively, in order to amplify all the light chain sequences in the Dyax library. A particular P1 primer needs to be selected based on the N-terminal light chain sequence of the phage clone. Two P2 reverse primers were designed to accommodate the sequence variation immediately downstream of the light chain coding region. For simplicity, both P2 primers are included in the PCR reactions for all LC amplifications. For the P3 primer that binds to the N-terminal VH sequence, only one primer is required, as the Dyax library we use is a semi-synthetic library based on a single heavy chain germline sequence (3–23). This primer is used in conjunction with a single-reverse primer (P4) which anneals to the start of the heavy chain constant region. All the primers have a 15 bp extension (underlined) at the 50 end, which is required for In-Fusion cloning. Please refer to the In-Fusion cloning manual for the appropriate design of this 15 bp extension region. In our case, the 15 bp extension regions also contain appropriate restriction sites to allow High-Throughput IgG Reformatting and Expression 459

cloning of the fragments using the cut-and-paste cloning method as a backup strategy. The 30 half of the primers are target specific and normal principles for PCR primer design should be followed to ensure they are specific and work well under a duplex-PCR condition, so that both LC and VH fragment can be efficiently amplified. We often use 17–20 bases for these target-specific regions in order to keep the overall length of primers as short as possible. 4. We recommend using PAGE-purified primers for the best quality possible as current oligonucleotide synthesis technologies always produce by-products that are either prematurely terminated, or contain internal deletions in the sequence. These by-products, especially the ones with internal deletions, could contaminate the final IgG construct DNA as our HTP IgG reformatting strategy relies on bulk cloning without single-colony isolation. As we see batch-to-batch variation in primer quality, obtaining of primers from the same or different manufacturers is sometimes necessary. We also recommend the primers to be dissolved and stored in Tris-EDTA (TE) buffer, pH 8.0 at À20 C for long-term usage. 5. Since the CmR InTag adaptor contains the Cm resistance marker used for the final recombinant clone selection, there is a potential for the adaptor plasmid to contaminate the ligation solution. Therefore, care needs to be taken to ensure that the plasmid digestion and gel extraction procedures are performed well so that there is no carryover of uncut plasmid prior to cloning. Quality control can be performed by transforming 12 ng of purified adaptor and culturing in LB/Cm to ensure there is no growth. Although we rarely observe InTag adaptor contamination, if it does happen, this contamination can be eliminated by selecting the cells with ampicillin since the CmR InTag adaptor plasmid we use lacks the AmpR gene and is therefore not resistant to ampicillin. 6. The Duplex PCR conditions used here were optimized for our primers and need to be modified accordingly when different primers are used. In our case, some of the primers have a short (17 bp) overlapping region to the targets and one primer even has two mismatched bases intentionally introduced to avoid off-targeting. Therefore, a relatively low annealing temperature (50 C) is initially used for the first 5 cycles of the PCR to allow these primers to anneal to the templates. The annealing temperature is subsequently increased to 60 C for another 25 cycles. 7. Agarose gel analysis of PCR products as a quality control step is optional but recommended, especially when the number of samples is not too large. On the agarose gel, apart from the 460 Chao-Guang Chen et al.

intended LC and VH bands amplified by Primers 1 and 2, and Primers 3 and 4 respectively, we can sometimes observe a larger PCR product representing the full-length Fab fragment amplified by Primer 1 and Primer 4 (Fig. 1). Although this Fab fragment can be cloned into the expression vector by itself, the recombinant clone containing this Fab fragment will be eliminated by InTag positive selection with chloramphenicol due to the absence of InTag adaptor in such clones. 8. This protocol can be adapted for automation using a liquid handler (e.g., Eppendorf epMotion). For accuracy, we recommend diluting the Cloning Enhancer-treated PCR product with ddH2O to a final volume of 24 μL and add 4 μLtoa 6 μL In-Fusion cloning master mix. 9. Extra care needs to be taken for the handling and storage of the In-Fusion HD Cloning Kit in order to avoid a decrease in cloning efficiency. It is recommended to keep small aliquots to minimize freezing and thawing, to which the Cloning Enhancer is particularly sensitive. PCR clean-up can be used as an alternative method to the Enhancer treatment. 10. We have tried various ratios of inserts, InTag adaptor and vector and found that the cloning method can tolerate wide variation in ratios. Therefore, quantitation and normalization of the PCR products is not required prior to cloning, thus facilitating the HTP operation. 11. It is important that the volume of In-Fusion reaction mix added to the competent cells does not exceed 10% of the volume of competent cells used to minimize transformation inhibition. We usually use the Stellar chemically competent cells provided with the In-Fusion HD Cloning Kit. We have also successfully used a range of other chemical competent cells such as TOP10 (Thermo Fisher Scientific) and Alpha-Select Gold Efficiency (BIOLINE). We strongly recommend the use of competent cells with a transformation efficiency 1 Â 108 cfu/μg, and naturally the competent cells need to be sensitive to ampicillin and chloramphenicol, which are used for clone selection. If low cloning efficiency is observed with certain bacterial strains, you may obtain better results by diluting the reaction mix with TE buffer 5–10 times prior to transformation. 12. For liquid cultures following transformation we use a 24-well deep well plate with a final volume of 5 mL/well. Cultures can just as easily be grown in tubes; however, the use of a 24-well deep well plate enables application of either automation or adjustable spacing pipettes for sample transfer. We usually grow the cultures for approximately 40 h to make sure we High-Throughput IgG Reformatting and Expression 461

have consistent DNA yields as overnight growth is often inadequate. 13. Due to the nature of bulk cloning without single-colony isolation, we sometimes observe double or multiple sequencing signals from the same DNA sample. One possible cause of this is that the starting phage material is not clonal. Another cause is that the primer is not 100% pure and contains a certain percentage of by-products with internal deletion(s) (also see Note 4). One solution to these problems is to plate the culture out and pick single colonies for subsequent sequencing analysis to isolate the correct clone.

References

1. Sarantopoulos S, Kao CY, Den W, Sharon J 5. Batonick M, Kiss MM, Fuller EP, Magadan CM, (1994) A method for linking VL and VH region Holland EG, Zhao Q, Wang D, Kay BK, Weiner genes that allows bulk transfer between vectors MP (2016) pMINERVA: a donor–acceptor sys- for use in generating polyclonal IgG libraries. J tem for the in vivo recombineering of scFv into Immunol 152:5344–5351 IgG molecules. J Immunol Methods 431:22–30 2. Jones ML, Seldon T, Smede M, Linville A, Chin 6. Chen CG, Fabri LJ, Wilson MJ, Panousis C DY, Barnard R, Mahler SM, Munster D, Hart D, (2014) One-step zero-background IgG refor- Gray PP, Munro TP (2010) A method for rapid, matting of phage-displayed antibody fragments ligation-independent reformatting of recombi- enabling rapid and high-throughput lead identi- nant monoclonal antibodies. J Immunol Meth- fication. Nucleic Acids Res 42(4):e26. https:// ods 354:85–90 doi.org/10.1093/nar/gkt1142. Epub 2013 3. Sanna PP, Samson ME, Moon JS, Nov 18 Rozenshteyn R, De Loqu A, Williamson RA, 7. Horton RM, Hunt HD, Ho SN, Pullen JK, Burton DR (1999) pFab-CMV, a single vector Pease LR (1989) Engineering hybrid genes system for the rapid conversion of recombinant without the use of restriction enzymes: gene Fabs into whole IgG1 antibodies. Immunotech- splicing by overlap extension. Gene 77:61–68 nology 4:185–188 8. Schimdt PM, Abdo M, Butcher RE, Yap MY, 4. Jostock T, Vanhove M, Brepoels E, Gool RV, Scotney PD, Ramunno ML, Martin-Roussety G, Daukandt M, Wehnert A, Hegelsom RV, Owczarek C, Hardy MP, Chen CG, Fabri LJ Dransfield D, Sexton D, Devlin M, Ley A, (2016) A robust robotic high-throughput anti- Hoogenboom H, Mullberg€ J (2004) Rapid gen- body purification platform. J Chromatogr A eration of functional IgG antibodies derived 455:9–19. https://doi.org/10.1016/j.chroma. from Fab-on-phage display libraries. J Immunol 2016.05.076 Methods 289:65–80 Chapter 26

Monitoring Phage Biopanning by Next-Generation Sequencing

Anna Vaisman-Mentesh and Yariv Wine

Abstract

Phage display has enabled the rapid isolation of antigen-speciﬁc antibodies from combinatorial libraries of the variable heavy chain gene (VH) and variable light chain gene (VL). The method is based on genetic engineering of bacteriophages and repeated rounds of antigen-guided selection by phage biopanning. Next-Generation Sequencing (NGS) coupled with bioinformatics are powerful tools for analyzing the large number of DNA sequences present in an immune library. Here, we describe a method that demonstrates how NGS analysis enhances phage biopanning of complex antibody libraries as well as facilitates the antibody discovery process.

Key words Antibody repertoire, B-Cell receptor repertoire, Immunoglobulin repertoire, VDJ, Next- Generation Sequencing, NGS, High-throughput sequencing, HTS, scFv, V gene, Variable region

1 Introduction

Phage display has enabled the rapid isolation of antigen-specific antibodies from combinatorial libraries of VH and VL genes obtained from lymphocytes and it is currently the most widespread method for the display and selection of large collections of antibodies [1]. It is based on libraries of phage particles expressing a great variety of exogenous polypeptides fused to phage surface proteins. The highly diverse library is reduced to a few leads by performing several rounds of selection, also known as biopanning [2]. Despite it being a powerful platform for antibody discovery, each step of phage display is prone to biases. Noteworthy, the antibodies are screened usually in their scFv configuration. Biases introduced during phage panning derive from library preparation as a result of PCR-based bias in amplification of different VH and VL templates; the subsequent PCR amplifications used for the assembly of scFvs and due to the fact that actual library size falls short of covering all possible combinations of VH and VL

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_26, © Springer Science+Business Media LLC 2018 463 464 Anna Vaisman-Mentesh and Yariv Wine

domains present in the cDNA [3, 4]. Second, some scFvs display better on phage than others [5–7]. Third, the affinity selection process itself is very sensitive to varying conditions and experimental variations [8, 9]. Recently, it was demonstrated that antigen-specific antibodies can be readily identified without screening, by exploiting Next- Generation Sequencing (NGS) of the V gene repertoire encoded by antibody secreting cells [10–14]. It was also shown that phage display results in the isolation of antibodies which are barely detected in the NGS reads from the same library, and conversely, mining of the V gene repertoire results in the identification of antigen-specific antibodies that had been missed by phage display. Moreover, clones identified by NGS exhibit higher selectivity to the antigen [7]. The integration of phage panning with NGS allows an in-depth evaluation of the enrichment landscapes during the panning process. By monitoring the diversity of the VH and VL from the scFv library following each panning cycle, it is possible to evaluate the enrichment progression and even recover antigen-specific antibodies that had been missed by phage display. Here, we describe a method for monitoring the progression of the enrichment of scFv during the phage biopanning pipeline, by harnessing the ability of NGS to analyze the diversity and abundance of the scFv VH and VL regions (Fig. 1). The method described herein focuses on the library preparation of V gene libraries from human B-cell subsets. The method of library preparation is optimized for deep sequencing in the MiSeq platform (Illumina, 2x300bp). The method can be readily trans- lated to generate MiSeq libraries based on VH and VL (cloned as scFv in a phagmid) as a template following each cycle of phage biopanning (Fig. 2).

Fig. 1 A schematic workﬂow for the preparation of V gene libraries. B cells from peripheral blood or other tissues are sorted and subjected to high-throughput immunoglobulin V-gene sequencing, resulting in the generation of a personal antibody sequence database Phage Biopanning and NGS 465

Fig. 2 Overview of the procedure used for adapter addition to antibody variable heavy chain amplicon libraries. The method requires a reverse transcription of antibody mRNA into cDNA which serves as template for the IgG gene-specific amplification by PCR. The primer extension method incorporates a GC-rich overhang into the library in the first PCR (PCR1), which results in uniformly high amplification in the second PCR (PCR2) using primers specific for the GC-rich overhang and containing the full-length Illumina adapters

2 Materials

Prepare all solutions using ddH2O (prepared by purifying deionized water, to attain a sensitivity of 18 MΩ-cm at 25 C) and HPLC analytical grade reagents. Prepare and store all reagents at room temperature (unless indicated otherwise). Diligently follow all waste disposal regulations when disposing waste materials.

2.1 Whole Blood 1. K2-EDTA tubes for blood collection (Thermo Fisher Scien- Separation and B-Cells tiﬁc, Waltham, USA). Isolation 2. Dulbecco’s PBS, no Calcium/Magnesium (DPBS) (Thermo Fisher Scientiﬁc). 3. Uni-SepMAXIþ tube (Novamed, Chicago, United States). 4. Sterile glass pipette. 466 Anna Vaisman-Mentesh and Yariv Wine

2.2 Sorting B-Cell 1. Antibody conjugated with fluorophores in accordance with Subsets B-cell markers. 2. FACS buffer (BioLegend, San Diego, Untied States). 3. Falcon™ Cell Strainers 40 μm (Thermo Fisher Scientific). ® 4. TRIzol LS Reagent (Thermo Fisher Scientific).

® 2.3 Total mRNA 1. RNaseZap RNase Decontamination Solution (Thermo Fisher Extraction and cDNA Scientiﬁc). Synthesis 2. Chloroform. 3. 70% Ethanol. 4. RNeasy micro Kit (Qiagen, Hilden, Germany). 5. mRNA template. 6. SuperScript™ II Reverse Transcriptase (Thermo Fisher Scientiﬁc).

2.4 VH and VL Library 1. Forward and reverse primers (see Fig. 3 and Table 1). Preparation 2. dNTP mix of 2.5 mM each. (Amplicon) 3. FastStart™ High Fidelity PCR System (Roche, Basel, Switzerland). 4. cDNA template. 5. Agencourt AMPure XP (Beckman Coulter, Brea, United States). ® 6. SeaKem LE Agarose (Lonza, Basel, Switzerland). 7. SYBR Safe DNA Gel Stain (Thermo Fisher Scientiﬁc).

Fig. 3 Human primer set used for amplifying a VH and VL amplicons. The colors representing different regions in the primers correspond to the colors in the sequences in primer Table 1 in sections B and C. All reverse primers should be reverse complement prior to synthesis Phage Biopanning and NGS 467

Table 1 Primer list for PCR 1 (a) and PCR 2 (b). The nucleotide sequences in the primers correspond to different regions as indicated in Fig. 3

Sequence Primers PCR CCCTCCTTTAATTCCCCAGGTCCAGCTKGTRCAGTCTGG CCCTCCTTTAATTCCCCAGGTGCAGCTGGTGSARTCTGG CCCTCCTTTAATTCCCTCAACACAACGGTTCCCAGTTA CCCTCCTTTAATTCCCCAGRTCACCTTGAAGGAGTCTG VH FW mix CCCTCCTTTAATTCCCGAGGTGCAGCTGKTGGAGWCY CCCTCCTTTAATTCCCCAGGTACAGCTGCAGCAGTCA CCCTCCTTTAATTCCCCAGGTGCAGCTGCAGGAGTCS CCCTCCTTTAATTCCCCAGGTGCAGCTACAGCAGTGGG CCCTCCTTTAATTCCCGACATCCRGDTGACCCAGTCTCC CCCTCCTTTAATTCCCGATATTGTGMTGACBCAGWCTCC VK FW mix CCCTCCTTTAATTCCCGAAATTGTRWTGACRCAGTCTCC CCCTCCTTTAATTCCCGAAACGACACTCACGCAGTCTC GAGGAGAGAGAGAGAGAGATGGTGCAGCCACAGTTC VK REV PCR 1 CCCTCCTTTAATTCCCCAGTCTGTSBTGACGCAGCCGCC CCCTCCTTTAATTCCCCAGCCTGTGCTGACTCARYC CCCTCCTTTAATTCCCCAGCCWGKGCTGACTCAGCCMCC CCCTCCTTTAATTCCCCAGTCTGYYCTGAYTCAGCCT CCCTCCTTTAATTCCCTCCTATGWGCTGACWCAGCCAA VL FW mix CCCTCCTTTAATTCCCTCCTCTGAGCTGASTCAGGASCC CCCTCCTTTAATTCCCTCCTATGAGCTGAYRCAGCYACC CCCTCCTTTAATTCCCAATTTTATGCTGACTCAGCCCC CCCTCCTTTAATTCCCCAGDCTGTGGTGACYCAGGAGCC GAGGAGAGAGAGAGAGTAGGACGGTSASCTTGGTCC VL REV mix GAGGAGAGAGAGAGAGGAGGACGGTCAGCTGGGTGC GAGGAGAGAGAGAGAGSGATGGGCCCTTGGTGGARGC VH REV

Sequence Primers PCR CAAGCAGAAGACGGCATACGAGATCGTGATGTGACTGGAGTTCAGA REV, index adapter CGTGTGCTCTTCCGATCTNNNNGAGGAGAGAGAGAGAG PCR 2 AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCT FW, universal adapter CTTCCGATCTNNNNCCCTCCTTTAATTCCC

8. Zymoclean™ Gel DNA Recovery Kit (Zymo Research, Irvine, United States). 9. Zymoclean™ DNA Clean & Concentrator™ (Zymo Research). ® 10. Qubit dsDNA HS Assay Kit (Thermo Fisher Scientiﬁc) Agi- lent DNA 1000 Kit (Agilent Technologies, Santa Clara, United States). 468 Anna Vaisman-Mentesh and Yariv Wine

2.5 Illumina 1. MiSeq Reagent Kit v3 (Illumina, San Diego, United States). Sequencing, Data 2. MiXCR - universal software for fast and accurate analysis of T- Analysis, and B-cell receptor repertoire sequencing data. Available at: and Statistics https://milaboratory.com/software/mixcr/

3 Methods

Carry out all the procedures at room temperature unless otherwise speciﬁed.

3.1 Whole Blood 1. Collect whole blood in K2-EDTA tubes (see Note 1). Separation and B-Cells 2. Carefully layer the blood into the Uni-SepMAXIþ tube and Isolation centrifuge at 1200 Â g for 10 min with the brake on. 3. Transfer the middle layer (buffy coat) comprising the peripheral blood mononuclear cells (PBMC) to a new 50 mL canonical tube using a sterile glass pipette and centrifuge at 180 rpm for 10 min at room temperature with the brake off. 4. Discard the supernatant and wash the pellet with 15 mL of DPBS. Resuspend gently pellet in a ﬁnal volume of 2 mL DPBS. 5. Filter the cells into a new 50 mL canonical tube using a sterile 40 μm ﬁlter.

3.2 Sorting B-Cell 1. Sort B-cell populations using any Fluorescence-activated cell Subsets sorting instrument according to the manufacturer’s protocols ® into TRIzol LS Reagent (see Note 1).

3.3 Total mRNA 1. Isolate total RNA according to the RNeasy micro Kit manu- Extraction and cDNA facturer’s protocol. Synthesis 2. Generate cDNA with 500 ng of isolated total RNA as template and Oligo(dT) primer using a SuperScript RT II kit following manufacturer’s protocol.

3.4 VH and VL Library 1. Prepare the following equimolar forward and reverse primer Preparation mixes to create a stock in the concentration of 10 μM of each (Amplicon) primer (Fig. 3 and Table 1):

PCR 1 10 μM forward primer mix for VH library PCR 1 10 μM reverse IgG primer

PCR 1 10 μM reverse primer mix for VK library

PCR 1 10 μM reverse primer mix for VL library (see Note 2) Phage Biopanning and NGS 469

2. Prepare the following reaction mix:

Forward primer mix 7.5 μL (final concentration of 0.25 μM) Reverse primer mix 7.5 μL (final concentration of 0.25 μM) dNTPs 6 μL (final concentration of 200 μM for each) Polymerase buffer X5 60 μL FastStart™ high Fidelity 3 μL ddH2O 211 μL

3. Remove 50 μLto1Â 0.2 mL PCR tube to serve as a negative control. 4. Add 5 μL of cDNA to the reaction mix and pipette (see Note 3). 5. Transfer 50 μLto5Â 0.2 mL PCR tube and spin down. 6. Run PCR program 1 with the following cycling conditions: 95 C denaturation for 3 min 95 C for 30 s 50 C for 30 s 68 C for 1 min >>4 cycles 95 C for 30 s 55 C for 30 s 68 C for 1 min >>4 cycles 95 C for 30 s 63 C for 30 s 68 C for 1 min >>20 cycles 68 C for 7 min 7. Perform PCR puriﬁcation using Agencourt AMPure XP beads (for recovery of double-stranded and single-stranded DNA templates greater than 100 bp), according to the manufacturer’s protocol. Final elution volume is 140 μL. 8. Prepare the following reaction mix (PCR2):

Forward primer 5 μL (final concentration of 0.25 μM) Reverse primer 5 μL (final concentration of 0.25 μM) dNTPs 4 μL (final concentration of 200 μM for each) Buffer X5 20 μL FastStart™ High Fidelity 2 μL ddH2O24μL 470 Anna Vaisman-Mentesh and Yariv Wine

Fig. 4 Amplicon library on 1% agarose gel. Shown band is after PCR 2. The band migrates between approximately 550 bp in length should be excised and gel-puriﬁed

9. Add 140 μL of the puriﬁcation product obtained in step 7 to the reaction mix and pipette. 10. Transfer 50 μLto4Â 0.2 mL PCR tube and spin down. 11. Run PCR 2 with the following cycling conditions: 95 C denaturation for 3 min 95 C for 30 s 40 C for 30 s 68 C for 1 min >>2 cycles; 95 C for 30 s 65 C for 30 s 68 C for 1 min >>7 cycles 68 C for 7 min. 12. Subject PCR products to 1% agarose gel (see Fig. 4) and purify using Zymoclean™ Gel DNA Recovery Kit according to the manufacturer’s instructions.

13. Measure VH and VL libraries concentrations using Qubit system (Thermo Fisher Scientiﬁc). 14. Measure library quality using bioanalyzer system (Agilent) or tapestation (Agilent). The expected product range for VH is ~550 bp and for the VL is ~500 bp (see Fig. 5). Phage Biopanning and NGS 471

Fig. 5 Quality control of VH NGS library using Bioanalyzer system (Agilent). Libraries giving a single-clear peak around 550 bp are submitted to Illumina MiSeq

3.5 Illumina 1. VH and VL libraries from sorted B-cell populations/affinity Sequencing, Data selected phages are subjected to NGS on the Illumina MiSeq Analysis, platform with a MiSeq Reagent Kit V3 2 Â 300 bp paired-end and Statistics (Illumina), using an input concentration of 16 pM with 2–5% PhiX (better results were obtained using 5% PhiX). 2. The processing of the Illumina output file can be carried out by several bioinformatic tools including Change-O toolkit [15] and FLASH [16]. The following processing pipeline utilizes the software for comprehensive adaptive immunity profiling (MiXCR) [17]. Paired-end FASTQ files (prefix R1 and R2) acquired from Illumina MiSeq are processed to consist of the following steps: trimming of low-quality reads, merging of paired-end reads, and length filtering using with input parameters set as follows:

“align” -f , --report, --library imgt, -a “exportAlignments” -f, --preset-file, -cloneIdWithMappingType “assemble” -f, --report, --index, -OseparateByC¼true, - OclonalFactoryParameters.vParameters. featureToAlign¼ VRegion, - OssemblingFeatures¼CDR3, -OminimalClonalSequenceLength¼6 “exportClones” -f, --chains, --preset-file, -readIds, -o –t 472 Anna Vaisman-Mentesh and Yariv Wine

4 Notes

1. To obtain highly polarized human immunized libraries, blood samples should be collected 7–10 days post vaccination in order to “capture” a large amount of antigen-specific plasma- blasts (see dynamics of B cells following vaccination and phy- notypes in [12]). 2. In phage biopanning, the constructed phagmid library is isotype-specific, phage-displayed scFv library so VH primers described in Fig. 3 and listed in Table 1 are IgG specific. The primers used to amplify VL region should be either VƘ or Vλ, according to the constructed library (Table 1).

3. For monitoring phage panning progression, VH and VL from the recovered phages from each panning cycle should be used as templates instead of using cDNA.

References

1. Hoogenboom HR (2005) Selecting and 8. de Wildt RM, Mundy CR, Gorick BD, Tom- screening recombinant antibody libraries. Nat linson IM (2000) Antibody arrays for high- Biotechnol 23:1105–1116. https://doi.org/ throughput screening of antibody-antigen 10.1038/nbt1126 interactions. Nat Biotechnol 18:989–994. 2. Pande J, Szewczyk MM, Grover AK (2010) https://doi.org/10.1038/79494 Phage display: concept , innovations , applica- 9. Smith J, Kontermann RE, Embleton J, Kumar tions and future. Biotechnol Adv 28:849–858. S (2017) Antibody phage display technologies https://doi.org/10.1016/j.biotechadv.2010. with special reference to angiogenesis. FASEB J 07.004 19:331–341. https://doi.org/10.1096/fj.04- 3. Sblattero D, Bradbury A (2000) Exploiting 2863rev recombination in single bacteria to make large 10. Boutz DR, Horton AP, Wine Y, Lavinder JJ, phage antibody libraries. Nat Biotechnol Georgiou G, Marcotte EM (2014) Proteomic 18:75–80. https://doi.org/10.1038/71958 identiﬁcation of monoclonal antibodies from 4. Holt LJ, Enever C, de Wildt RM, Tomlinson serum. Anal Chem 86:4758–4766. https:// IM (2000) The use of recombinant antibodies doi.org/10.1021/ac4037679 in proteomics. Curr Opin Biotechnol 11. DeKosky BJ, Ippolito GC, Deschner RP, 11:445–449. https://doi.org/10.1016/ Lavinder JJ, Wine Y, Rawlings BM, S0958-1669(00)00133-6 Varadarajan N, Giesecke C, Do¨rner T, Andrews 5. Pavoni E, Monteriu` G, Cianfriglia M, Mine- SF, Wilson PC, Hunicke-Smith SP, Willson nkova O (2007) New display vector reduces CG, Ellington AD, Georgiou G (2013) High- biological bias for expression of antibodies in throughput sequencing of the paired human E. coli. Gene 391:120–129. https://doi.org/ immunoglobulin heavy and light chain reper- 10.1016/j.gene.2006.12.009 toire. Nat Biotechnol 31:166–169. https:// 6. Scott N, Reynolds CB, Wright MJ, Qazi O, doi.org/10.1038/nbt.2492 Fairweather N, Deonarain MP (2008) Single- 12. Lavinder JJ, Wine Y, Giesecke C, Ippolito GC, chain Fv phage display propensity exhibits Horton AP, Lungu OI, Hoi KH, DeKosky BJ, strong positive correlation with overall expres- Murrin EM, Wirth MM, Ellington AD, sion levels. BMC Biotechnol 10:1–10. https:// Do¨rner T, Marcotte EM, Boutz DR, Georgiou doi.org/10.1186/1472-6750-8-97 G (2014) Identiﬁcation and characterization of 7. Saggy I, Wine Y, Shefet-carasso L (2012) Anti- the constituent human serum antibodies eli- body isolation from immunized animals: com- cited by vaccination. Proc Natl Acad Sci U S parison of phage display and antibody A 111:2259–2264. https://doi.org/10. discovery via V gene repertoire mining. Protein 1073/pnas.1317793111 Eng Des Sel 25:539–549. https://doi.org/10. 13. Wine Y, Boutz DR, Lavinder JJ, Miklos AE, 1093/protein/gzs060 Hughes RA, Hoi KH, Jung ST, Horton AP, Murrin EM, Ellington AD, Marcotte EM, Phage Biopanning and NGS 473

Georgiou G (2013) Molecular deconvolution large-scale B cell immunoglobulin repertoire of the monoclonal antibodies that comprise the sequencing data. Bioinformatics polyclonal serum response. Proc Natl Acad Sci 31:3356–3358. https://doi.org/10.1093/bio U S A 110:2993–2998. https://doi.org/10. informatics/btv359 1073/pnas.1213737110 16. Magocˇ T, Salzberg SL (2011) FLASH: fast 14. Reddy ST, Ge X, Miklos AE, Hughes RA, Kang length adjustment of short reads to improve SH, Hoi KH, Chrysostomou C, Hunicke- genome assemblies. Bioinformatics Smith SP, Iverson BL, Tucker PW, Ellington 27:2957–2963. https://doi.org/10.1093/bio AD, Georgiou G (2010) Monoclonal antibo- informatics/btr507 dies isolated without screening by analyzing 17. Bolotin DA, Poslavsky S, Mitrophanov I, the variable-gene repertoire of plasma cells. Shugay M, Mamedov IZ, Putintseva EV, Chu- Nat Biotechnol 28:965–969. https://doi. dakov DM (2015) MiXCR: software for com- org/10.1038/nbt.1673 prehensive adaptive immunity proﬁling. Nat 15. Gupta NT, Vander Heiden JA, Uduman M, Methods 12:380–381. https://doi.org/10. Gadala-Maria D, Yaari G, Kleinstein SH 1038/nmeth.3364 (2015) Change-O: a toolkit for analyzing Part IV

Phage Display for Epitope Mapping and Identiﬁcation of Biomarkers Chapter 27

ORFeome Phage Display

Jonas Zantow, Gustavo Marc¸al Schmidt Garcia Moreira, Stefan Dubel,€ and Michael Hust

Abstract

ORFeome phage display allows the efficient functional screening of entire proteomes or even metapro- teomes to identify immunogenic proteins. For this purpose, randomly fragmented, whole genomes or metagenomes are cloned into a phage-display vector allowing positive selection for open reading frames (ORF) to improve the library quality. These libraries display all possible proteins encoded by a pathogen or a microbiome on the phage surface. Consequently, immunogenic proteins can be selected from these libraries using disease-related immunoglobulins from patient serum. ORFeome phage display in particular allows the identification of immunogenic proteins that are only expressed in the host-pathogen interaction but not in cultivation, as well as the detection of very low expressed and very small immunogens and immunogenic proteins of non-cultivable organisms. The identified immunogenic proteins are potential biomarkers for the development of diagnostic assays or vaccines. These articles will give an introduction to ORFeome phage- display technology and give detailed protocols to identify immunogenic proteins by phage display.

Key words Oligopeptide phage display, Genomic libraries, cDNA libraries, ORFeome phage display, Biomarker identiﬁcation, Immunogenic proteins, Open reading frame (ORF) selection, Antigen

1 Introduction

Phage display is a powerful tool for the identification of protein- protein interactions and has been used for the identification of allergens [1–3] and the identification of immunogenic proteins of pathogens, e.g., Mycobacterium tuberculosis [4], Mycoplasma mycoides [5], the rickettsia Cowdria ruminantium [6], or the eukaryotic pathogen Taenia solium [7] by cloning cDNA or whole genome fragments. However, non-directional cloning of cDNA or whole genome fragments result in a majority of sequences that do not encode any protein. Only 1 out of 18 randomly cloned DNA fragments statistically results in a correct open reading frame (ORF). This means, only fragments that are cloned in the correct orientation, reading frame and with the correct fragment length do not lead to frameshifts. Additional “junk” sequences originate from

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_27, © Springer Science+Business Media LLC 2018 477 478 Jonas Zantow et al.

noncoding DNA like gene regulatory elements. The “junk” content in libraries derived from eukaryotic genomic DNA is even more increased as introns account for a large portion of the eukaryotic genome. The selection of specifically bound clones is a single- binding event in the background of millions of nonspecific clones. The propagation of clones without insert and non-ORF inserts is more favored and in consequence affecting phage library composition, complexity, and the proportion of protein displaying particles which dramatically hampers the panning success [8, 9]. Therefore, decreasing the portion of DNA fragments in the library that do not code for a correct protein sequence substantially increases the chance to eventually identify specific protein interactions. To address this problem, the DNA libraries should be enriched for ORFs. Cloning cDNA upstream of a selection marker (e.g., β-lactamase conferring ampicillin resistance) in an E. coli vector allows the enrichment of ORFs as only DNA fragments without stop codons and cloned in-frame with the β-lactamase gene allow the expression of β-lactamase and the corresponding E. coli clone can grow in selection medium supplemented with ampicillin [10]. However, this method is tedious since the inserts have to be subcloned into a phage-display vector after ORF enrichment [11] or the β-lactamase gene has to be removed by CRE-mediated recombination [12] to allow selection of protein interaction partners. For phage display, the protein of interest is displayed by fusion to an M13 phage coat proteins, typically minor coat protein 3 (pIII). A phage particle has about five pIII copies and this protein provides the initial binding of the phage to the E. coli F pilus to initiate infection. Consequently, a phage particle without pIII is non-infectious and cannot be amplified. Due to their independent replication and regulation, phage-display vectors are usually phagemids, i.e., plasmids that can be packaged into phage particles and contain the coding DNA for the foreign protein fused to the pIII gene but no other phage proteins. In order to complement the missing phage proteins like major coat protein pVIII and other necessary non-structural proteins, the E. coli cells have to be infected with a helperphage. Different antibiotics resistances and origins of replication allow the coexistence of phagemid and helperphage genome in the same E. coli cell. However, an attenuated packaging signal in the helperphage genome ensures that the phagemid is preferably packaged into the phage particle. Using a helperphage with a trypsin-sensitive pIII but cloning the fragmented DNA upstream of a trypsin-resistant pIII gene, ORFs can be enriched by tryptic digest of the phage-display library and infection of E. coli [13]. The simultaneous enrichment for correct ORFs and functional selection using phage display without the need for protease treatment was enabled by another special helperphage, referred to as ORFeome Phage Display 479

Hyperphage [14, 15]. Hyperphage is a unique helperphage that has À a deletion in the pIII gene (genotype pIII ) but carries functional pIII proteins and, therefore, is infectious (phenotype pIII+). Infec- tion of E. coli provides it with all phage protein except pIII but does not result in the production of infectious helperphage particles. Therefore, infection with Hyperphage allows the enrichment of ORFs encoded in suitable phagemids, as cloning of randomly fragmented DNA upstream of the pIII gene only results in the production of pIII if the added DNA is in-frame with the pIII gene and without stop codon. Consequently, when using Hyperph- age as helperphage, the only pIII source is the fusion protein. As sequences outside of ORFs, inverted or frame-shifted sequences longer than about 100 base pairs statistically contain one or more stop codons, infectious phage particles are only produced if the inserted DNA represents an actual protein coding sequence [9]. - ORF-enriched phage-display libraries can directly be used to select immunogenic proteins in a panning using patient sera. Thus, ORFeome phage display combines the positive selection of ORFs from whole genomes and metagenomes with the functional display of protein fragments coded by these sequences. The procedure is given in Fig. 1. The ORFeome phage-display technology was used for the first time to identify novel immunogenic protein from two different Mycoplasma species [16, 17], animal pathogens that cause economic damage in cattle breeding (e.g., contagious bovine pleuro- pneumonia caused by M. mycoides in East Africa). For Salmonella Typhimurium, which causes severe diarrheal disease that can be life threatening, described immunogenic proteins were confirmed and also novel ones were identified using the ORFeome phage display [18]. Further, the technology was able to identify novel immunogenic proteins from Neisseria gonorrhoeae [19]. Recently, the technology was also validated to work with eukaryotic cDNA, as immunogenic saliva proteins were identified from Ixodes scapularis (tick) that are involved in the tick feeding [20]. Since ORFeome phage display is independent of cultivation and does not rely on sequenced or annotated genomes, it can also be applied on complex microbial communities. The technology applying another gIII- deficient helperphage was used to analyze a ruminal metasecretome by enriching the library for secreted proteins similar to ORFeome phage display and analyzing the metasecretome by next-generation sequencing [21]. In another recent study, the ORFeome phage display was used not only to construct libraries from gut microbiota-derived metagenomic DNA but also to successfully identify immunogenic proteins based on specific immunoglobulin binding using serum antibodies of a mouse model of experimental ileitis [22]. The proteins identified by ORFeome phage display are interesting candidates for potential future diagnostic or vaccination applications. 480 Jonas Zantow et al.

Fig. 1 ORFeome phage display combines the enrichment of open reading frames (ORF) and the functional display of proteins on the surface of phage particles. Randomly fragmented genomic DNA from pathogens or microbial communities is cloned into a phage display vector (phagemid; ori: origin of replication) upstream of the minor coat protein III (pIII) gene. When using a special helperphage (“Hyperphage,” deleted pIII gene), the ORFeome Phage Display 481

2 Materials

2.1 Isolation 1. DNA isolation kit. of Genomic DNA

2.2 Ampliﬁcation 1. illustra Ready-To-Go GenomiPhi V3 DNA Ampliﬁcation Kit of Genomic DNA (GE Healthcare, Freiburg, Germany). 2. DNA-free water. 3. PCR reaction tubes. 4. Thermocycler. 5. Agarose (Peqlab, Erlangen, Germany). 6. TAE-buffer 50Â: 2 M Tris–HCl, 1 M acetic acid, 0.05 M EDTA pH 8. 7. Electrophoresis chamber.

2.3 Fragmentation 1. Sonopuls HD2200 and MS72 Sonotrode (Bandelin, Berlin, of DNA Germany). 2. Amicon Ultra Centrifugal Filters (30k) (Merck Millipore, Tul- lagree, Ireland). 3. Gel and PCR puriﬁcation kit. 4. Agarose. 5. TAE-buffer 50Â. 6. Electrophoresis chamber.

2.4 DNA End Repair 1. Fast End Repair Kit (Thermo Fisher Scientiﬁc, Waltham, USA). 2. Gel and PCR puriﬁcation kit.

2.5 Library 1. Phagemid (in this protocol pHORF3). Construction 2. PmeI-HF (NEB, Frankfurt am Main, Germany). 3. CIP (NEB, Frankfurt am Main, Germany). 4. CutSmart Buffer (NEB, Frankfurt am Main, Germany). 5. NucleoSpin Gel and PCR clean-up (Macherey-Nagel, Duren,€

Germany). ä

Fig. 1 (Continued) phagemid encoded pIII fusion protein is the only pIII source. Infectious phage particles are only assembled if the cloned DNA sequence is in frame with the pIII gene and does not contain any stop codon leading to an enrichment of ORFs whereas the encoded protein is displayed at the same time on the phage particle. In a panning procedure using immobilized immune sera, ORF-enriched libraries can be screened for proteins that have induced immune responses. The coding sequence of the selected protein is contained in the phage particles and can subsequently be identiﬁed by DNA sequencing 482 Jonas Zantow et al.

6. T4 DNA ligase (Promega, Mannheim, Germany). 7. Amicon Ultra Centrifugal Filters (30k) (Merck Millipore, Tul- lagree, Ireland). 8. Glycerol. 9. 0.1 cm electroporation cuvette. 10. Electroporator. 11. SOC medium pH 7.0: 2% (w/v) tryptone, 0.5% (w/v) yeast extract, 0.05% (w/v) NaCl, 20 mM Mg solution, 20 mM glucose (sterilize magnesium and glucose separately, add solutions after autoclavation). 12. Polystyrene dish with lid (245 mm Â 245 mm Â 25 mm). 13. 2xTY medium pH 7.0: 1.6% (w/v) tryptone, 1% (w/v) yeast extract, 0.5% (w/v) NaCl. 14. 2xTY-GA agar: 2xTY, 100 mM glucose, 100 μg/mL ampicillin, 1.2% (w/v) agar-agar. 15. Electrocompetent E. coli TOP10F’ (Invitrogen, Carlsbad, USA) (F’[lacIq Tn10(tetR)] mcrA Δ(mrr-hsdRMS-mcrBC) φ80lacZΔM15 ΔlacX74 deoR nupG recA1 araD139 Δ(ara-- leu)7697 galU galK rpsL(StrR) endA1 λ-). 16. Liquid Nitrogen. 17. Single-use Drigalsky spatulas. 18. 2 mL cryo vials. 19. 10 cm Petri dishes. 20. Agarose. 21. TAE-buffer 50Â. 22. Electrophoresis chamber. 23. Optional: QIAxcel Advanced System (QIAGEN, Hilden, Germany).

2.6 Antigen Library 1. 2xTY medium pH 7.0. Packaging 2. 2xTY-T: 2xTY þ 20 μg/mL tetracycline. 3. 2xTY-GA: 2xTY þ 100 mM glucose þ100 μg/mL ampicillin. 4. 2xTY-GA agar plates. 5. 2xTY-AK: 2xTY þ 100 μg/mL ampicillin þ50 μg/mL kanamycin. 6. 10 cm Petri dishes. 7. Hyperphage (Progen, Heidelberg, Germany). 8. 1 mL cuvettes and spectrophotometer 600 nm wavelength. 9. 100 and 1000 mL glass shake ﬂasks. 10. 50 mL tubes. ORFeome Phage Display 483

Table 1 Oligonucleotide primers

Oligonucleotide primer Sequence 50-30

MHLacZ-Pro_f GGCTCGTATGTTGTGTGG MHgIII_r CTAAAGTTTTGTCGTCTTTCC

11. 0.45 μM syringe filters. 12. Syringe. 13. Incubator for shake flasks. 14. Eppendorf centrifuge (Eppendorf, Hamburg, Germany). 15. Sorval Centrifuge RC5B Plus, rotor F9S and SS34 (Thermo Fisher Scientific, Waltham, USA) and respective tubes. 16. Polyethylenglycol-Sodium Chloride (PEG-NaCl) solution: 20% (w/v) PEG 6000, 2.5 M NaCl. 17. Phosphate-buffered saline (PBS) pH 7.4:8.0 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4·2H2O, 0.24 g KH2PO4 in 1 L.

2.7 Colony PCR 1. Oligonucleotide primer (Table 1). 2. GoTaq DNA Polymerase and buffer (Promega, Frankfurt am Main, Germany). 3. dNTP Mix.

4. DNA-free H2O. 5. Thermocycler.

2.8 Antigen Panning 1. 96-well ELISA Costar plate (Corning, Corning, USA). and Screening 2. Phosphate-buffered saline (PBS) pH 7.4. 3. PBS-T: PBS þ Tween 20 0.05% (v/v). 4. 2% MPBS-T: skimmed milk powder 2% (w/v) diluted in PBS-T. 5. Tecan plate washing machine. 6. E. coli TOP10F’ (Invitrogen, Carlsbad, USA) (F’[lacIq Tn10 (tetR)] mcrA Δ(mrr-hsdRMS-mcrBC) φ80lacZΔM15 ΔlacX74 deoR nupG recA1 araD139 Δ(ara-leu)7697 galU galK rpsL(StrR) endA1 λ-). 7. 1 mL cuvettes and spectrophotometer 600 nm wavelength. 8. Trypsin (1 mg/mL stock). 9. Eppendorf centrifuge. 10. 2xTY medium pH 7.0. 11. 2xTY-T. 484 Jonas Zantow et al.

12. 2xTY-GA. 13. 2xTY-GA agar plates. 14. 2xTY-AK. 15. 96-well U-shaped polypropylene plate. 16. Hyperphage (Progen, Heidelberg, Germany). 17. PEG-NaCl. 18. Anti-Ig Fc-speciﬁc antibody HRP-conjugated. 19. Anti-M13 phage (pVIII) HRP-conjugated (GE Healthcare, Freiburg, Germany). 20. TMB solution.

21. 1 N H2SO4.

3 Methods

3.1 Isolation 1. Prepare genomic DNA from a bacterial culture (or other of Genomic DNA sources of bacteria) using any commercial kit suitable for the preparation of genomic DNA.

3.2 Amplification Approximately 20 μg template DNA is needed for library construc- of Genomic DNA tion. If DNA amounts are insufficient for library construction they can isothermally be amplified by multiple displacement amplification (MDA) using commercial kits like the illustra Ready-To-Go GenomiPhi V3 DNA Amplification Kit:

1. Dilute 10–100 ng of template DNA in 10 μL DNA-free H2O Milli-Q and mix with 10 μL2Â Denaturation Buffer provided with the kit. 2. Denature DNA for 3 min at 95 C and subsequently cool down on ice. 3. Transfer the denatured DNA to the reaction cake, ensure complete reconstitution of the cake, and incubate for 2 h at 30 C using a thermocycler. 4. Inactivate the polymerase for 10 min at 65 C. 5. Analyze the ampliﬁed DNA on a 1% (w/v) TAE agarose gel (see Note 1).

3.3 DNA 1. Dilute 20 μg template DNA in 2 mL H2O Milli-Q. Fragmentation 2. Fragment DNA by sonication (6Â 120 s, 50% intensity, HD2200 Sonopuls MS72 sonotrode) keeping the sample on ice. Chill on ice between fragmentation cycles. Check fragment size distribution by analyzing 20 μL of fragmented DNA using agarose gelelectrophoresis (1% (w/v) agarose gels). ORFeome Phage Display 485

Table 2 Reagents to be added on reaction for DNA-ends repair

Fragmented DNA (ﬁnal amount 5 μg) X μL 10Â end repair reaction mix 5 μL End repair enzyme mix 2.5 μL

H2O Milli-Q Up to 50 μL

Table 3 Reagents to be added on the linearization of the phagemid

Phagemid (total 5 μg) X μL Buffer CutSmart 10Â (NEB, Frankfurt am Main, Germany) 2 μL PmeI (10 U/μL, NEB, Frankfurt am Main, Germany) 1 μL

H2O Milli-Q Up to 20 μL

® 3. Concentrate the DNA fragments using Amicon Ultra-0.5 mL ® Centrifugal Filters Ultracel -30K and extract the desired fragment sizes by gel extraction or directly use a PCR puriﬁcation kit and adjust fragment size cutoff by diluting the high salt binding buffer according to the manufacturer’s information.

3.4 Removal Sonication of DNA results in DNA fragments with 50 or 30 over- of Cohesive Ends hangs. The cohesive ends have to be repaired and fragments have to be phosphorylation to allow blunt-end cloning in the linearized phage-display vector. Removal of cohesive ends and phosphorylation can be performed using the Fast End Repair Kit or any other commercial kit (Table 2): 1. Incubate the reaction for 15 min at 20 C (do not let it stand longer) and purify using a PCR puriﬁcation kit. Elute in 30 μL of the provided elution buffer or H2O Milli-Q. 2. Determine the DNA concentration.

3.5 Phagemid- The preparation of the phagemid varies with the kind of phage- Fragment Ligation display method used. In this protocol, it is necessary to use a and Library phagemid that allows the cloning in a blunt end, such as Construction pHORF3, which has a PmeI as a cloning site. Thus, perform the digestion as described in Table 3. 1. Incubate the reaction for 2 h at 37 C and add 1 μL of calf- intestinal alkaline phosphatase (10 U/μL, NEB, Frankfurt am Main, Germany). 486 Jonas Zantow et al.

Table 4 Reagents to be added on the ligation of gene fragments with phagemid

Digested ~4-kb phagemid (total 1 μg) X μL DNA fragments up to 1.5 kb (total 1.4 μg) Y μL T4 DNA ligase buffer 10Â (Promega) 10 μL T4 DNA ligase (3 U/μL, Promega) 3.33 μL

H2O Milli-Q Up to 100 μL

2. Incubate for 1 h at 37 C and purify the reaction using the NucleoSpin Gel and PCR clean-up kit (see Note 2). Elute in 20 μL Milli-Q water. 3. Perform the ligation reaction for 16 h at 16 C (Table 4).

1. Inactivate the ligase for 10 min at 65 C(see Note 3) and ® perform a buffer exchange using Amicon Ultra-0.5 mL Cen- ® trifugal Filters Ultracel -30K. For this, add 400 μL of Milli-Q water in the reaction and centrifuge (10 min, 10,000 Â g). Repeat this buffer exchange with 500 μL of Milli-Q water for four more times before collecting the final volume as instructed by the manufacturer (Place the centrifugal filter inverted in a fresh collection tube and centrifuge for 2 min at 1000 Â g.) 2. Split the eluted ligation volume (approximately 30 μL) into four individual transformations and mix with each 25 μL electrocompetent E. coli TOP10F’ in a 1.5-mL Eppendorf tube, transfer the volume to a prechilled 0.1 mm cuvette, and keep it on ice for 5 min. 3. Perform electroporation for bacteria (1.8 kV; pulse ~4.8 ms) and immediately add 1 mL of SOC medium pre-warmed at 37 C. 4. Transfer the cells to a 1.5 mL tube and incubate at 37 C for 1 h and 650 rpm. 5. Take 10 μL of the tube and add into 10 mL of 2xYT (first dilution). From this last tube, transfer 100 μL to 1 mL of 2xYT (second dilution) and from the second dilution another 100 μL to 1 mL of 2xYT (third dilution). Finally, plate 100 μL of each À À À dilution (final dilution factor 10 4,10 5,10 6) onto a 2xYT- GA agar 10 cm plate and grow it overnight at 37 C(see Note 4). 6. Plate the remaining ~990 μL of the transformation onto a 245 Â 245 Â 25 mm plate (“pizza plate”) with 2xYT-GA agar and incubate at 37 C overnight. 7. Perform the colony counting on the 10 cm plates. ORFeome Phage Display 487

Table 5 Composition of a colony PCR

Solution or component Volume Final concentration

dH2O 7.5 μL GoTaq buffer (5Â)2μL1Â dNTPs (10 mM each) 0.2 μL 200 μM each MHLacZPro_f 10 μM 0.1 μL 0.1 μM MHgIII_r10 μM 0.1 μL 0.1 μM GoTag (5 U/μL) 0.1 μL 0.5 U Template Picked colonies from dilution plate

8. On the “pizza plate”, add 20 mL of 2xYT and incubate on a shaker for 20 min. 9. With a Drigalsky spatula, carefully scrape the cells from the medium surface (see Note 5). Then, collect the liquid containing cells with a serological pipette in a 50 mL tube and supplement with 20% (v/v) glycerol and distribute 1 mL in each of 6 cryovials. 10. Flash freeze the cells in liquid nitrogen and wait for 5 min. Then, carefully take the tubes with proper protection gloves and store the tubes at À80 C promptly.

3.6 Library Quality 1. From the 10 cm plates used for counting on the previous topic, Control and Packaging take at least 20 colonies to perform a colony PCR. For this PCR, make one tube containing the empty phagemid used for the library construction as a negative control (Table 5). 2. Check the size of each fragment by electrophoresis (see Note 6). 3. Count the number of positives (those above the band of the negative control) expecting to have at least 80% (16/20) of the clones positive (this quality measurement is called “insert rate”). If the number is much below 80%, consider repeating previous steps, mainly the phagemid preparation or ligation (see Note 7).

3.7 Library 1. Gently thaw the library previously stored at À80 C on ice and Packaging and ORF pool the individual transformations (here n ¼ 4). Inoculate Enrichment 400 mL of 2xYT-GA in a 1000 mL shake ﬂask with the library (OD600 ¼ 0.1). 2. Incubate the shake ﬂask at 37 C, 250 rpm until logarithmic growth is reached (OD600 0.5). Then, transfer 25 mL 488 Jonas Zantow et al.

(1.25 Â 1010 cells) of the culture to a 50 mL tube and add 2.5 Â 1011 cfu (MOI 1:20) of Hyperphage. 3. Incubate the tube for 30 min at 37 C without shaking, and another 30 min at 37 C and 250 rpm. 4. Pellet the cells at 3220 Â g for 10 min, RT. Discard the supernatant, suspend the cells in 10 mL of 2xYT-AK, and transfer them to a 1000 mL shake flask containing 390 mL of the same medium. Incubate the flask at 30 C, 250 rpm for 24 h. 5. Transfer the culture to a 1000 mL centrifuge tube and centrifuge for 10 min at 10,000 Â g,4C. The supernatant contains the phage. Collect the supernatant into another 1000 mL centrifuge tube (see Note 8), add 1/5 volume (100 mL) of PEG-NaCl solution, mix thoroughly and incubate the tube at 4 C overnight. In parallel, inoculate a 100 mL shake flask containing 25 mL of 2xYT-T with E. coli TOP10F’ and incubate at 37 C, 250 rpm overnight. 6. Centrifuge the tube containing the supernatant with PEG-NaCl 10,000 Â g, 1 h, 4 C and discard the supernatant. 7. Suspend the pellet containing phage in 10 mL of prechilled PBS, filter the suspension with a 0.45 μm filter, and transfer to another 50 mL centrifuge tube. 8. Add 1/5 volume (2.5 mL) of PEG-NaCl solution and incubate for 1 h on ice on a rocker. 9. Centrifuge the suspension 20,000 Â g, 30 min, 4 C and discard the supernatant. 10. Suspend the pellet in 1 mL of Phage dilution buffer, transfer to a 1.5 mL tube, and centrifuge at 16,000 Â g, 30 min, 4 Cto remove remaining bacteria. 11. Transfer the supernatant to a cryovial and store it at 4 C until further use. 12. Take the E. coli TOP10F’ overnight culture, make another 30 mL 2xYT-T culture in a 100 mL shake flask with initial OD600 0.1, and incubate at 37 C, 250 rpm until OD600 0.5. 13. Prepare six 1.5 mL tubes for phage dilution, three with 990 μL and three with 900 μL of PBS. First, use 10 μL of the phage prepared on step 11 to make the three 100-fold dilutions on the tubes with 990 μL. Then, make three tenfold dilutions by adding 100 μL of the last tube on the remaining three tubes À À À À with 900 μL (these will be dilutions 10 2,10 4,10 6,10 7, À À 10 8, and 10 9). 14. Prepare four 1.5 mL tubes with 50 μLofE. coli TOP10F’ cells in each and transfer 10 μL of the last four phage dilutions to ORFeome Phage Display 489

À À À À each tube (these will be dilutions 10 8,10 9,10 10,10 11 on the plate) (see Note 9). 15. Incubate the tubes at 37 C for 30 min without shaking. 16. Divide one 2xYT-GA agar plate into 4 parts and make three 10 μL droplets of each of the four dilutions on each part. Let the droplets dry under the biological cabinet for 5 min and incubate the plate at 37 C for 16 h. 17. Spread the remaining volume (30 μL) of the two intermedi- À À ary dilutions (10 9, and 10 10) on 2xYT-GA agar plates. 18. Count the colonies on countable droplets and calculate the titer as the arithmetic mean of the three droplets and multiply per six, so the ﬁnal result will be in cfu/mL. This quality measurement is called “library titer.” 19. From the other plate, pick at least 20 colonies, analyze insert- rate and size by colony PCR, and sequence the DNA expecting to have at least 50% (10/20) of the clones with in-frame and correct sequence. This quality measurement is called “in-frame rate.” 20. Analyze the colony PCR by electrophoresis (see Note 10).

3.8 Colony PCR Choose 10–20 single colonies per transformation. Set up the 10 μL PCR reaction per colony as follows (see Table 1 for primer sequences): Suggested PCR program: 95 C, 120 s þ 95 C, 15 s; 54 C, 20 s; 72 C, 120 s (25 cycles) þ 72 C 10 min þ 4 C forever.

3.9 Antigen Panning 1. Inoculate an overnight culture of E. coli TOP10F’ in 30 mL 2xYT-T in a 100 mL ﬂask. 2. Coat 2 wells of a 96-well high-binding ELISA plate with 2 μg of a puriﬁed antibody suitable to capture the desired antibody isotype and species (diluted in 200 μL of PBS) (following referred to as “selection wells”). In parallel, coat 4 wells with 5 Â 1010 cfu Hyperphage (in 150 μL PBS) (following referred to as “pre-clearance wells”). Coating can be performed at 4 C overnight (see Note 11). 3. On the next day, remove the solutions and add 350 μL2% MPBST in each of the wells to saturate the protein binding capacity (1 h at room temperature). 4. Wash the pre-clearance wells three times with PBST. 5. Dilute the serum 1:100–1:1000 with PBST (2 Â 150 μL), transfer to two of the pre-clearance wells, and incubate for 1 h at RT. 6. Transfer the sera to the second two pre-clearance wells and incubate for 1 h at RT. 490 Jonas Zantow et al.

7. Wash the selection wells three times with PBST and transfer the pre-cleared sera to the selection wells. Incubate for 1 h at RT to capture the serum antibodies. 8. Remove excess serum in the selection wells by three washing steps with PBST. Dilute 1 Â 1011 cfu (or at least 100-fold excess of library size) of the library in 150 μL MPBST and apply the libraries to the captured serum antibodies. Incubate for 2 h at RT. 9. After approximately 1 h, inoculate 30 mL 2xYT-T with the E. coli TOP10F’ overnight culture (OD600 ¼ 0.1) and cultivate at 37 C and 250 rpm until logarithmic growth is reached (OD600 ¼ 0.5, approximately 1.5 h). 10. Remove non-bound antigen-phage in ten washing cycles using PBST (see Note 12). 11. Elute the bound phage using 200 μL Trypsin solution (10 μg/ mL in PBS) for 30 min at 37 C and pool the elutions (The phagemid encoded pIII fusion protein harbors a trypsin site). 12. Prepare three 1.5 mL tubes with 90 μL PBS for phage dilution. First, use 10 μL of the phage prepared on step 11 to make the three tenfold dilutions on the tubes with 90 μL (these will be À À À dilutions 10 2,10 3,10 4). The dilutions have to be adjusted each panning round according to the expected elution titers (see Note 13). 13. Prepare four 1.5 mL tubes with 50 μLofE. coli TOP10F’ cells in each and transfer 10 μL of the non-diluted phage and each of À À the respective dilutions (these will be dilutions 10 2,10 3, À À 10 4,10 5 on the plate). 14. Incubate the tubes at 37 C for 30 min without shaking. 15. Divide one 2xYT-GA agar plate into four parts and make three 10 μL droplets of each of the four dilutions on each part. Let the droplets dry under the biological cabinet for 5 min and incubate the plate at 37 C for 16 h. 16. Use the remaining 380 μL of eluted phage to infect 5 mL of the E. coli TOP10F’ culture and incubate for 30 min at 37 C without shaking. 17. Pellet the cells for 10 min at 3,220 Â g and discard the supernatant. Suspend the pellet in up to 500 μL and plate on a 15 cm 2xYT-GA agar plate. Incubate at 37 C overnight. 18. On the 15 cm plate, add 5 mL of 2xYT and incubate on a shaker for 20 min. 19. With a Drigalsky spatula, carefully scrape the cells from the medium surface (see Note 14). Collect the liquid containing the cells and inoculate 30 mL 2xYT-GA (OD600 ¼ 0.1) and ORFeome Phage Display 491

incubate at 37 C and 250 rpm until logarithmic growth is reached (OD600 ¼ 0.5). 20. Collect 5 mL of the culture (~2.5 Â 109 cells) and infect with 5 Â 1010 cfu Hyperphage (MOI 1:20). Incubate for 30 min at 37 C without shaking and for another 30 min at 37 C and 250 rpm. 21. Pellet the cells for 10 min at 3,220 Â g and discard the supernatant. Suspend the cells in 1 mL 2xYT-AK and transfer to a 100 mL shaking ﬂask with 29 mL 2xYT-AK. Incubate at 30 C and 250 rpm overnight. 22. Pellet the cells for 10 min at 3,220 Â g and transfer the supernatant to another tube (the supernatant contains the phage). 23. Add 1/5 volume (ca. 6 mL) PEG-NaCl, mix thoroughly, and incubate for 1 h on ice to precipitate phage. 24. Pellet the phage for 1 h at 3,220 Â g and 4 C. Discard and completely remove the supernatant. 25. Suspend the pellet in 1 mL of phage dilution buffer, transfer to a 1.5 mL tube, and centrifuge at 16,000 Â g, 30 min, 4 Cto remove remaining cells. Transfer the supernatant to a cryovial and store at 4 C until used. 26. Repeat the steps above for another 1 two 2 rounds of panning. Stop at step 14 in the last panning round. 27. Instead of applying 10 μL droplets to the agar plates, plate the whole 60 μL for each dilution on an individual 2xYT-GA agar plate in order to allow screening of individual clones.

3.10 Monoclonal 1. In a 96-well U-bottom propylene plate, add 180 μL/well of Phage Production 2xYT-GA. and Screening ELISA 2. Pick 92 colonies from the plates described on the last step of the previous part. In this same plate, include 2 wells (H3 and H9) with medium only, 1 well (H6) with a colony to produce a non-related phage, and 1 well (H12) with the same colony added on H11. 3. Add a breathable sticker over the plate and incubate at 37 C, 800 rpm, for 6 h (this will be called “Master plate”) (see Note 15). 4. In another 96-well U-shaped propylene plate, add 180 μL/well of 2xYT-GA, and transfer 10 μL of the previously grown plate to this new one. Store the Master plate at 4 C, and incubate the new one at 37 C, 800 rpm, for 2 h. 5. Dilute puriﬁed Hyperphage in 2xYT to the concentration of 1 Â 1011 cfu/mL, and add 50 μL/well (5 Â 109 cfu/well). 492 Jonas Zantow et al.

6. Incubate for 30 min at 37 C without shaking, followed by 30 min at 37 C under 800 rpm. 7. Centrifuge the plate 3,220 Â g for 10 min at RT, remove the supernatant, and add 190 μL/well of 2xYT-AK. 8. Incubate the plate overnight at 30 C, 800 rpm. 9. Centrifuge the plate at 3,220 Â g for 10 min, RT, transfer 150 μL of each supernatant to a new plate, and add 40 μL/ well of PEG-NaCl solution. 10. Incubate the plate 1 h at 4 C, and centrifuge 3,220 Â g for1hat 4 C. 11. Completely remove the supernatant ensuring not to touch the pellet and suspend each pellet in 150 μL of PBS. 12. Shake the plate for 5 min under 500 rpm, and centrifuge 3,220 Â g for 10 min at 4 C to pellet remaining bacteria. 13. Coat each well of a high-binding ELISA plate with an anti- M13 (pVIII speciﬁc) antibody of a species different to the used serum at 4 C overnight (see Note 16). 14. Discard the content on the ELISA plate and add 350 μL/well of 2% MPBS-T. 15. Add 50 μL/well of 2% MPBST (except well H9), and then add 50 μL of the supernatant from the phage production plate (step 12), diluting the phage 1:2. On well H9, add 3 Â 108 cfu of Hyperphage as a negative control and incubate for 1.5 h at RT to capture the monoclonal oligopeptide phage. 16. Wash the plate three times with PBST. 17. Dilute the serum according to a previously determined dilution (e.g., titration ELISA on cell lysate) in 2% MPBS-T, and add 100 μL/well on each well, except H12. On H12, add 100 μL of 2% MPBS-T only (see Note 17). 18. Incubate for 1.5 h at room temperature, and wash the plate three times with PBST. 19. Dilute an appropriate detection antibody-HRP conjugate in 2% MPBST, and add 100 μL/well on each well, except H12. On H12, add 100 μL of an anti-M13 (pVIII) HRP-conjugated antibody in 2% MPBST. 20. Incubate for 1 h at room temperature, and wash the plate three times with PBST. 21. Add 100 μL/well of TMB ELISA developing solution and incubate at room temperature until single wells exhibit a sig- niﬁcant blue color (5–30 min). Stop the reaction by adding 100 μL/well of 1 N H2SO4 (the blue color will turn yellow). Acquire the data with an ELISA plate reader at 450 nm, using 620 nm as a reference wavelength. ORFeome Phage Display 493

4 Notes

1. The amplified DNA appears as smear between about 15 kb and 1 kb. The genomic template DNA appears as rather distinct band >20 kb. 2. Check the manufacturer’s FAQ’s for the use of CIP on blunt ends. Some manufacturers recommend adjusted protocols for the dephosphorylation of blunt ends (NEB recommends 50 C for dephosphorylation of blunt ends). 3. Ligase inactivation is crucial. Skipping this step will negatively influence the transformation rates. 4. Transformation rates between 107 and 108 clones per transformation are expected. 5. It is important to scrape all colonies. 6. Colony PCR can be analyzed by agaraose gel electrophoresis or capillary electrophoresis like on Qiaxcel Advance system. There will be a certain distribution of insert size depending on cutoffs used for library construction. The better the resolution of the used technique the more precise is the estimation of mean insert size. We usually prefer to use capillary electrophoresis. 7. There is one rule of thumb: The higher the insert rate the better the performance of ORF enrichment. However, we found the ORF filtering to be quite efficient even with libraries of 50% insert rate and less. 8. If the supernatant still contains bacteria consider another centrifugation step as this will alleviate the filtration step after the first precipitation. 9. Depending on the library you should expect phage titers between 1010 and 1012 cfu/mL. 10. Mean insert size often decreases after ORF enrichment. 11. Optional: Pre-clearance of the library with non-relevant serum antibodies to remove nonspecific binders from the library. Therefore, immobilize the serum capture antibody in another 2 wells and perform antibody capturing in parallel to the capturing in the selection wells. Perform pre-clearance of the library in parallel to the pre-clearance step of the serum used for selection (1 h immobilization and 2 h pre-clearance). 12. Washing cycles can be increased with the panning rounds (1st round: 10 cycles, 2nd round: 20 cycles, 3rd round: 30 cycles). 13. Typical elutions are 103–104 total cfu in panning round 1 and 106–107 total cfu in panning round 2. 14. Consider storing 2Â 1 mL of the scraped at À80 C (supplemented with 20% (v/v) glycerol). 494 Jonas Zantow et al.

15. Alternatively, this step can be performed at 34 C overnight. 16. This can be prepared in parallel with the phage production. 17. If you experience high background consider further dilution of the serum and competitive addition of E. coli cell lysate.

Acknowledgments

The introduction of this article contains updated and extended parts of the German language article Zantow et al. [23] and the dissertation of Jonas Zantow, TU Braunschweig.

References

1. Rhyner C, Weichel M, Fluckiger€ S, 9. Hust M, Meysing M, Schirrmann T, Selke M, Hemmann S, Kleber-Janke T, Crameri R Meens J, Gerlach G-F, Dubel€ S (2006) Enrich- (2004) Cloning allergens via phage display. ment of open reading frames presented on bac- Methods (San Diego, Calif.) 32(3):212–218 teriophage M13 using hyperphage. 2. Crameri R, Walter G (1999) Selective enrich- BioTechniques 41(3):335–342 ment and high-throughput screening of phage 10. Seehaus T, Breitling F, Dubel€ S, Klewinghaus I, surface-displayed cDNA libraries from complex Little M (1992) A vector for the removal of allergenic systems. Comb Chem High deletion mutants from antibody libraries. Gene Throughput Screen 2(2):63–72 114(2):235–237 3. Kodzius R, Rhyner C, Konthur Z, Buczek D, 11. Faix PH, Burg MA, Gonzales M, Ravey EP, Lehrach H, Walter G, Crameri R (2003) Rapid Baird A, Larocca D (2004) Phage display of identification of allergen-encoding cDNA cDNA libraries: enrichment of cDNA expres- clones by phage display and high-density sion using open reading frame selection. Bio- arrays. Comb Chem High Throughput Screen techniques 36(6):1018–1022, 1024, 6(2):147–154 1026–1029 4. Liu S, Han W, Sun C, Lei L, Feng X, Yan S, 12. Di Niro R, Sulic AM, Mignone F, D’Angelo S, Diao Y, Gao Y, Zhao H, Liu Q, Yao C, Li M Bordoni R, Iacono M, Marzari R, Gaiotto T, (2011) Subtractive screening with the Myco- Lavric M, Bradbury ARM, Biancone L, Zevin- bacterium tuberculosis surface protein phage Sonkin D, De Bellis G, Santoro C, Sblattero D display library. Tuberculosis (Edinb) 91 (2010) Rapid interactome profiling by massive (6):579–586 sequencing. Nucleic Acids Res 38(9): 5. Miltiadou DR, Mather A, Vilei EM, Plessis e110–e110 DHD (2009) Identification of genes coding 13. Gupta A, Shrivastava N, Grover P, Singh A, for B cell antigens of Mycoplasma mycoides Mathur K, Verma V, Kaur C, Chaudhary VK subsp. mycoides Small Colony (MmmSC) by (2013) A novel helper phage enabling con- using phage display. BMC Microbiol 9:215 struction of genome-scale ORF-enriched 6. Fehrsen J, du Plessis DH (1999) Cross-reactive phage display libraries. PLoS One. 8(9): epitope mimics in a fragmented-genome phage e75212 display library derived from the rickettsia, 14. Rondot S, Koch J, Breitling F, Dubel€ S (2001) Cowdria ruminantium. Immunotechnology 4 A helper phage to improve single-chain anti- (3–4):175–184 body presentation in phage display. Nat Bio- 7. Gonza´lez E, Robles Y, Govezensky T, Bobes technol 19(1):75–78 RJ, Gevorkian G, Manoutcharian K (2010) 15. Soltes G, Hust M, Ng KKY, Bansal A, Field J, Isolation of neurocysticercosis-related antigens Stewart DIH, Dubel€ S, Cha S, Wiersma EJ from a genomic phage display library of Taenia (2007) On the influence of vector design on solium. J Biomol Screen 15(10):1268–1273 antibody phage display. J Biotechnol 127 8. Stratmann T, Kang AS (2005) Cognate (4):626–637 peptide-receptor ligand mapping by directed 16. Kugler€ J, Nieswandt S, Gerlach GF, Meens J, phage display. Proteome Sci 3:7 Schirrmann T, Hust M (2008) Identification of ORFeome Phage Display 495

immunogenic polypeptides from a Myco- 20. Becker M, Felsberger A, Frenzel A, Shattuck plasma hyopneumoniae genome library by WMC, Dyer M, Kugler€ J, Zantow J, Mather phage display. Appl Microbiol Biotechnol 80 TN, Hust M (2015) Application of M13 phage (3):447–458 display for identifying immunogenic proteins 17. Naseem S, Meens J, Jores J, Heller M, Dubel€ S, from tick (Ixodes scapularis) saliva. BMC Bio- Hust M, Gerlach G-F (2010) Phage display- technol 15(1):43 based identification and potential diagnostic 21. Ciric M, Moon CD, Leahy SC, Creevey CJ, application of novel antigens from Mycoplasma Altermann E, Attwood GT, Rakonjac J, Gagic mycoides subsp. mycoides small colony type. D (2014) Metasecretome-selective phage dis- Vet Microbiol 142(3–4):285–292 play approach for mining the functional poten- 18. Meyer T, Schirrmann T, Frenzel A, Miethe S, tial of a rumen microbial community. BMC Stratmann-Selke J, Gerlach GF, Strutzberg- Genomics 15(1):356 Minder K, Dubel€ S, Hust M (2012) Identifica- 22. Zantow J, Just S, Lagkouvardos I, Kisling S, tion of immunogenic proteins and generation Dubel€ S, Lepage P, Clavel T, Hust M (2016) of antibodies against Salmonella typhimurium Mining gut microbiome oligopeptides by func- using phage display. BMC Biotechnol 12(1):29 tional metaproteome display. Sci Rep 6:34337 19. Connor DO, Zantow J, Hust M, Bier FF, von 23. Zantow J, Dubel€ S, Hust M (2016) Funktio- Nickisch-Rosenegk M (2016) Identification of nales Proteom-Display zur Identifikation von novel immunogenic proteins of Neisseria Biomarkern. BIOspektrum 22(3):256–259 gonorrhoeae by phage display. PLoS One 11 (2):e0148986 Chapter 28

Epitope Mapping by Phage Display

Gustavo Marc¸al Schmidt Garcia Moreira, Viola Fuhner,€ and Michael Hust

Abstract

Among the molecules of the immune system, antibodies, particularly monoclonal antibodies (mAbs), have been shown to be interesting for many biological applications. Due to their ability to recognize only a unique part of their target, mAbs are usually very speciﬁc. These targets can have many different compositions, but the most common ones are proteins or peptides that are usually from outside the host, although self-proteins can also be targeted in autoimmune diseases, or in some types of cancer. The parts of a mAb that interact with its target compose the paratope, while the recognized parts of the target compose the epitope. Knowing the epitope is valuable for the improvement of a biological product, e.g., a diagnostic assay, a therapeutic mAb, or a vaccine, as well as for the elucidation of immune responses. The current techniques for epitope mapping rely on the presentation of the target, or parts of it, in a way that it can interact with a certain mAb. Even though there are several techniques available, each has its pros and cons. Thus, the choice for one of them is usually dependent on the preference and availability of the researcher, opening possibility for improvement, or development of alternative techniques. Phage display, for example, is a versatile technology, which allows the presentation of many different oligopeptides that can be tested against different antibodies, ﬁtting the need for an epitope mapping approach. In this chapter, a protocol for the construction of a single-target oligopeptide phage library, as well as for the panning procedure for epitope mapping using phage display is given.

Key words Epitope mapping, Phage display, Antigen, Oligopeptide phage display, Genomic library, Single-gene library

1 Introduction

The key molecules of the adaptive immune system are antibodies/ B-cell receptors, and T-cell receptors, which are involved in the recognition of potentially harmful structures, called antigens [1, 2]. These antigens can have different compositions (e.g., lipids, carbohydrates, etc.), but the most common ones are proteins or peptides usually derived from pathogens. Under certain circumstances, self-antigens can also be targeted by the immune system, characterizing autoimmune diseases or certain types of cancer [3, 4]. Among the molecules of the immune system, the antibodies play a key role in the host organism, since they speciﬁcally recognize

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_28, © Springer Science+Business Media LLC 2018 497 498 Gustavo Marc¸ al Schmidt Garcia Moreira et al.

antigens and either act directly or trigger further immune responses against it [5]. The binding of an antibody to its target is usually highly specific, meaning that one antibody can only recognize a certain part of the antigen. It is known that antibodies interact with their target through a defined part of their structure, called “paratope”, which consists of amino acids contained in the complementarity-determining regions (CDRs). The target, in turn, also has a small part of its structure, called “epitope”, which is recognized by the paratope [6, 7]. Thus, an epitope is a small amino acid sequence that allows the interaction with the paratope via non-covalent interactions (i.e., ionic interactions, hydrogen bounds, hydrophobic interactions, etc.). In principle, there are two possible kinds of epitopes: linear, and conformational. The former type is characterized by amino acids that are very close to each other in the protein sequence (usually among a sequence of 4–30 amino acids), while the latter contains amino acids that are far from each other on the primary structure, but very close on the tertiary or quaternary structure [8, 9]. Due to their high specificity and stability, antibodies, especially monoclonal antibodies (mAbs), are valuable molecules for therapy against cancer, autoimmune diseases, or infections [10–12]. In an infectious disease context, knowing the target of antibodies raised by the host against a pathogen allows using this target as a vaccine [13, 14], as well as permits the use of a mAb against this target as a therapeutic molecule [15]. In the diagnostic field, mAbs are used for the direct detection of pathogens, or other biomarkers with diagnostic value. In most works with mAbs, it is interesting to define their epitope, since it may be crucial not only to enhance the efficacy of diagnostics, therapy, or vaccines [16, 17], but also to understand immune responses [18]. The principle of all techniques available for epitope mapping is to provide a source of a target protein or oligopeptide, which can be tested against a certain mAb. In this way, it is possible to define a minimum number of amino acids of the target as essential for the interaction, and, thus, determine the epitope. The most used techniques for this purpose are site-directed mutagenesis, high- throughput mutagenesis, array-based oligopeptide scanning, and X-ray co-crystallography. The site-directed mutagenesis consists of adding mutations on a gene in a way that some amino acids will be changed [19]. In this way, the role of this amino acid can be verified by testing the protein against the studied antibody and check for loss of reactivity, for example. Although it allows the study of both linear and conformational epitopes, only few mutations can be added at a time. Thus, it is essential to have previous information of the binding region in order to perform the mutations. Further- more, the time to obtain the protein for each mutated variant is quite long, turning this method very laborious and time- consuming. The high-throughput mutagenesis tries to overcome Epitope Mapping by Phage Display 499 these problems, since a library is generated containing mutations on every position of a certain target [20]. In this way, each variant is expressed and tested against the studied mAb. Although it is a very interesting and successful approach, most of the works involve viral antigens interesting for therapy, and require an advanced structure for high-throughput expression of proteins (usually in mammalian cells), as well as a good set of software for data interpretation. In an array-based oligopeptide scanning, overlapping and nonoverlap- ping peptides are synthesized and immobilized on a surface (e.g., on plates, or nitrocellulose membranes) that allows the test against antibodies [21]. There is also the possibility of combining different peptides, or modifying them, in a way that conformational epitopes can also be mapped. Although the immobilized peptides can be used to characterize different antibodies, this technique is relatively expensive for basic research. From the four aforementioned techniques, the X-ray co-crystallography is considered the gold standard for epitope mapping, since it can give interaction information not only on amino acid level, but also in atomic level [22, 23]. The method consists of mixing both the target and the studied antibody in optimal concentrations with a buffer for the development of a two-protein (antigen-antibody) crystal, which is than diffracted with an X-ray source to determine the tridimensional structure of the protein complex. Even though crystallography is the approach that gives the most refined and reliable data, the development and diffraction of crystals is still a limitation, mainly when the target has special characteristics (e.g., when it is a lipophilic, or a highly flexible protein). So far, although these techniques have been shown to be effective, none is considered an easy-to-do method that can be performed in most of the situations. This often leads to the search for new alternatives, such as bioinformatics analysis [24] or adaptation of other procedures (e.g., H/De-exchange mass spectrometry) [25, 26]. In this way, since most of the techniques still have room for improvement, there is still demand for new ones, which can help in solving the current problems, or complement the information for epitope mapping. Phage-display technology has been extensively used to generate useful antibodies for diagnostics, therapy, or basic research [27–29]. Due to the high versatility of the technique regarding the displayed molecules on phage surface, not only antibody sequences are used, but also random peptides with binding properties that can show applications similar to antibodies [30, 31]. The principle of presenting random peptides or parts of proteins, instead of complete ones, enabled the use of phage display on pathogen research, mainly for the identification of novel biomarkers for diagnostics, therapy, or even vaccine applications [32]. The same approach allows protein-protein interaction studies, such as antigen-antibody binding. Moreover, the main advantage of this technique is that both the phenotype (oligopeptide on phage 500 Gustavo Març al Schmidt Garcia Moreira et al.

surface) and genotype (coding sequence inside the phage) are present on the same system. In this way, it was possible to adapt the technology for the display of oligopeptides encoded by DNA fragments from different organisms and sources [33, 34]. Because phage-display technology allows the presentation of many different oligopeptides in a library scale with a single source of DNA, it is attractive for epitope mapping applications, since different parts of a certain target can be displayed [35]. Further- more, by having coupled phenotype and genotype, further techniques, such as the site-directed mutagenesis, can be easily used together with phage display to refine the results [36]. The use of phage display for epitope mapping is based on two strategies regarding the used library: the use of a random peptide library; or a library containing parts of only one already defined target [37]. In the former, small oligopeptides (~20 amino acids) with random sequences are displayed and used to perform panning with the studied mAb. Since the oligopeptides are random and not related to the actual target, the resulting sequences show conserved properties, but have to be analyzed carefully to determine the corresponding parts on the target [38]. On the other hand, the use of libraries containing sequences of a single target (called “single-target library”) can provide more reliable information regarding the recognized epitope, once parts of the antigen can be directly defined as the epitope without further complex analysis. Besides this single-target approach, it has been shown that a genome library, built with the ORFeome phage-display protocol described in the previous chapter of this book, also allows mapping epitopes [16]. In any case, it is worth considering that, although the literature describes the epitope mapping by phage display as being more reliable for linear epitopes, there are also data describing its effectiveness for conformational ones [39, 40]. Additionally, since many possibilities for the improvement of this technique for epitope mapping were not tested yet, its applicability is still an open field of study. In this chapter, we describe a protocol for epitope mapping using single-target libraries, beginning with the construction of the library, and going until the interpretation of the results to find the epitope. Considering that epitope mapping is a complex field, which can show many possible results depending on the mAb-target combination used, the presented technique does not rule out its refinement or confirmation with additional approaches.

2 Materials

2.1 Antigen Library 1. Primers for gene ampliﬁcation (designed by the researcher). Construction (for 2. Phusion DNA polymerase þ buffer 5Â (NEB, Frankfurt, Subheadings 3.1–3.3) Germany). Epitope Mapping by Phage Display 501

3. dNTP mix (10 mM each). 4. Agarose. 5. TAE-buffer 50Â: 2 M Tris–HCl, 1 M acetic acid, 0.05 M EDTA, pH 8.0 6. NanoDrop spectrophotometer (Thermo Fisher Scientific, Wal- tham, USA). 7. Gel and PCR purification kit (Macherey-Nagel, Duren,€ Germany). ® 8. Sonicator Bioruptor Plus Sonication System (Diagenode, Seraing, Belgium). ® ® 9. Amicon Ultra-0.5 mL Centrifugal Filters Ultracel -30K (Millipore). 10. Fast DNA End Repair kit (Thermo Fisher Scientific). 11. Phagemid (pHORF3 [32] is used in this protocol). 12. PmeI endonuclease þ buffer (NEB). 13. Calf intestine phosphatase (CIP; NEB). 14. T4 ligase þ buffer (Promega). 15. E. coli TOP10 F0 (Thermofisher), genotype: F0{lacIq, Tn10 (TetR)} mcrA Δ(mrr-hsdRMS-mcrBC) Φ80lacZΔM15 ΔlacX74 recA1 araD139 Δ(ara leu) 7697 galU galK rpsL (StrR) endA1 nupG. 16. MicroPuser Electroporator (Bio-Rad). 17. SOC medium pH 7.0: 2% (w/v) tryptone, 0.5% (w/v) yeast extract, 0.05% (w/v) NaCl, 20 mM Mg solution, 20 mM glucose (sterilize magnesium and glucose separately, add solutions after autoclavation). 18. Ampicillin (100 mg/mL stock). 19. Kanamycin (50 mg/mL stock). 20. Tetracyclin (20 mg/mL stock). 21. 2 M Glucose (autoclaved). 22. 2Â TY medium: 1.6% (w/v) tryptone, 1% (w/v) yeast extract, 0.5% (w/v) NaCl. 23. 2Â TY-glycerol: 1.6% (w/v) tryptone, 1% (w/v) yeast extract, 0.5% (w/v) NaCl, 16% (v/v) glycerin. 24. 2Â TY-T: 2Â TY, 20 μg/mL tetracycline. 25. 2Â TY-GA: 2Â TY, 100 mM glucose, 100 μg/mL ampicillin. 26. 2Â TY-GA agar plates: 2Â TY-GA, 1.5% (w/v) agar-agar. 27. 2Â TY-AK: 2Â TY, 100 μg/mL ampicillin, 50 μg/mL kanamycin. 28. Single-use Drigalsky spatulas. 502 Gustavo Març al Schmidt Garcia Moreira et al.

29. 10 cm Petri dishes. 30. 24.5 Â 24.5 Â 2.5 cm plates. 31. 2 mL cryovials. 32. Liquid Nitrogen. 33. À80 C freezer. 34. Hyperphage for oligovalent display (Progen, Heidelberg, Germany). 35. 1 mL cuvettes and spectrophotometer with 600 nm wavelength. 36. Taq DNA polymerase þ buffer 5Â (Promega, Heidelberg, Germany). 37. 100 and 500 mL glass shake flasks. 38. 50 mL tubes. 39. Incubator for shake flasks. 40. Refrigerated centrifuge with holders for 15 and 50 mL tubes, and plates. 41. Sorval Centrifuge RC5B Plus, rotor GS3 and SS34 (Thermo Fisher Scientific) and respective tubes. 42. Polyethylenglycol-Sodium Chloride (PEG-NaCl) solution: 20% (w/v) PEG 6000, 2.5 M NaCl. 43. Phosphate-buffered saline (PBS) pH 7.4: 8.0 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4∙2H2O, 0.24 g KH2PO4 in 1 L. 44. Phage elution buffer (PEB) pH 7.5: 10 mM Tris–HCl, 20 mM NaCl, 2 mM EDTA. 45. E. coli XL1-Blue MRF‘(Agilent, Santa Clara, CA, USA), genotype: Δ(mcrA)183 Δ(mcrCB-hsdSMR-mrr)173 endA1 supE44 thi-1 recA1 gyrA96 relA1 lac [F0 proAB lacIqZΔM15 Tn10 (Tetr)].

2.2 Antigen Panning 1. 96-well ELISA Costar plate (Corning). and Screening (for 2. PBS. Subheadings 3.4 3. PBS-T (PBS, Tween 20 0.05% (v/v)). and 3.5) 4. Panning block solution (skimmed milk powder 1% (w/v), bovine serum albumine (BSA) 1% (w/v) diluted in PBS-T). 5. Columbus Pro plate washer (Tecan, Crailsheim, Germany). 6. E. coli TG1 (Lucigen, Middleton, WI, USA), genotype: [F0 traD36 proAB lacIqZ ΔM15] supE thi-1 Δ(lac-proAB) Δ(mcrB-hsdSM)5(rK - mK -). 7. E. coli XL1-Blue MRF0. 8. 1 mL cuvettes and spectrophotometer 600 nm wavelength. 9. 24-deep well plate. Epitope Mapping by Phage Display 503

10. VorTemp 56 incubator (Labnet, Edison, USA). 11. Trypsin (1 mg/mL stock). 12. M13K07 Helperphage for monovalent display (Agilent). 13. Refrigerated centrifuge for 15 and 50 mL tubes, and plates (Eppendorf, Hamburg, Germany). 14. 2Â TY-T. 15. 2Â TY-GA. 16. 2Â TY-GA agar. 17. 2Â TY-AK. 18. 96-well U-shaped polypropylene plate. 19. Hyperphage for oligovalent display (Progen). 20. Non-related phage (for negative control). 21. PEG-NaCl. 22. 2% MPBS-T (skimmed milk powder 2% (w/v), diluted in PBS-T). 23. 96-well ﬂat-bottom polystyrene ELISA plate. 24. Anti-Fc-speciﬁc HRP-conjugated (Sigma Aldrich, Munchen,€ Germany). 25. Anti-M13 phage (pVIII) HRP-conjugated (GE Healthcare, Munchen,€ Germany). 26. TMB solutions: TMB-A: 50 mM citric acid, 30 mM potassium citrate, pH 4.1; TMB-B: 90% (v/v) ethanol, 10% (v/v) acetone; 10 mM tetramethylbenzidine; 1 mL 30% H2O2; mix 19 parts of TMB-A with 1 part of TMB-B.

27. 1 N H2SO4. 28. ELISA plate reader with 450 nm ﬁlter (Tecan).

3 Methods

3.1 Gene 1. Design primers for the gene of interest depending on the DNA Amplification, source used (see Note 1). Fragmentation, 2. Amplify the gene using polymerase chain reaction in duplicates and End-Repair (Table 1). 3. Run an agarose gel to check the amplification (band size, specificity, etc.). 4. Mix the two duplicate reactions and purify using NucleoSpin Gel and PCR clean-up kit (Macherey-Nagel), eluting the DNA with Milli-Q water twice in different tubes, first with 30 μL, and then with 20 μL. 504 Gustavo Març al Schmidt Garcia Moreira et al.

Table 1 Reagents to be added on the gene amplification PCR

DNA (50 ng/μL plasmid, or 200 ng/μL genome) 1 μL dNTP mix (10 mM each) 1 μL HF buffer 5Â 10 μL Primer forward þ reverse (10 μM each) 2.5 μL þ 2.5 μL ® Phusion DNA polymerase (2 U/μL) 0.5 μL

H2O Milli-Q 32.5 μL Total volume 50 μL Suggested PCR program: 98 C, 30 s þ 98 C, 10 s; Tm,15s;72 C, 30 s/1 kb (30 cycles) þ 72 C 5 min þ 4 C forever

Table 2 Reagents to be added on reaction for DNA-ends repair

Fragmented DNA (ﬁnal amount 0.8–1 μg) X μL 10Â End repair reaction mix 5 μL End Repair Enzyme Mix 2.5 μL

H2O Milli-Q Up to 50 μL

5. Quantify the eluted DNA using NanoDrop and mix 1 μgina total volume of 100 μL of Milli-Q water. ® 6. Fragment the DNA using the Bioruptor Plus Sonication Sys- tem (Diagenode) equipment following the manufacturer’s instructions to obtain fragments with 150 bp. Normally, it is set as: 70 times 30 s sonication with 30 s interval at low power, all at 4 C in water bath (see Note 2). 7. Run a 1.5% agarose gel loading 5 μL of the sample to check the actual size of the fragments (see Note 3). 8. Concentrate the fragments using Amicon Ultra-0.5 mL Cen- trifugal Filters Ultracel-30K (Millipore) following the manufacturer’s instructions. 9. Quantify the DNA using NanoDrop. 10. Repair the ends of the fragment using Fast DNA End Repair kit (Thermo Scientiﬁc) according to the manufacturer’s instructions (Table 2). 11. Incubate the reaction at 20 C for 15 min (do not let it stand longer) and purify using the NucleoSpin Gel and PCR clean-up kit (Macherey-Nagel). Elute in 20 μL Milli-Q water. Epitope Mapping by Phage Display 505

Table 3 Reagents to be added on the linearization of the phagemid

Phagemid (total 5 μg) X μL Buffer CutSmart 10Â (NEB) 2 μL PmeI (10 U/μL, NEB) 1 μL

H2O Milli-Q Up to 20 μL

Table 4 Reagents to be added on the ligation of gene fragments with phagemid

Digested 4-kb phagemid (total 1 μg) X μL Gene fragments 150–500 bp (total 0.75 μg)a Y μL T4 DNA ligase buffer 10Â (Promega) 10 μL T4 DNA ligase (3 U/μL, Promega) 3.5 μL

H2O Milli-Q Up to 100 μL aThe range 150–500 is considered because the described sonication procedure usually results in a smear between two different sizes. In this case, the considered average size of the fragments is 325 bp and, thus, 0.75 μg should be added. However, if the obtained average size is different, it is important to maintain the molar ratio of 1:10 (vector:insert)

3.2 Phagemid- 1. The preparation of the phagemid varies with the kind of phage- Fragment Ligation display method used. In this protocol, it is necessary to use a and Library phagemid that allows the cloning in a blunt end, such Construction pHORF3 (see Note 4), which has PmeI as cloning site. Thus, perform the digestion as described in Table 3. 2. Incubate the reaction for 2 h at 37 C and add 1 μL of calf- intestinal alkalyne phosphatase (10 U/μL, NEB). 3. Incubate for more 1 h at 37 C and purify the reaction using the NucleoSpin Gel and PCR clean-up kit (Macherey-Nagel). Elute in 20 μL Milli-Q water. 4. Perform ligation reaction for 16 h at 16 C (Table 4). 5. Inactivate the ligation for 10 min at 65 C and clean the ® reaction using Amicon Ultra-0.5 mL Centrifugal Filters ® Ultracel -30K (Millipore). For this, add 400 μL of Milli-Q water in the reaction and centrifuge (5 min, 14,000 Â g). Repeat this washing with 400 μL of Milli-Q water more three times before collecting the final volume as instructed by the manufacturer. 6. Mix 15 μL of the purified ligation with 25 μL electrocompetent E. coli TOP10F0 (Thermo Scientific) in a 0.2 mL tube, transfer the volume to a 0.1 mm cuvette, and keep it on ice for 1 min. 506 Gustavo Març al Schmidt Garcia Moreira et al.

7. Perform electroporation for bacteria (1.8 kV; pulse 5.3 ms long) and immediately add 1 mL of SOC medium pre-warmed at 37 C. 8. Transfer the cells to a 1.5 mL tube and incubate at 37 C for 1h. À 9. Take 10 μL of the tube and make tenfold dilutions until 10 5 in 2Â YT. 10. Plate 50 μL of the dilution 10–3 and 100 μL of the dilution 10–5 onto 2Â YT-GA agar 10 cm plates, and grow it overnight at 37 C. 11. Plate the remaining 990 μL of the transformation onto a 24.5 Â 24.5 Â 2.5 cm plate with 2Â YT-GA agar and incubate at 37 C for 16 h. 12. Perform the colony counting on the 10 cm plates (see Note 5). 13. On the 24.5 Â 24.5 Â 2.5 cm plate, add 20 mL of 2Â - YT-glycerol 16% and incubate on a shaker for 10 min. 14. With a Drigalsky spatula, carefully remove the cells from the medium surface. Then, take the liquid containing cells with a serological pipette and distribute 1 mL in each of six cryovials. 15. Immerse the cryovials containing the cells into liquid nitrogen and wait for 5 min. Then, carefully take the tubes with proper protection gloves and store the tubes at À80 C promptly.

3.3 Library Quality 1. From the 10 cm plates used for counting on the previous topic, Control and Packaging take at least 20 colonies to perform a colony PCR. For this PCR, make one tube containing the empty phagemid used for the library construction as a negative control (Table 5, see Note 6). 2. To check the size of each fragment, prepare a 1.5% agarose gel and run the samples at 80 V for 1 h to increase the resolution (see Note 7). 3. Count the number of positives (those above the band of the negative control) expecting to have at least 80% (16/20) of the clones positive (this quality measurement is called “insert rate”). If the number is much below 80%, consider repeating previous steps, mainly the phagemid preparation or ligation. 4. With an insert rate 80%, add 200 mL of 2Â YT-GA in a 500 mL shake ﬂask. Then, take the library stored at À80 C prepared on the previous topic and inoculate 200–500 μL into the medium until OD600 ¼ 0.1. 5. Incubate the shake ﬂask at 37 C, 250 rpm until OD600 0.5. Then, transfer 25 mL (1.25 Â 1010 cells) of the culture to a 50 mL tube and add 2.5 Â 1011 CFU (MOI 1:20) of Hyperphage. Epitope Mapping by Phage Display 507

Table 5 Reagents to be added on the PCR for insert rate calculation

dNTP mix 0.2 μL

MgCl2 25 mM 0.8 μL ® GoTaq Flexi Buffer 5Â 2 μL Primer forward þ reverse (10 mM each) 0.5 μL þ 0.5 μL ® GoTaq DNA polymerase (5 U/μL) 0.05 μL

H2O Milli-Q 5.95 μL Total volume 10 μL Suggested PCR program: 95 C, 5 min þ 95 C, 30 s; Tm,30s;72 C, 1 min/1 kb (30 cycles) þ 72 C 5 min þ 4 C forever

6. Incubate the tube for 30 min at 37 C without shaking, followed by 30 min at 37 C, 250 rpm. 7. Centrifuge the tube 3,220 Â g, 10 min, RT. Then, discard the supernatant, suspend the cells in 10 mL of 2Â YT-AK, and transfer them to a 500 mL shake flask containing 190 mL of the same medium. Incubate the flask at 30 C, 250 rpm for 20–24 h. 8. Transfer the culture to a 500 mL centrifuge tube and centrifuge 10,000 Â g, 10 min, 4 C. Collect the supernatant into another 500 mL centrifuge tube, add 1/5 volume (40 mL) of PEG-NaCl solution and incubate the tube at 4 C on ice overnight. In parallel, inoculate a 100 mL shake flask containing 25 mL of 2Â YT-T with E. coli XL1-Blue MRF’ and incubate at 37 C, 250 rpm overnight. 9. Centrifuge the tube containing the supernatant with PEG-NaCl 10,000 Â g, 1 h, 4 C and discard the supernatant. 10. Suspend the pellet containing phage in 10 mL of prechilled PBS (or PEB) and transfer the volume to a 50 mL centrifuge tube. 11. Centrifuge the suspension 20,000 Â g, 10 min, 4 C and collect the supernatant. 12. Filter the suspension with a 0.45 μm filter and transfer to another 50 mL centrifuge tube. 13. Add 1/5 volume (2 mL) of PEG-NaCl solution and incubate for 30 min on ice, mixing manually every 5 min. 14. Centrifuge the suspension 20,000 Â g, 30 min, 4 C and discard the supernatant. 15. Suspend the pellet in 1 mL of PBS, transfer to a 1.5 mL tube and centrifuge 16,000 Â g, 30 min, 4 C. 508 Gustavo Març al Schmidt Garcia Moreira et al.

16. Transfer the supernatant to a cryovial and store it at 4 C for further use. 17. Take the E. coli XL1-Blue MRF’ culture, make another 25-mL 2Â YT-T culture in a 100 mL shake ﬂask with initial OD600 0.1, and incubate at 37 C, 250 rpm until OD600 0.5. 18. Take 10 μL of the phage suspension and make tenfold dilutions À until 10 9 in PBS. 19. Prepare four 1.5 mL tubes with 50 μLofE. coli XL1-Blue MRF’ cells in each and transfer 10 μL of the last four phage À À dilutions to each tube (these will be dilutions 10 8,10 9, À À 10 10,10 11 on the plate). 20. Incubate the tubes at 37 C for 30 min without shaking. 21. Divide one 2Â YT-GA agar plate into four parts and make three 10 μL droplets of each of the four dilutions on each part. Let the droplets dry under the biological cabinet for 5 min and incubate the plate at 37 C for 16 h. 22. Divide another 2Â YT-GA agar plate into two parts and make spread the remaining volume (30 μL) of the two intermedi- À À ary dilutions (10 9, and 10 10). 23. Count the colonies on countable droplets and calculate the titer as the arithmetic mean of the three droplets and multiply per 6, so the ﬁnal result will be in CFU/mL. This quality measurement is called “library titer” (see Note 8). 24. From the other plate, pick at least 20 colonies and send them for sequencing expecting to have at least 60% (12/20) of the clones with in-frame and correct sequence. This quality measurement is called “in-frame rate” (see Note 9).

3.4 Antigen Panning Panning round 1 1. Coat two wells of a 96-well ELISA plate with 1.5 μgofa purified monoclonal antibody diluted in 150 μL of PBS (recommended wells A1 and B1, called “mAb wells”) (see Note 10). 2. Add 300 μL of Panning Block solution in another two wells (recommended wells A3 and B3, called “block wells”) and incubate the plate at 4 C overnight. 3. In parallel, inoculate 25 mL of 2Â YT in a 100 mL shake flask with E. coli TG1 and incubate overnight at 37 C, 250 rpm. 4. On the next day, mix the single-target library with Panning Block solution to a final volume of 300 μL in a 1.5 mL tube (the final amount of phage should be 1 Â 1010). 5. Take the contents out of the block and mAb wells. In the block wells, add 150 μL of the library (or 5 Â 109/well). In the Epitope Mapping by Phage Display 509

mAb wells, add 300 μL of Panning Block solution and incubate the plate for 30 min at room temperature. 6. Take the blocking solution out of the antibody well and wash three times with 300 μL of PBS-T (using the Columbus Pro plate washer is recommended). 7. Transfer the phage library preincubated in the block well to the mAb well and incubate for 1.5 h at RT. 8. After 1 h, inoculate 300 μL of the E. coli TG1 overnight culture in 25 mL 2Â YT (initial OD600 ¼ 0.08–0.1), incubate at 37 C, 250 rpm for 1.5 h until OD600 0.5, and use on step 12. While the cells are growing, perform steps 9–11. 9. Remove the library from the mAb well and wash the well roughly, i.e., 40 times with 300 μL of PBS-T (using the Colum- bus Pro plate washer is recommended). 10. Elute the binding phage by adding 160 μLof10μg/mL Trypsin diluted in PBS for 30 min at 37 C. 11. Store 60 μL of eluted phage at 4 C in 0.2 mL tubes. 12. In a 24-deep well plate, add 1 mL of E. coli TG1 into two wells. Then add the remaining 100 μL of eluted phage in each well, and incubate for 30 min at 37 C without shaking, followed by 30 min at 37 C under 500 rpm. 13. Centrifuge the plate 2500 Â g for 10 min at RT. Discard the supernatant, add 5 mL of pre-warmed 2Â YT-GA, and incubate at 37 C under 500 rpm for 30 min until OD600 0.5 is reached. Then, add the Helper phage M13K07 (5 Â 1010 total, MOI 1:20) for 30 min at 37 C without shaking. Then, incubate for 30 min at 37 C under 500 rpm. 14. Centrifuge the plate 2500 Â g for 10 min at RT. Remove the supernatant completely (be careful with the pellet). Add 5 mL 2Â YT-AK, suspend the pellet, and incubate at 30 C under 500 rpm overnight. 15. For the next panning round, coat 1 well of the ELISA plate the same way as done on step 1 with both antibody and Panning block solution. Moreover, inoculate E. coli E. coli XL1-Blue MRF’ the same way as described for E. coli TG1 in step 1, but in 2Â YT-T instead. Panning round 2 16. On the next day, centrifuge the 24-well plate (3220 Â g, 10 min, RT), collect, and mix the supernatants in a 15 mL tube. 17. Mix 50 μL of the mixed supernatant from the ﬁrst panning round with 100 μL of Panning block solution. 510 Gustavo Marc¸ al Schmidt Garcia Moreira et al.

18. Take the contents out of the block and mAb wells. In the block wells, add 150 μL of the supernatant mixed with Panning block solution. In the mAb wells, add 300 μL of Panning Block solution, and incubate the plate 30 min at room temperature. 19. Take the blocking solution out of the antibody well and wash three times with 300 μL of PBS-T (using the Columbus Pro plate washer is recommended). 20. Transfer the supernatant in Panning block solution preincubated in the block well to the mAb well and incubate for 1.5 h at RT. 21. After 40 min, inoculate 300 μL of the E. coli XL1-Blue MRF0 overnight culture in 25 mL 2Â YT-T (initial OD600 ¼ 0.08–0.1), incubate at 37 C, 250 rpm for 1.5 h until OD600 0.5, and use on step 27. 22. While the cells are growing, perform steps 23–25. 23. Remove the library from the mAb well and wash the well roughly, i.e., 60 times with 300 μL of PBS-T (using the Columbus Pro plate washer is recommended). 24. Elute the binding phage by adding 160 μLof10μg/mL Trypsin diluted in PBS for 30 min at 37 C. 25. Store all volume of eluted phage at 4 C in 0.2 mL tubes. 26. In a 24-deep well plate, add 1 mL of E. coli XL1-Blue MRF0 at OD600 0.5 into two wells. Then, add the 60 μL of eluted phage from panning round 1 (stored on step 12) in one well, and 100 μL of the eluted phage from panning round 2 (stored in the previous step) in the other well. 27. Incubate for 30 min at 37 C without shaking, followed by 30 min at 37 C under 500 rpm. 28. From each well, make 2 tenfold dilutions by adding 20 μLof the infected E. coli XL1-Blue MRF’ into 180 μLof2Â YT. 29. Spread 25 μL from the ﬁrst dilution, and 50 μL from the second onto 2Â YT-GA agar in 10 cm plates. Store the dilutions at 4 C until the next day (see Note 11). 30. Centrifuge the 24-well plate 2500 Â g for 10 min at RT, discard 1 mL of the supernatant, suspend the pellet on the remaining medium volume, and spread it onto one 10 cm plate with 2Â YT-GA agar as a backup (see Note 12). 31. Incubate all the plates at 37 C overnight, and then store them at 4 C or start the next part directly.

3.5 Monoclonal 1. In a 96-well U-bottom propylene plate, add 150 μL/well of Phage Production 2Â YT-GA. and Screening 2. Use 200 μL pipette tips to pick 92 colonies (46 from each panning round) from the plates described on the last step of Epitope Mapping by Phage Display 511

the previous part. In this same plate, include two wells (H3 and H9) with medium only, one well (H6) with a colony to produce a non-related phage, and one well (H12) with the same colony added on H11. 3. Add a breathable membrane over the plate and incubate at 37 C, 800 rpm, for 6 h (this will be called “Master plate”) (see Note 13). 4. In another 96-well U-shaped propylene plate, add 180 μL/well of 2Â YT-GA, and transfer 20 μL of the previously grown plate to this new one. Store the Master plate at 4 C, and incubate the new one at 37 C, 800 rpm, for 1.5 h. 5. Dilute puriﬁed Hyperphage in 2Â YT to the concentration of 1 Â 1010 CFU/mL, and add 50 μL/well (5 Â 109 CFU/well). 6. Incubate for 30 min at 37 C without shaking, followed by 30 min at 37 C under 800 rpm. 7. Centrifuge the plate 3200 Â g for 10 min at RT, remove the supernatant by inverting the plate very quickly over a discard, and add 190 μL/well of 2Â YT-AK. 8. Incubate the plate at overnight 30 C, 800 rpm. 9. Centrifuge the plate 3200 Â g for 10 min at RT, transfer 150 μL of each supernatant to a new plate, and add 40 μL/ well of PEG-NaCl solution. 10. Incubate the plate 1 h at 4 C, and centrifuge 3200 Â g for 1 h at 4 C. 11. Remove the supernatant by inverting the plate very quickly over a discard, and suspend each pellet in 150 μL of PBS. 12. Shake the plate for 5 min under 500 rpm, and centrifuge 3200 Â g for 10 min at 4 C. 13. On an ELISA plate, add 50 μL/well of PBS (except well H9), and then add 50 μL of the supernatant from the plate centrifuged in the previous step on each well, diluting the phage 1:2. On well H9, add 3 Â 108 cfu of Hyperphage as a negative control (see Note 14). 14. Incubate the ELISA plate overnight at 4 C. 15. Discard the content on the ELISA plate and add 300 μL/well of 2% MPBS-T. 16. Incubate for 30 min at room temperature, and wash the plate three times with 300 μL/well of PBS-T. 17. Dilute the studied mAb to 1 μg/mL in 2% MPBS-T, and add 100 μL/well on each well, except H12. On H12, add 100 μL of 2% MPBS-T only. 18. Incubate for 1 h at room temperature, and wash the plate three times with 300 μL/well of PBS-T. 512 Gustavo Marc¸ al Schmidt Garcia Moreira et al.

19. Dilute an anti-Fc speciﬁc HRP-conjugated antibody (see Note 15) in 2% MPBS-T, and add 100 μL/well on each well, except H12. On H12, add 100 μL of goat anti-M13 (pVIII) HRP-conjugated antibody diluted 1:40,000 in 2% MPBS-T. 20. Incubate for 1 h at room temperature, and wash the plate three times with 300 μL/well of PBS-T. 21. Add 100 μL/well of TMB ELISA developing solution and let it stand for 15 min at room temperature. Then, stop the reaction by adding 100 μL/well of 1 N H2SO4. Acquire the data with an ELISA plate reader at 450 nm, using 620 nm as a reference wavelength (see Note 16).

3.6 Selection 1. Based on the signal obtained in the ELISA, set the maximum and Sequencing signal as 100%. of Positive Hits, 2. Then, classify the clones according to their signal as low and Epitope (10–30%), medium (30–70%), and high (70–100%) reactions. Determination 3. For each panning round, select four clones: one from the “low” group (the lowest signal is recommended); two from the “medium” group (the lowest and highest signals are recommended); and one from the “high” group (the highest signal is recommended). Thus, eight clones will be selected in total (four from panning round 1, and four form panning round 2). 4. Take the Master plate stored on step 4 of the previous part and use it as a source of the selected colonies to send them for sequencing. 5. Analyze the DNA sequences checking for their quality, i.e., if there are out-of-frame sequences or premature stop codons, and discard these sequences (see Note 17). 6. With the remaining sequences, perform their translation with TranSeq tool (EBI, https://www.ebi.ac.uk/Tools/st/emboss_ transeq) to obtain the corresponding amino acid sequence (see Note 18). 7. Align all the amino acid sequences with ClustalOmega tool (EBI, http://www.ebi.ac.uk/Tools/msa/clustalo) and observe for the regions with high identity and similarity. 8. Select a region with no more than 25 amino acids as the ﬁnal epitope (see Note 19).

4 Notes

1. If using a plasmid as source of the gene, it is recommended to design primers annealing to the plasmid, but near the gene. In this way, it is possible to amplify different genes if the same Epitope Mapping by Phage Display 513

plasmid is used. If using a genomic DNA as source of the gene, it is recommended to design primers annealing to the gene. 2. Since this protocol is focused on the epitope mapping, the size of the fragments follows the principle that the smaller the fragment is, closer to the real epitope the result is. In this way, by setting the sonicator to 150 bp, it is expected to have most of the displayed peptides with 50 amino acid, but also some smaller (20 amino acids) and some bigger (100 amino acids). However, it is worth considering that certain structures of the epitope can only be formed by augmenting the size of the displayed peptide. If it occurs, an epitope can only be mapped by building libraries with bigger DNA fragments. In this case, it is recommended to build multiple libraries in parallel, e.g., one with 150 bp fragments, other with 400 bp, and another with 1250 bp. Although this alternative can allow the detection of fragments containing epitopes, it is essential to know which kind of epitope is being searched (linear, conformational, etc.) and determine if this technique will be useful. 3. Usually, the DNA is seen as a smear with a concentrated band on the expected size. However, it can happen that the smear is broader with no concentrated band, or the concentrated band is slightly above the expected size. In these cases, it is accepted to work in a range of up to 500 bp (not more than this). If the smear or a band has >500 bp, the sonication conditions should be optimized. 4. This protocol is written based on the pHORF3 system for ORFeome display [32]. Brieﬂy, this phagemid allows cloning fragments in a blunt PmeI restriction site. In this way, cloned fragments that are in frame will be expressed in fusion with pIII, allowing the peptide corresponding to the cloned fragment to be displayed on the surface of the phage particles. 5. The expected amount of independent clones is 106–108. This amount is actually above the number of clones necessary to cover the gene length. For example, a gene with 1200 bp would need approximately 2102 cloned fragments of 150 bp to completely cover its sequence when walking one nucleotide upstream at a time. It means that a titer of 2.102 Â 103 is enough to cover the gene. This number is obtained by the following formula: N ¼ 2 Â ðÞa À b þ 1 N is the number of cloned fragments (independent clones); a is the size of the gene; and b is the average size of the fragmented DNA. The calculation considers that the fragments can be cloned in two possible orientations in the blunt-end ligation. This is why the number of cloned fragments needed to cover the gene (a À b þ 1) 514 Gustavo Marc¸ al Schmidt Garcia Moreira et al.

is multiplied by 2. This formula is just to guide the researcher in a simple manner. If specific applications are needed, the formula may have to be modified accordingly. 6. The pair of primers used for colony PCR (including negative control) depends on the phagemid system used. However, regardless the phagemid used, it is recommended to have primers annealing on the phagemid backbone, and approximately 130 bp distant from the restriction site used for cloning. In this way, the negative control will have about 250 bp, and all the positive clones should be above this size. 7. This step can sometimes need better resolution. Initially, it is possible to optimize the running and imaging conditions to detect small size differences between the fragments, i.e., increase agarose gel concentration, or running with lower volt- age. Nevertheless, the optimal evaluation of the band size is done with capillary electrophoresis (e.g., using QIAxcel Advanced System, QIAGEN, Hilden, Germany). 8. The expected titer is 1010–1012 per mL. If the titer is much below this value, consider repeating the phage production (steps 4–16). 9. The way the sequencing is performed depends on the workflow of each laboratory. Usually, the plasmids of individual colonies have to be extracted and premixed with a proper primer for sequencing. Otherwise, a colony PCR can be performed and purified prior to sequencing. Anyway, the sequences should be analyzed regarding the reading frame of the cloned fragments, which should be in-frame with the gIII contained in the vector (e.g., frame “þ2” in pHORF3 vector). Sequences containing only parts of the vector, if present, should be counted as negative. After the analysis, the number of positives (i.e., in-frame sequences) should be 60%. If this number is much below 60%, considering repeating the sequencing with more 20 clones. If the bad results persist, probably the whole procedure for building libraries should be repeated. 10. Normally, coating mAbs directly onto the ELISA plate surface is not a problem. Nevertheless, some antibodies can have their activity considerably reduced or even eliminated after attaching onto the plastic surface. If this is the case, coat the plate for 1 h at RT with 100 μL/well of an antibody against the Fc part of the studied mAb diluted in PBS to 2–4 μg/mL, and block by adding 300 μL/well of 2% MPBS-T for 30 min prior to adding the mAb. 11. Considering that the number of phage containing the epitope, or epitopes, of the studied antibody can be highly variable (especially in the panning round 1), there is no rule for the Epitope Mapping by Phage Display 515

number of phage eluted on the first and second rounds. In this way, the number of colonies obtained after diluting and spreading the infected bacteria on agar plates is variable. The protocol suggests a volume that is more likely to give the researcher enough colonies to start a screening, which is at least 46 isolated colonies for each panning round. However, if the plates have too few or too much colonies, the spreading step should be repeated using the dilutions stored at 4 C on this same step. In this case, the spread volume should be adjusted accordingly. 12. The backup plates can be stored at 4 C until the end of the screening. If the epitope mapping procedure is considered successful, the backup plates contain most of the selected clones. Thus, it is recommended to store these clones long- term, considering that further studies can become interesting as the researcher’s work goes on. For this, add 5 mL of 2Â YT þ 16% glycerol over the plate and use a Drigalski spatula to scrape the cells from the plate. Then, take 1 mL of the scraped cells, add into cryovials (make at least two cryovials), and store at À80 C promptly. 13. The growth of colonies in the 96-well U-shaped polypropylene plate can also be done overnight. The step described in the protocol uses 6 h of incubation to reduce the time of the procedure by 1 day, but it is optional to the researcher. 14. During the monoclonal phage production, although most of the E. coli cells contents (e.g., cell debris, proteins, etc.) are removed by precipitation and centrifugation, it is possible that it is still present in the final preparation. Usually, the high specificity of the studied mAb allows ignoring these “undesired” protein contents. However, if high levels of background are noticed after the screening (i.e., high reactions with the non-related phage used as negative control), it is recommended to immobilize the phage using anti-phage antibody. In this case, coat the plate for 1 h at RT with 100 μL/well of goat anti-M13 (pVIII) antibody (Sigma Aldrich) diluted in PBS to 2–4 μg/mL. Then, block by adding 300 μL/well of 2% MPBS-T for 30 min prior to adding the phage. 15. The specifications of the used anti-Fc-specific antibody depend on the properties of the mAb used in the procedure. Usually, the studied mAbs are IgG and have human or mouse Fc parts, allowing the use of goat anti-mouse or human IgG Fc HRP-conjugated antibodies. If the studied mAb has another isotype (IgA, IgE, IgM, etc.), or if the Fc part is from another species (rat, rabbit, etc.), the researcher should choose the best HRP-conjugated option for the work. 516 Gustavo Març al Schmidt Garcia Moreira et al.

16. As mentioned on the introduction of this chapter, the complete applications of this technique are not completely defined. At the moment, it is known that linear epitopes have more chance to be mapped compared to conformational ones. In this way, if the result of the screening ELISA shows negative (even in the repetition), it is more likely that the studied mAb recognizes a conformational epitope that cannot be mapped by this procedure. However, it is worth trying the alternative described on Note 2 before discarding the use of this method. 17. Usually, it is expected that every sequence is useful for the analysis. In some cases, however, few sequences show problems and have to be discarded. To obtain a good analysis for the epitope mapping, make sure that most of the discarded sequences are from the “low” and “medium” groups. Since the “high” group is the one that will give the most refined information, especially those from the second panning round, it is recommended to have these samples in the analysis. If for some reason the “high” group sequences showed to be bad, try repeating the sequencing, or even picking other colonies from the Master plate for a new sequencing. If most of the sequences show to be bad, try to repeat the sequencing or pick new colonies too. 18. In this step, it is important to observe the size of the obtained amino acid sequences. The smaller the fragments are, more refined the epitope mapping will be. Thus, having at least one sequence with not more than 50 amino acids (one with <25 amino acids is optimal) is helpful for the analysis. In this way, the next step, which involves the alignment of obtained amino acid sequences, will give better information about the epitope. If only big sequences (with >100 amino acids) are obtained, it is suggested to select new colonies for sequencing from the Master plate. 19. For the definition of the final epitope, it is better if the analyzed sequences contain at least one with <25 amino acids. In this way, the confirmation of the epitope is simplified. If it is not the case, the manual selection of a 25-amino acid sequence has to be validated by another method (e.g., ELISA against a synthetic peptide, or using peptide membranes that cover the selected sequence). The epitope size for most of the antibodies is from 8 to 22 amino acids, thus, the definition of the epitope relies on the data set obtained by the presented technique. If there is a need to refine the obtained data, other techniques should be performed either by using phage display or not. Epitope Mapping by Phage Display 517

Acknowledgments

We thank Dr. Jonas Zantow for critical comments and suggestions on the chapter. We also thank CNPq for providing the scholarship of GMSGM (process 204693/2014-4).

References

1. Heesters BA, van der Poel CE, Das A, Carroll the pathogenic mechanism and a possible ther- MC (2016) Antigen presentation to B cells. apy for autoimmune TTP syndrome. Proc Natl Trends Immunol. https://doi.org/10.1016/ Acad Sci U S A 112:9620–9625. https://doi. j.it.2016.10.003 org/10.1073/pnas.1512561112 2. Krueger A, Zie˛tara N, Łyszkiewicz M (2016) T 12. Pin E, Henjes F, Hong M-G et al (2016) Iden- Cell development by the numbers. Trends tification of a novel autoimmune peptide epi- Immunol 38(2):128–139. https://doi.org/ tope of prostein in prostate cancer. J Proteome 10.1016/j.it.2016.10.007 Res 16(1):204–216. https://doi.org/10. 3. Soliman C, Yuriev E, Ramsland PA (2016) 1021/acs.jproteome.6b00620 Antibody recognition of aberrant glycosylation 13. Lanzavecchia A, Fruhwirth€ A, Perez L, Corti D on the surface of cancer cells. Curr Opin Struct (2016) Antibody-guided vaccine design: iden- Biol 44:1–8. https://doi.org/10.1016/j.sbi. tification of protective epitopes. Curr Opin 2016.10.009 Immunol 41:62–67. https://doi.org/10. 4. Rock KL, Reits E, Neefjes J (2016) Present 1016/j.coi.2016.06.001 yourself! By MHC class I and MHC class II 14. Aghebati-Maleki L, Bakhshinejad B, Baradaran molecules. Trends Immunol 37(11):724–737. B et al (2016) Phage display as a promising https://doi.org/10.1016/j.it.2016.08.010 approach for vaccine development. J Biomed 5. Gunn BM, Alter G (2016) Modulating anti- Sci 23:66. https://doi.org/10.1186/s12929- body functionality in infectious disease and 016-0285-9 vaccination. Trends Mol Med 22 15. Miethe S, Rasetti-Escargueil C, Liu Y et al (11):969–982. https://doi.org/10.1016/j. (2014) Development of neutralizing scFv-Fc molmed.2016.09.002 against botulinum neurotoxin A light chain 6. Sela-Culang I, Kunik V, Ofran Y (2013) The from a macaque immune library. MAbs structural basis of antibody-antigen recogni- 6:446–459. https://doi.org/10.4161/mabs. tion. Front Immunol 4:302. https://doi.org/ 27773 10.3389/fimmu.2013.00302 16. Mendonça M, Moreira GMSG, ConceiçaõFR 7. Kaur H, Salunke DM (2015) Antibody promis- et al (2016) Fructose 1,6-bisphosphate aldol- cuity: understanding the paradigm shift in anti- ase, a novel immunogenic surface protein on gen recognition. IUBMB Life 67:498–505. Listeria species. PLoS One 11:e0160544. https://doi.org/10.1002/iub.1397 https://doi.org/10.1371/journal.pone. 8. Kringelum JV, Nielsen M, Padkjær SB, Lund O 0160544 (2013) Structural analysis of B-cell epitopes in 17. Jiang X, Totrov M, Li W et al (2016) Rationally antibody:protein complexes. Mol Immunol designed immunogens targeting HIV-1 gp120 53:24–34. https://doi.org/10.1016/j. V1V2 induce distinct conformation-specific molimm.2012.06.001 antibody responses in rabbits. J Virol 90 9. Gupta S, Ansari HR, Gautam A et al (2013) (24):11007–11019. https://doi.org/10. Identification of B-cell epitopes in an antigen 1128/JVI.01409-16 for inducing specific class of antibodies. Biol 18. He W, Tan GS, Mullarkey CE et al (2016) Direct 8:27. https://doi.org/10.1186/1745- Epitope specificity plays a critical role in regu- 6150-8-27 lating antibody-dependent cell-mediated cyto- 10. Cherryholmes GA, Stanton SE, Disis ML toxicity against influenza A virus. Proc Natl (2015) Current methods of epitope identifica- Acad Sci 113:11931–11936. https://doi.org/ tion for cancer vaccine design. Vaccine 10.1073/pnas.1609316113 33:7408–7414. https://doi.org/10.1016/j. 19. Benjamin P (1996) Site-directed mutagenesis vaccine.2015.06.116 in epitope mapping. Methods 9:508–515 11. Casina VC, Hu W, Mao J-H et al (2015) High- 20. Davidson E, Doranz BJ (2014) A high- resolution epitope mapping by HX MS reveals throughput shotgun mutagenesis approach to 518 Gustavo Març al Schmidt Garcia Moreira et al.

mapping B-cell antibody epitopes. Immunol- 31. Wang L, Hu Y, Li W et al (2016) Identification ogy 143:13–20. https://doi.org/10.1111/ of a peptide specifically targeting ovarian cancer imm.12323 by the screening of a phage display peptide 21. Katz C, Levy-Beladev L, Rotem-Bamberger S library. Oncol Lett 11:4022–4026. https:// et al (2011) Studying protein-protein interac- doi.org/10.3892/ol.2016.4549 tions using peptide arrays. Chem Soc Rev 32. Kugler€ J, Zantow J, Meyer T, Hust M (2013) 40:2131–2145. https://doi.org/10.1039/ Oligopeptide m13 phage display in pathogen c0cs00029a research. Virus 5:2531–2545. https://doi. 22. Augustin T, Cehlar O, Skrabana R et al (2015) org/10.3390/v5102531 Unravelling viral camouflage: approaches to 33. Hust M, Meysing M, Schirrmann T et al the study and characterization of conforma- (2006) Enrichment of open reading frames tional epitopes. Acta Virol 59:103–116. presented on bacteriophage M13 using https://doi.org/10.4149/av_2015_02_103 hyperphage. BioTechniques 41:335–342 23. Malito E, Carfi A, Bottomley MJ (2015) Pro- 34. Zantow J, Just S, Lagkouvardos I et al (2016) tein crystallography in vaccine research and Mining gut microbiome oligopeptides by func- development. Int J Mol Sci 16:13106–13140. tional metaproteome display. Sci Rep 6:34337. https://doi.org/10.3390/ijms160613106 https://doi.org/10.1038/srep34337 24. Gershoni JM, Roitburd-Berman A, Siman-Tov 35. Wu C-H, Liu I-J, Lu R-M, Wu H-C (2016) DD et al (2007) Epitope mapping: the first step Advancement and applications of peptide in developing epitope-based vaccines. Bio- phage display technology in biomedical sci- Drugs 21:145–156 ence. J Biomed Sci 23:8. https://doi.org/10. 25. Gallagher ES, Hudgens JW (2016) Mapping 1186/s12929-016-0223-x protein-ligand interactions with proteolytic 36. Rojas G, Tundidor Y, Infante YC (2014) High fragmentation, hydrogen/deuterium throughput functional epitope mapping: revi- exchange-mass spectrometry. Methods Enzy- siting phage display platform to scan target mol 566:357–404. https://doi.org/10. antigen surface. MAbs 6:1368–1378. https:// 1016/bs.mie.2015.08.010 doi.org/10.4161/mabs.36144 26. Pra˛dzin´ska M, Behrendt I, Astorga-Wells J et al 37. Rowley MJ, O’Connor K, Wijeyewickrema L (2016) Application of amide hydrogen/deute- (2004) Phage display for epitope determina- rium exchange mass spectrometry for epitope tion: a paradigm for identifying receptor-ligand mapping in human cystatin C. Amino Acids 48 interactions. Biotechnol Annu Rev (12):2809–2820. https://doi.org/10.1007/ 10:151–188. https://doi.org/10.1016/ s00726-016-2316-y S1387-2656(04)10006-9 27. Schirrmann T, Meyer T, Schutte€ M et al (2011) 38. Luzar J, Sˇtrukelj B, Lunder M (2016) Phage Phage display for the generation of antibodies display peptide libraries in molecular allergol- for proteome research, diagnostics and therapy. ogy: from epitope mapping to mimotope- Molecules 16:412–426. https://doi.org/10. based immunotherapy. Allergy 3390/molecules16010412 71:1526–1532. https://doi.org/10.1111/all. 28. Kuhn P, Fuhner€ V, Unkauf T et al (2016) 12965 Recombinant antibodies for diagnostics and 39. Pande J, Szewczyk MM, Grover AK (2010) therapy against pathogens and toxins generated Phage display: concept, innovations, applica- by phage display. Proteomics Clin Appl tions and future. Biotechnol Adv 10:922–948. https://doi.org/10.1002/prca. 28:849–858. https://doi.org/10.1016/j.bio 201600002 techadv.2010.07.004 29. Frenzel A, Schirrmann T, Hust M (2016) 40. Cariccio VL, Domina M, Benfatto S et al Phage display-derived human antibodies in (2016) Phage display revisited: epitope clinical development and therapy. MAbs mapping of a monoclonal antibody directed 8:1177–1194. https://doi.org/10.1080/ against Neisseria meningitidis adhesin A using 19420862.2016.1212149 the PROFILER technology. MAbs 8:741–750. 30. Reverdatto S, Burz DS, Shekhtman A (2015) https://doi.org/10.1080/19420862.2016. Peptide aptamers: development and applica- 1158371 tions. Curr Top Med Chem 15:1082–1101 Chapter 29

Metasecretome Phage Display

Milica Ciric, Filomena Ng, Jasna Rakonjac, and Dragana Gagic

Abstract

Metasecretome is a collection of cell-surface and secreted proteins that mediate interactions between microbial communities and their environment. These include adhesins, enzymes, surface structures such as pili or flagella, vaccine targets or proteins responsible for immune evasion. Traditional approaches to exploring matasecretome of complex microbial communities via cultivation of microorganisms and screening of individual strains fail to sample extraordinary diversity in these communities, since only a limited fraction of microorganisms are represented by cultures. Advances in culture-independent sequence analysis methods, collectively referred to as metagenomics, offer an alternative approach that enables the direct analysis of collective microbial genomes (metagenome) recovered from environmental samples. This protocol describes a method, metasecretome phage display, which selectively displays the metasecretome portion of the metagenome. The metasecretome library can then be used for two purposes: (1) to sequence the entire metasecretome (using PacBio technology); (2) to identify metasecretome proteins that have a specific function of interest by affinity-screening (bio-panning) using a variety of methods described in other chapters of this volume.

Key words Phage display, Metagenome, Metasecretome, Secreted proteins, Next-generation sequencing

1 Introduction

Microbial secretome (a collection of secreted, surface and integral membrane proteins) comprises a wide range of proteins that mediate interactions with the environment, including membrane- anchored or surface-attached receptors, adhesins, transporters, proteins that form complex cell surface structures such as pili or secreted enzymes, toxins, and virulence factors. In bacteria that colonize the human organism, secreted proteins mediate attachment to the host, destruction of the host tissue, or interference with the immune response [1–3]. In pathogenic bacteria, variation of a surface protein between strains of a species indicates its role in evading the immune response [4–7]; conversely, conserved surface proteins that are capable of inducing a protective immune response are sought-after as vaccine candidates [8].

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_29, © Springer Science+Business Media LLC 2018 519 520 Milica Ciric et al.

Secretome proteins are traditionally studied in vitro using biochemical approaches (2D gel electrophoresis and/or liquid chromatography coupled with mass spectrometric methods [9]), and in silico using bioinformatic tools. Surface display screening methods and reporter fusion systems [10–15], as well as phage-display-based systems [16–19], have also been used for screening, identifying, and characterizing secretome proteins. While biochemical methods allow for elucidation and direct functional characterization of identified secretome proteins, they are usually technically demanding and limited only to cultivable cells. Likewise, in silico analysis requires complete sequenced genomes or deep-sequenced metagenomes for the prediction of secretome proteins. Numerous algorithms for the prediction of secretome proteins have been developed and are mostly based on specific conserved features of secretome proteins (signal sequences and transmembrane α-helices). The accuracy of these methods, however, varies, and is highly dependent on accuracy of the genomic annotation. A culture-independent approach that combines functional and bioinformatics methods for the identification of secretome proteins within microbial metagenomes, metasecretome phage display, will be presented here [20]. This approach is based on a method for direct selection, expression, and display of the bacterial secretome published by Jankovic et al. in 2007 [21]. The principle underlying all display methods, including the secretome-selective phage display, is the physical link of a polypeptide and its coding sequence. A typical phage-display system consists of two components: phagemid vector and a helper phage [22, 23]. The phagemid vectors most commonly encode the carboxy-terminal domain of pIII, preceded by a signal sequence. Inserts are placed between the signal sequence and mature portion of pIII. If an insert is translationally in-frame with both the signal sequence and the mature portion of pIII, then the encoded protein will be displayed on the surface of the phage. In the secretome- selective phage-display system a phagemid vector, pDJ01, contains a pIII C-domain cloning cassette from which the signal sequence was deleted (Fig. 1)[21]. In addition, phagemid pDJ01 contains a phage-infection-inducible promoter psp [24] which ensures very low level of background expression; thereby decreasing the potential toxic effects that could be introduced by foreign protein fusions to pIII. Other features that are downstream of the psp promoter include: Shine-Dalgarno site, multiple cloning cassette, and coding sequence for the C-domain of pIII. The helper phage component of a phage-display system is normally used to provide the f1 replication protein pII that mediates the rolling circle replication of the phagemid vector from the f1 origin, resulting in a single-stranded DNA (ssDNA) genome that is packaged into the virion [25]. The helper phage also provides other phage-encoded proteins essential for packaging of the phagemid ssDNA into the virion, to form Metasecretome Phage Display 521

Fig. 1 Schematic overview of the bacterial secretome-selective phage-display system. The display cassette of the secretome-selective phagemid vector pDJ01 contains: ﬁlamentous phage infection-inducible promoter (psp), ribosome-binding site (RBS), multiple cloning site (MCS), the sequence encoding a common peptide tag, c-myc (tag) followed by a single-amber stop codon (amber) and the sequence encoding C-terminal domain of phage protein pIII (C-gIII), which is used as a display platform. This vector does not have a signal sequence (ss). Infection of E. coli cells harboring recombinant secretome-selective vectors that contain cloned inserts without (A) or with (B) native ss with a helper phage (VCSM13d3), from which the the entire pIII coding sequence was deleted, results in the generation of incomplete recombinant virions (unstable phagemid particles) without the pIII cap at one end (A) or complete virions with recombinant fusion proteins providing a cap (B)

phagemid particles (PPs). However, in the secretome-selective system, the entire coding sequence for pIII (gIII) is deleted from the helper phage [26]. Consequently, the only pIII protein produced in the system is the phagemid vector-encoded pIII that lacks a signal sequence, which in turn is necessary for correct targeting of pIII to the inner membrane and incorporation into the virion [23]. More- over, assembly of pIII into the virion is required to complete the 522 Milica Ciric et al.

Fig. 2 Overview of the metasecretome library construction and selection. (A) Shotgun metagenomic library construction. Some metagenomic inserts contain endogenous signal sequences, represented by red ovals. (B) Shotgun metagenomic library infection with the gIII-deleted helper phage VCSM13d3 (ΔgIII helper phage). (C) Metagenomic phage-display library contains a mix of virions capped by insert-pIII fusion proteins (signal sequence-positive clones) that are resistant to sarkosyl (SarkosylR) and uncapped virions (signal sequence- negative clones) that are sensitive to sarkosyl (SarkosylS). Sarkosyl resistance is used as a basis for selection. (D) Single-stranded DNA (ssDNA) purified from SarkosylR virions after the selection can be used for the identification of metasecretome by next-generation sequencing (NGS) (E) or to obtain the metasecretome plasmid library which can be used in downstream application such as affinity screening (F)

phage assembly. When pIII is absent, virions either stay associated with the host cells as long filaments composed of multiple sequentially packaged genomes, or are broken off by mechanical shearing. Protein pIII is required for the formation of the stabilizing cap structure at the terminus of the virion; hence, the broken-off pIII- deficient virions are structurally unstable and are easily disassembled by sarkosyl, a detergent to which pIII-containing virions are resistant [27, 28]. Therefore, in secretome-selective phage display, only DNA inserts encoding signal sequences can be assembled into stable recombinant phagemid particles (Fig. 2). Traditionally, after the library construction, a phage-display library can be screened for desirable binding properties using an affinity selection procedure, usually referred to as panning or biopanning [29]. During panning, recombinant phage particles are exposed to ligands of interest to selectively capture a phage displaying the binding peptide. Phage-display systems are highly flexible, and affinity selection can be performed against proteins as well as Metasecretome Phage Display 523 other immobilized inorganic [30, 31] and organic molecules [32, 33], or cells in vitro [34] and in vivo [35]. Through successive rounds of binding, washing, elution, and amplification, the original highly diverse phage population (i.e., phage libraries displaying up to 1010 peptide or protein variants) is rapidly enriched for the phage library clones with specificity for the binding target molecule. Due to some unspecific binding occurring between phage and components of the system used for ligand immobilization, at least three rounds of panning (typically 3–5) are needed to eliminate “background” nonspecific binding and enrich monoclonal recombinant phage populations with the desired ligand specificity. Dis- played proteins can be identified by sequencing the inserted DNA encoding the displayed peptide, and can be easily purified and subsequently functionally analyzed. DNA inserts of the individual candidate binders enriched through panning must be sequenced in order to determine the corresponding peptide sequence. Next-generation sequencing (NGS) technologies have been increasingly applied to capture the diversity of phage binding variants enriched after one or two rounds of panning on a ligand of interest [36–40]. NGS has been recently combined with metasecretome-selective phage display for the use at a metagenomic scale, to identify secretome proteins from a microbial community [20]. The information on secretome proteins’ coding sequences in the metasecretome protein database could also be used for transcriptome analysis to examine the expression of the secretome genes in microbial communities using microarrays or nanostring probing methods. Among the NGS technologies that are currently being used to characterize microbial communities are Illumina (Illumina, USA) and PacBio (Pacific Biosciences, USA). In PacBio sequencing, similar to other NGS methods, the DNA template must be within a suitable size range prior to sequencing [41]. This technology produces much longer reads (average of 2000 nt) in comparison to Illumina. As metasecretome inserts range from 0.4 to 4 kb, PacBio sequencing presents an ideal platform for sequencing the metasecretome libraries. The following protocol describes methodology for constructing metasecretome phage-display libraries and indicates potential downstream applications that can be adopted according to the requirements of the project. 524 Milica Ciric et al.

2 Materials

2.1 Preparation 1. DNA shearing buffer [55 mM Tris–HCl (pH 8.0), 15 mM of Metagenomic DNA MgCl2, 25% glycerol]. Inserts 2. Disposable nebulizers (Unomedical Inc., USA). and Secretome- 3. N gas cylinder with ﬁne tuning gauge (10 psi). Selective Phagemids 2 4. Centrifuge with swinging bucket rotor and buckets. 5. Vivaspin sample concentrator (100 kDa cutoff; GE Healthcare Lifesciences, USA) (see Note 1). 6. DNA end-repair and 50-phosphorylation enzymes cocktail (T4 DNA Polymerase; Roche Applied Science, Germany), Kle- now Enzyme (Roche Applied Science, Germany), USB Opti- KinaseTM (Afﬁmetryx, USA). 7. Phenol:Chloroform:Isoamyl alcohol mixture (ratio 25:24:1). 8. Ethanol [100% (v/v) and 70% (v/v)]. 9. Tris–HCl, 10 mM (pH 8.0). 10. pDJ01 phagemid [21]. 11. Restriction endonuclease SmaI. 12. rAPid Alkaline Phosphatase (Roche Applied Science, Germany). 13. QIAEX II Gel Extraction Kit (Qiagen, Germany).

2.2 Construction 1. T4 ligase (Roche Applied Science, Germany). of Shotgun 2. Escherichia coli TG1, genotype: supE thi-1 Δ(lac-proAB) À À Metagenomic Library Δ(mcrBhsdSM)5 (rK mK )[F0 traD36 proABlacIqZΔM15] (Lucigen, USA). ® 3. Gene Pulser II Electroporation System and 0.1 cm gap cuvettes (BioRad, USA), (conditions: 1.8 kV, 25 μF, 200 Ω). 4. SOC media [0.5% (w/v) yeast extract, 2% (w/v) tryptone, 10 mM NaCl, 2.5 mM KCl, 10 mM MgCl2, 10 mM MgSO4, 20 mM glucose]. 5. 2Â YT media pH 7.0 [1.6% (w/v) tryptone, 1% (w/v) yeast extract, 0.5% (w/v) NaCl].

6. 2Â YT Cm25 pH 7.0 (2Â YT, containing 25 μg/ml chloramphenicol). 7. Helper phage VCSM13d3 [26]. 8. 2Â YTagar [2Â YT, containing 1.5% (w/v) bacteriological agar]. 9. 2Â YT soft agar [2Â YT, containing 0.6% (w/v) molecular grade agarose]. 10. Cm double-layer selective plates (see Note 2)[42]. 11. Dimethylsulfoxide (DMSO). Metasecretome Phage Display 525

Table 1 Primers for PCR and sequencing

Primer name Sequence (50–30)

pspF03 ATGTTGCTGTTGATTCTTCA pspR03 TGCCTTTAGCGTCAGACTGTAGC

2.3 Selection 1. Sarkosyl (Sodium lauroyl sarcosinate). of Metasecretome 2. DNase I (Roche Applied Science, Germany). Phage-Display Library 3. MgCl . and Isolation 2 of Single- 4. EDTA. Stranded DNA 5. PVDF membranes, 0.45 μm (Millipore, USA). 6. The PEG/NaCl phage precipitation solution, 6Â stock sterilized by autoclaving [30% (w/v) polyethylene glycol (PEG) 8000, 3 M NaCl]. 7. TN buffer [10 mM Tris–HCl (pH 7.6), 50 mM NaCl]. ® 8. E.Z.N.A. M13 DNA Kit (Omega Bio-tek, USA). 9. Phage disassembly SDS-containing buffer, 4Â concentrated (4% SDS, 4Â TAE, 20% glycerol, 1% bromophenol blue). 10. Plasmid Miniprep Kit.

2.4 Preparation 1. Oligonucleotide primers (see Table 1; primers are unique for of Metasecretome pDJ01 phagemid). ® Phage-Display Library 2. PrimeSTAR Max DNA Polymerase premix (Takara Bio Inc., ® Inserts for PacBio Japan). Premix (2Â) contains the PrimeSTAR Max DNA Sequencing Polymerase, reaction buffer, 2 mM Mg2+ and dNTP mixture. 3. PCR cycler. 4. NucleoMag NGS Clean-up and Size Select magnetic beads (Macherey-Nagel GmbH & Co., Germany).

3 Methods

3.1 Construction 1. Mix 10–20 μg of high molecular weight metagenomic DNA of Microbial Shotgun with 1.5 ml shearing buffer, add the mixture to disposable Metagenomic nebulizer and mechanically shear by subjecting the sample to (Primary) Library a pressure of 10 psi for 1 min (see Note 3). 2. Spin the nebulizer in a centrifuge with a swinging bucket rotor at 300 Â g for 3 min, to collect the dispersed liquid (see Note 4). 3. Size-fractionate and concentrate the sheered DNA using a Vivaspin sample concentrator (100 kDa cutoff) (see Note 1). 526 Milica Ciric et al.

4. Repair and 50-phosphorylate the termini of sheared and frac- tionated DNA fragments using an enzyme cocktail containing T4 DNA Polymerase, Klenow Enzyme and USB OptiKinase™ (see Note 5). 5. Purify the repaired DNA by phenol:chloroform:isoamyl alcohol (volume ratio 25:24:1) extraction and precipitate using 100% ice-cold ethanol, followed by a wash with 70% ice-cold ethanol. Resuspend DNA pellet in 80 μl of 10 mM Tris–HCl (pH 8.0). 6. Linearize vector pDJ01 by digesting with the SmaI restriction endonuclease and dephosphorylate the 50 ends using rAPid Alkaline Phosphatase. 7. Purify the linearized and dephosphorylated vector from the remaining uncut vector by preparative agarose gel electrophoresis; extract DNA from the gel slice using a QIAEX II Gel Extraction Kit (Qiagen). 8. Perform ligation trials to optimize conditions for library construction. Set up a number of test ligations using a series of vector to insert DNA molar ratios (see Note 6). 9. Purify test ligations using phenol:chloroform extraction and ethanol precipitation as described above. Resuspend DNA in 50 μl 10 mM Tris–HCl pH 8.0. 10. Transform into E. coli TG1 competent cells by electroporation, and determine the number of transformants. The ratio of insert to vector that generates the largest number of transformants should be used for the construction of the library. 11. Set up the library ligation using a total of 10 μg of the end-repaired metagenomic DNA fragments and linearized pDJ01 (at the optimal ratio as determined in the previous step). Set up multiple ligation reactions, each containing up to 0.5 μg DNA and T4 ligase in the ligation buffer; incubate at 16 C overnight. Mix the contents of multiple ligation tubes before puriﬁcation. 12. Purify the ligated DNA (as in step 9) and dissolve in 50 μl 10 mM Tris–HCl pH 8.0. 13. Transform high-competency E. coli TG1 electrocompetent cells (>109 CFU/μg DNA) with the ligated and puriﬁed DNA using electroporation. Use no more than 1 μg DNA per electroporation cuvette. After electroporation, add 950 μl of SOC medium to each cuvette, transfer to a microfuge tube, and allow the cells to recover for 1 h at 37 C. 14. Each of the transformations should be processed separately throughout the whole metasecretome selection procedure and the template preparation for next-generation sequencing. Metasecretome Phage Display 527

15. To estimate the primary shotgun library size, titer the number of transformants by plating 10 μl of serial dilutions from each transformation on 2Â YT Cm25 plates. Mix the remaining portion of each transformation with 9 ml of 2Â YT Cm25 broth and incubate for 8 h at 37 C with aeration (180 rpm) to amplify the library in a plasmid form and thereby generate a master shotgun plasmid metagenomic library in pDJ01. 16. Use 1 ml of each ampliﬁed master shotgun metagenomic library aliquots (each derived from a separate transformation) for the production of phagemid particles (PPs). Freeze the remainder at À80 C in 7% DMSO (see Note 7).

3.2 Phagemid To minimize dominance of PPs derived from fast-growing recom- Particle Production binant phagemids, PPs derived from the amplified master metage- Using ΔgIII Helper nomic library are produced using a plate method (instead of growth Phage in a liquid medium) [21]. 1. Mix each of the 1 ml aliquots of amplified master metagenomics library (from Subheading 3.1, step 16) with 100 ml of 2Â YTCm25 media and grow until exponential phase. 2. Infect the cells in the exponential phase of growth [(optical density of a sample measured at a wavelength of 600 nm (OD600nm) ~ 0.2)] with the VCSM13d3 helper phage at a multiplicity of infection (m.o.i.) of 50 phage per cell and incubate for 1 h at 37 C without agitation. 3. Pellet infected cells by centrifugation at 3200 Â g for 10 min at room temperature to remove the remaining unabsorbed helper phage. 4. Resuspend the pellet from each batch of infected cells in 1 ml of 2Â YT, mix quickly with 10 ml of 2Â YT soft agar, and pour over four Cm double-layer selective plates (~ 2.5 ml per plate; see Note 2). Incubate the plates overnight at 37 C. 5. Extract PPs from each plate by adding 5 ml of 2Â YT media on the top of the soft agar surface and incubating with rotary agitation for 1 h at room temperature. 6. Combine PPs from the four plates derived from the same amplified library aliquot, purify and concentrate using PEG precipitation (see Note 8).

3.3 Selection The selection procedure combines sarkosyl treatment to release of Metasecretome DNA from sarkosyl-sensitive PPs, followed by the removal of Phage-Display Library released DNA using DNase I, then de-activation of DNase I and Isolation of ssDNA using EDTA and puriﬁcation of DNA from the remaining PPs containing the metasecretome-encoding recombinant phagemids. Each step of selection is monitored using agarose gel 528 Milica Ciric et al.

electrophoresis of native and disassembled PPs to ensure that both sarkosyl and DNase I treatments were successful (see Note 9). 1. Add 1/10 volume of 1% sarkosyl to each aliquot of puriﬁed PPs and incubate for 10 min at room temperature to disassemble structurally unstable PPs, derived from the recombinant library clones that do not belong to metasecretome, and to release recombinant DNA. PPs of the metasecretome library clones are resistant to sarkosyl and remain intact. Remove 20 μl and store in a sterile tube.

2. To each aliquot of sarkosyl-treated PPs add MgCl2 to a final concentration of 5 mM, followed by DNase I (100 μg/ml). Incubate for 1 h at room temperature, followed by the addition of EDTA (to a final concentration of 25 mM) to inactivate DNase I. 3. Precipitate the sarkosyl-resistant PPs in PEG/NaCl buffer (see Note 8) and resuspend in 10 mM Tris–HCl (pH 7.6). Remove 20 μl and store in a sterile tube. 4. Incubate purified PPs at 75 C for 10 min to ensure the complete inactivation of DNase I prior to the extraction of ssDNA from sarkosyl-resistant PPs. Remove 20 μl and store in a sterile tube. ® 5. Purify ssDNA from sarkosyl-resistant PPs using E.Z.N.A. M13 DNA Kit according to the manufacturer’s recommendations (see Note 10). Remove 20 μl and store in a sterile tube. 6. Analyze the samples from each step by agarose gel electrophoresis (see Fig. 3) The ssDNA isolated from PPs after secretome selection represents the metasecretome library. This library can be used for two main applications, metasecretome sequencing and bio-panning. PacBio sequencing platform can sequence long amplicons (up to 10 kb) without additional amplification or shearing. Given that the insert size in metasecretome library is up to 4 kb, this method is much more suitable than other next-generation sequencing platforms (e.g., Illumina), where the sequence read lengths are below 500 nt and the required template size is 600–800 nt. The PacBio sequencing can be complemented by Illumina sequencing to minimize the sequencing error rate observed in PacBio (<10%). If so, the inserts must be randomly sheared to the required template size range [43]. The non-infectious PPs from the metasecretome selection experiment (Subheading 3.3, step 4) can be subjected to a single round of panning followed by ssDNA purification (Subheading 3.3, step 5) and subjected to PacBio sequencing. Enriched recombinant clones can be identified by comparison of the metasecretome library sequences before and after the panning round. Metasecretome Phage Display 529

Fig. 3 Demonstration of the sarkosyl resistance selection step. (A) Free phagemid DNA (samples were loaded directly on an agarose gel); (B) Total DNA, the sum of the free DNA and DNA encapsulated in the phagemid particles (samples were heated at 70 C in 1.2% SDS for 10 min before loading, to disassemble the sarkosyl-resistant phagemid particles). Lanes: 1, library phagemid particles (PPs) before incubation with sarkosyl; 2, after incubation with sarkosyl; 3, after incubation with sarkosyl and DNase I, followed by inactivation of DNase I

A metasecretome library of infectious PPs can be generated by transforming metasecretome ssDNA library into E. coli to generate a metasecretome master phagemid library, followed by infection with helper phage VCSM13 that encodes full-length pIII and therefore assembles infectious PPs. This phagemid phage-display library is suitable for standard afﬁnity screening as described elsewhere in this volume.

3.4 PacBio 1. PCR-amplify metasecretome insert sequences using the meta- Sequencing secretome library ssDNA (Subheading 3.3, step 5) as a tem- of Metasecretome plate and the primer pair pspF03/pspR03 complementary to Phage-Display Inserts the vector sequences ﬂanking the inserts (25–50 pg/μl template, 0.5 μM primers and 1Â PrimeSTAR Max premix; please 530 Milica Ciric et al.

Table 2 Thermal profile for PCR amplification of metasecretome phage-display library inserts

PCR step Temperature, C Time, s

1 Denature 98 10 2 Anneal 55 5 3 Extend 72 45 4 Repeat steps 1–3 total 35Â

refer to Table 1 for the primer sequences and Table 2 for the PCR cycling parameters). Set up separate PCR reactions for each ssDNA aliquot derived from the original master metagenomic library. 2. Visualize the PCR amplicons by agarose gel electrophoresis [1% (w/v) agarose gels]. If amplicons below 0.4 kb are present, excise DNA fragments of larger sizes (running above the 0.4 kb marker) and purify using the commercially available agarose gel clean-up kits or size selection kits (e.g., NucleoMag NGS Clean-up and Size Select magnetic beads, Macherey-Nagel GmbH & Co., Germany). The amplicons are now ready to submit for the PacBio sequencing.

4 Notes

1. Alternative methods for size separation are preparative gel electrophoresis followed by the extraction of DNA, or Sephar- ose CL-4B 200 (e.g., Sigma Aldrich) size exclusion resin. The gel extraction method, however, often results in a considerable loss of DNA. If gel extraction is the only possibility, increase the starting amount of metagenomic DNA (increase proportionally the number of used nebulizers). 2. Cm double-layer plates allow in-agar infection of the indicator strain TG1 prior to exposure to Cm. First, prepare 2Â YT agar plates (21 ml per plate) with 25 μg/ml chloramphenicol. Once solidiﬁed, these plates are overlaid with 9 ml chloramphenicol- free 2Â YT agar shortly before use. 3. Maximum of 10 μg of DNA should be used per nebulizer for the conditions given in the protocol. More than 10 μg DNA per nebulizer requires the increase in shearing time. Shearing time for each DNA sample requires experimental optimization. In the ﬁrst instance try 10 s, 30 s, 1 min, and 2 min to gauge the shearing time that gives DNA fragments ranging from 0.4 to 4 kb. Metasecretome Phage Display 531

4. Disposable nebulizers usually do not fit into standard centrifuge tubes; therefore, the swinging rotor buckets should be filled with cotton to secure the nebulizers. Use only the buckets with lids to avoid damaging the centrifuge. 5. Instead of combining end-repair enzymes, a DNATerminator End Repair Kit (Lucigen, USA) that already contains all three combined enzymes can be used. 6. Several molar ratios of vector (V) and insert (I) should be tested. For example V:I ¼ 1:1 or 1:2, 1:3, 1:5 and 2:1 are good starting points. The maximum amount of insert DNA per ligation should not exceed 500 ng. 7. The master shotgun metagenomic library can be also used for the production of PPs containing inserts corresponding to the whole metagenome (not just the metasecretiome) using a wild- type helper phage [e.g., VCSM13 (Stratagene, USA)]. This approach is used to avoid transformation of E. coli with ssDNA, which is of low efficiency and may result in decreased library diversity. 8. To separate PPs from the cells, chill collected mixture for 20 min at 4 C. Pellet chilled cells by centrifugation at 13,200 Â g for 20 min at 4 C. Collect the supernatant containing PPs and filter (pore size 0.45 μm) to eliminate any remaining bacterial cells. Concentrate PPs by precipitation in PEG/NaCl buffer (5% (w/v) PEG, 500 mM NaCl) for 2 h on ice (or overnight at 4 C) and pellet PPs by centrifugation at 13,200 Â g for 20 min at 4 C. As PPs-containing pellet will be spread over the wall of the centrifuge tubes. Carefully and thoroughly remove the buffer (e.g., by vacuum suction). Pres- ence of PEG/NaCl buffer in PP preparations can interfere with subsequent steps in the selection process; therefore, it is very important to remove it completely (however take care not to remove the pellet). Resuspend PPs in 500 μl of TN buffer by repeated pipetting of the buffer down phage pellet that is spread along the centrifuge tube wall. PPs can be stored short term at 4 C or long term at À80 C in 7% DMSO. 9. To monitor the sarkosyl and DNase I treatment steps in the selection process, prepare an agarose gel [0.6–0.8% (w/v)]. Split each aliquot from steps 1–5, Subheading 3.3, into two aliquots (2 Â 10 μl). Mix one set of aliquots with a DNA gel-loading dye and load directly onto the gel to visualize free DNA present in the samples. Mix the second set of aliquots with 4Â phage disassembly SDS-containing buffer at a 3:1 (PPs to buffer) volume ratio and disassemble by incubation at 70 C for 20 min. Load disassembled PPs onto a gel and subject to agarose gel electrophoresis. Total viral ssDNA can be visualized by post-staining with a solution of 0.5 μg/ml ethidium bromide (see Fig. 3 [21]). 532 Milica Ciric et al.

10. An alternative method for purifying ssDNA from sarkosyl- resistant PPs is by using a plasmid mini prep kit (e.g., High Pure Plasmid Isolation Kit, Roche). If this method is used, disassemble the PPs by heating at 70 C for 10 min in the presence of 1.2% (w/v) SDS and then continue the plasmid mini-prep protocol by adding Solution 3 (no addition of solutions 1 or 2).

Acknowledgment

This work was funded by the New Zealand Ministry of Business, Innovation and Employment (contract C10X0803), The Royal Society of New Zealand through a Marsden Fast Start grant and Palmerston North Medical Research Foundation. M.C. was partially supported by the Institute of Fundamental Sciences (Massey University, New Zealand) and F.N. was supported by a Common- wealth Fellowship and a Massey University Bursary. We would like to thank DSMZ Sequencing facility (Germany) and Dr. Boyke Bunk for help in development and optimization of protocol for PacBio sequencing of phage-display libraries.

References

1. Lamont RJ, Jenkinson HF (1998) Life below loci in pathogenic bacteria. Curr Biol 4 the gum line: pathogenic mechanisms of Por- (1):24–33 phyromonas gingivalis. Microbiol Mol Biol Rev 8. Maione D, Margarit I, Rinaudo CD, 62(4):1244–1263 Masignani V, Mora M, Scarselli M, 2. Orth K, Xu Z, Mudgett MB, Bao ZQ, Palmer Tettelin H, Brettoni C, Iacobini ET, Rosini R, LE, Bliska JB, Mangel WF, Staskawicz B, D’Agostino N, Miorin L, Buccato S, Dixon JE (2000) Disruption of signaling by Mariani M, Galli G, Nogarotto R, Nardi Yersinia effector YopJ, a ubiquitin-like protein Dei V, Vegni F, Fraser C, Mancuso G, Teti G, protease. Science 290(5496):1594–1597 Madoff LC, Paoletti LC, Rappuoli R, Kasper 3. Schwarz-Linek U, Hook M, Potts JR (2006) DL, Telford JL, Grandi G (2005) Identiﬁca- Fibronectin-binding proteins of gram-positive tion of a universal Group B streptococcus vac- cocci. Microbes Infect 8(8):2291–2298 cine by multiple genome screen. Science 4. Lipsitch M, O’Hagan JJ (2007) Patterns of 309:148–150 antigenic diversity and the mechanisms that 9. Yang XY, Lu J, Sun X, He QY (2012) Applica- maintain them. J R Soc Interface 4 tion of subproteomics in the characterization (16):787–802 of Gram-positive bacteria. J Proteome 75 5. Bayliss CD, Field D, Moxon ER (2001) The (10):2803–2810 simple sequence contingency loci of Haemo- 10. Lee SY, Choi JH, Xu Z (2003) Microbial cell- philus inﬂuenzae and Neisseria meningitidis.J surface display. Trends Biotechnol 21 Clin Invest 107(6):657–662 (1):45–52 6. Areschoug T, Carlsson F, Stalhammar- 11. Chen W, Georgiou G (2002) Cell-surface dis- Carlemalm M, Lindahl G (2004) Host- play of heterologous proteins: from high- pathogen interactions in Streptococcus pyogenes throughput screening to environmental appli- infections, with special reference to puerperal cations. Biotechnol Bioeng 79(5):496–503 fever and a comment on vaccine development. 12. Georgiou G, Stathopoulos C, Daugherty PS, Vaccine 22(Suppl 1):S9–S14 Nayak AR, Iverson BL, Curtiss R III (1997) 7. Moxon ER, Rainey PB, Nowak MA, Lenski RE Display of heterologous proteins on the surface (1994) Adaptive evolution of highly mutable of microorganisms: from the screening of Metasecretome Phage Display 533

combinatorial libraries to live recombinant vac- 25. Dotto GP, Enea V, Zinder ND (1981) Func- cines. Nat Biotechnol 15(1):29–34 tional analysis of bacteriophage f1 intergenic 13. Liu R, Yang C, Xu Y, Xu P, Jiang H, Qiao C region. Virology 114(2):463–473 (2013) Development of a whole-cell biocata- 26. Rakonjac J, Jovanovic G, Model P (1997) Fila- lyst/biosensor by display of multiple heterolo- mentous phage infection-mediated gene gous proteins on the Escherichia coli cell surface expression: construction and propagation of for the detoxification and detection of organo- the gIII deletion mutant helper phage phosphates. J Agric Food Chem 61 R408d3. Gene 198:99–103 (32):7810–7816 27. Rakonjac J, Feng J, Model P (1999) Filamen- 14. A˚ vall-J€a€askel€ainen S, Lindholm A, Palva A tous phage are released from the bacterial (2003) Surface display of the receptor-binding membrane by a two-step mechanism involving region of the Lactobacillus brevis S-layer pro- a short C-terminal fragment of pIII. J Mol Biol tein in Lactococcus lactis provides nonadhesive 289(5):1253–1265 lactococci with the ability to adhere to intesti- 28. Rakonjac J, Model P (1998) Roles of pIII in nal epithelial cells. Appl Environ Microbiol 69 filamentous phage assembly. J Mol Biol 282 (4):2230–2236 (1):25–41 15. Werne´rus H, Sta˚hl S (2004) Biotechnological 29. Gagic D, Ciric M, Wen WX, Ng F, Rakonjac J applications for surface-engineered bacteria. (2016) Exploring the secretomes of microbes Biotechnol Appl Biochem 40(3):209–228 and microbial communities using filamentous 16. Rosander A, Bjerketorp J, Frykberg L, Jacobs- phage display. Front Microbiol 7:429 son K (2002) Phage display as a novel screen- 30. Ploss M, Facey SJ, Bruhn C, Zemel L, ing method to identify extracellular proteins. J Hofmann K, Stark RW, Albert B, Hauer B Microbiol Methods 51(1):43–55 (2014) Selection of peptides binding to metal- 17. Jacobsson K, Rosander A, Bjerketorp J, Fryk- lic borides by screening M13 phage display berg L (2003) Shotgun phage display-selection libraries. BMC Biotechnol 14:12 for bacterial receptins or other exported pro- 31. Naik RR, Brott LL, Clarson SJ, Stone MO teins. Biol Proced Online 5(1):123–135 (2002) Silica-precipitating peptides isolated 18. Rosander A, Guss B, Pringle M (2011) An from a combinatorial phage display peptide IgG-binding protein A homolog in Staphylococ- library. J Nanosci Nanotechnol 2(1):95–100 cus hyicus. Vet Microbiol 149(1):273–276 32. Guo J, Catchmark JM, Mohamed MN, Benesi 19. Bjerketorp J, Rosander A, Nilsson M, AJ, Tien M, Kao TH, Watts HD, Kubicki JD Jacobsson K, Frykberg L (2004) Sorting a (2013) Identification and characterization of a Staphylococcus aureus phage display library cellulose binding heptapeptide revealed by against ex vivo biomaterial. J Med Microbiol phage display. Biomacromolecules 14 53(10):945–951 (6):1795–1805 20. Ciric M, Moon CD, Leahy SC, Creevey CJ, 33. Dennissen MA, Jenniskens GJ, Pieffers M, Ver- Altermann E, Attwood GT, Rakonjac J, Gagic steeg EM, Petitou M, Veerkamp JH, van Kup- D (2014) Metasecretome-selective phage dis- pevelt TH (2002) Large, tissue-regulated play approach for mining the functional poten- domain diversity of heparan sulfates demon- tial of a rumen microbial community. BMC strated by phage display antibodies. J Biol Genomics 15:356 Chem 277(13):10982–10986 21. Jankovic D, Collett MA, Lubbers MW, Rakon- 34. Pavoni E, Vaccaro P, Anastasi AM, Minenkova jac J (2007) Direct selection and phage display O (2014) Optimized selection of anti-tumor of a Gram-positive secretome. Genome Biol 8 recombinant antibodies from phage libraries (12):R266 on intact cells. Mol Immunol 57(2):317–322 22. Rakonjac J, Bennett NJ, Spagnuolo J, Gagic D, 35. Babickova J, Tothova L, Boor P, Celec P Russel M (2011) Filamentous bacteriophage: (2013) In vivo phage display—a discovery biology, phage display and nanotechnology tool in molecular biomedicine. Biotechnol applications. Curr Issues Mol Biol 13 Adv 31(8):1247–1259 (2):51–76 36. Dias-Neto E, Nunes DN, Giordano RJ, Sun J, 23. Russel M (1995) Moving through the mem- Botz GH, Yang K, Setubal JC, Pasqualini R, brane with filamentous phages. Trends Micro- Arap W (2009) Next-generation phage display: biol 3(6):223–228 integrating and comparing available molecular 24. Model P, Jovanovic G, Dworkin J (1997) The tools to enable cost-effective high-throughput Escherichia coli phage shock protein operon. analysis. PLoS One 4(12):e8338 Mol Microbiol 24:255–261 37. Di Niro R, Sulic AM, Mignone F, D’Angelo S, Bordoni R, Iacono M, Marzari R, Gaiotto T, 534 Milica Ciric et al.

Lavric M, Bradbury AR, Biancone L, Zevin- understanding of antibody repertoires. Front Sonkin D, De Bellis G, Santoro C, Sblattero Immunol 4:265 D (2010) Rapid interactome profiling by mas- 41. Travers KJ, Chin CS, Rank DR, Eid JS, Turner sive sequencing. Nucleic Acids Res 38(9):e110 SW (2010) A flexible and efficient template 38. Ravn U, Gueneau F, Baerlocher L, Osteras M, format for circular consensus sequencing and Desmurs M, Malinge P, Magistrelli G, SNP detection. Nucleic Acids Res 38(15):e159 Farinelli L, Kosco-Vilbois MH, Fischer N 42. Russel M (1993) Protein-protein interactions (2010) By-passing in vitro screening--next during filamentous phage assembly. J Mol Biol generation sequencing technologies applied to 231(3):689–697 antibody display and in silico candidate selec- 43. Ciric M. Massey University metasecretome tion. Nucleic Acids Res 38(21):e193 phage display: a new approach for mining sur- 39. Hoen PA, Jirka SM, Ten Broeke BR, Schultes face and secreted proteins from microbial com- EA, Aguilera B, Pang KH, Heemskerk H, munities: a thesis presented in partial Aartsma-Rus A, van Ommen GJ, den Dunnen fulfillment of the requirements for the degree JT (2012) Phage display screening without of Doctor of Philosophy in Biochemistry at repetitious selection rounds. Anal Biochem Massey University, Palmerston North, 421(2):622–631 New Zealand 40. Mathonet P, Ullman CG (2013) The application of next generation sequencing to the Chapter 30

Phagekines: Screening Binding Properties and Biological Activity of Functional Cytokines Displayed on Phages

Gertrudis Rojas and Tania Carmenate

Abstract

The current chapter focuses on the use of ﬁlamentous phages to display, modify, and characterize cytokines, which are proteins belonging to a versatile group of essential mediators involved in cell-cell communication. Cytokines exhibit a considerable diversity, both in functions and in structural features underlying their biological effects. A broad variety of cytokines have been successfully displayed on phages, allowing the high-throughput study of their binding properties and biological activities and the discovery of novel therapeutics through directed evolution. The technical singularities and some potential applications of cytokine phage display are illustrated here with the case of Interleukin-2, a prototypic member of the four- alpha-helix bundle cytokine family playing a pivotal role in the immune response and having a long history of therapeutic use.

Key words CD25, CD122, Cell proliferation assay, CTLL-2, ELISA screening, IL-2 receptor, Inter- leukin-2, Kunkel mutagenesis, Phage display

1 Introduction

Even though phage display has evolved for more than 30 years as a powerful tool for manipulating proteins [1], most applications are concentrated in the ﬁelds of antibody engineering and peptide screening (widely described in this and other volumes). The availability of universal single-pot libraries of antibody fragments [2] and random oligopeptides [3], from which binders to virtually any target can be readily selected, has resulted in such a preponderance. Nevertheless, the versatility of the phage-display platform is high enough to allow engineering other proteins of diverse length and structural properties that have evolved to perform different functions [4]. Among them, cytokines are a prominent subset. Cytokines form a group of protein mediators involved in cell- to-cell communication. They are secreted by speciﬁc cell types in a highly regulated manner and exert a variety of paracrine and auto- crine effects on target cells armed with suitable cytokine receptors,

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4_30, © Springer Science+Business Media LLC 2018 535 536 Gertrudis Rojas and Tania Carmenate

playing an important role in the regulation of essential processes such as inflammation and immune responses. Natural cytokine functions have been exploited in multiple ways for the therapy of human disorders [5, 6]. Engineering cytokines to develop agonists, antagonists, and superagonists has become in an additional source of novel therapeutics [7–10]. Phage display of cytokines implies technical challenges due to the diversity of their folding requirements and structural features, the complexity of their interactions with cytokine cell surface receptors (often multi-chain complexes), and the broad variety of assays that are required to characterize their biological properties beyond mere binding. However, successful display of several cytokines has been achieved, including chemokines [11], Interleukin-3 [12], Interferon-alpha [13], and Tumor necrosis factor-alpha [14]. Some phage-displayed cytokines have been shown to repro- duce the functional properties of the original molecules [12, 14]. Human Interleukin-2 (IL-2) was one of the first functional cytokines displayed on filamentous phages [15, 16]. More recently, this initial experience has been greatly expanded in our laboratory. Hundreds of mutated IL-2 variants derived from both human and mouse IL-2 through Kunkel mutagenesis [17] were displayed in order to identify the residues contributing to recognition by several anti-IL-2 antibodies [18, 19]. Beyond epitope mapping, phage display was used to investigate binding to receptor subunits and biological activity of IL-2-derived muteins [20]. The ability of phage-displayed IL-2 to signal through cell surface receptors and induce proliferation of IL-2-dependent cells allows studying biological properties and comparing multiple variants in a simple way. Besides its central role in immune system biology and the above-described experience, IL-2 phage display is particularly interesting for protein engineers because IL-2 belongs to the family of four-alpha-helix bundle cytokines, which are similar in folding and include a plethora of immunologically relevant molecules like Interleukins-4, -7, -9, -15, and -21, among others. Il-2 can be considered a model for these structurally related proteins. The protocols described here for display, modification, and characterization of IL-2-derived molecules could thus be adapted to the study and manipulation of other cytokines.

2 Materials

2.1 Displaying 1. Use XL-1 Blue Escherichia coli (E.coli) strain (recA1 endA1 0 Human/Mouse IL-2 gyrA96 thi-1 hsdR17 supE44 relA1 lac F proAB lacIqZ_M15 on Filamentous Tn10 Tetr) to obtain plasmid DNA during cloning, sequenc- Phages ing, and storage of genetic constructs. Phage Display of Cytokines 537

2. Use TG1 E. coli strain (K12_(lac-pro), supE, thi, hsdD5/F0 traD36, proA+B+, lacIq, lacZ_M15) to rescue phagemid- containing viral particles displaying IL-2. 3. Use M13KO7 helper phage to rescue phagemid-containing viral particles. 4. Prepare 2Â TY by dissolving tryptone (16 g/L), yeast extract (10 g/L), and sodium chloride (5 g/L) in water. Sterilize by autoclaving during 20 min at 120 C and 1 atm. 5. Add 1.5% (w/v) of bacteriological agar to liquid 2Â TY to obtain solid 2Â TY. Sterilize by autoclaving during 20 min at 120 C and 1 atm. 6. Prepare stock solutions of the following additives in sterilized water and filter through 0.2 μm disposable filters: l Glucose (40% w/v). l Ampicillin (100 mg/mL). l Kanamycin (70 mg/mL). 7. Prepare 2Â TY/AG by supplementing 2Â TY with ampicillin (100 μg/mL) and glucose (2% w/v) before use. 8. Prepare 2Â TY/AK by supplementing 2Â TY with ampicillin (100 μg/mL) and kanamycin (70 μg/mL) before use. 9. Always use aerosol-resistant filtered pipette tips to handle samples containing phages. This applies to both M13KO7 helper phage and phagemid-containing viral particles. 10. Use disposable gamma-irradiated plastic materials for E. coli culture, DNA manipulation, and phage production and purification (pipette tips, aerosol-resistant filtered pipette tips, vials, 50 mL conical-bottom tubes, Petri dishes, and 0.2 μm syringe filters and centrifugal filtering devices). 11. Leave all the glassware (like erlenmeyer flasks) and non-disposable plastic materials (such as centrifuge bottles) that have been in contact with phage samples (M13KO7 helper phage or phagemid-containing viral particles) in 2% bleach overnight to decontaminate. 12. Sterilize decontaminated and washed glassware and non-disposable plastic materials during 20 min at 120 C and 1 atm before use. 13. Prepare PEG/NaCl phage precipitation solution by dissolving polyethylenglycol 6000 or 8000 (20% w/v) and sodium chloride (2.5 mol/L) in water. Sterilize by autoclaving during 20 min at 120 C and 1 atm. 14. Prepare 10Â phosphate-buffered saline (PBS), by dissolving sodium chloride (1.5 mol/L), dibasic sodium phosphate (80 mmol/L), and potassium monobasic phosphate 538 Gertrudis Rojas and Tania Carmenate

(20 mmol/L) in water. Adjust pH to the range 7.2–7.4. Steril- ize by autoclaving during 20 min at 120 C and 1 atm. 15. Prepare PBS by diluting 100 mL of 10Â PBS up to 1 L with sterilized water in a previously sterilized bottle. 16. Prepare 60% glycerol solution (v/v) in water. Sterilize by autoclaving during 20 min at 120 C and 1 atm.

À À 2.2 Producing 1. Use CJ236 E. coli strain (dut ung thi-1 relA1 spoT1 mcrA/ 0 Single-Stranded DNA pCJ105 (F camr)) to obtain single-stranded DNA template for Templates for IL- site-directed mutagenesis of the IL-2 genes. 2 Kunkel Mutagenesis 2. Use XL-1 Blue E. coli strain (see the previous Subheading) as an alternative host to obtain single-stranded DNA template for site-directed mutagenesis of the IL-2 genes. 3. Use M13KO7 helper phage to rescue template-containing viral particles. 4. Prepare the following media and additives as described in the previous Subheading: l 2Â TY. l Solid 2Â TY. l glucose (40% w/v). l 2Â TY/AG. l 2Â YT/AK. 5. Prepare stock solutions of the following additives in sterilized water and filter through 0.2 μm disposable filters: l Chloramphenicol (5 mg/mL). l Uridine (0.25 mg/mL). 6. Dissolve tetracycline in ethanol (5 mg/mL) to obtain a stock solution. 7. Supplement 2Â TY with the additives described above immediately before use (at the final concentrations described in Subheading 3.2). 8. Always use aerosol-resistant filtered pipette tips to handle samples containing phages, as described in the previous Subheading. 9. Use disposable gamma-irradiated plastic materials for E. coli culture, phage production and purification, and single- stranded DNA purification (pipette tips, aerosol-resistant filtered pipette tips, vials, 50 mL conical-bottom tubes, and Petri dishes). 10. Decontaminate, wash, and sterilize all the glassware and non-disposable plastic materials as described in the previous Subheading. Phage Display of Cytokines 539

11. Prepare PEG/NaCl phage precipitation solution and PBS as described in the previous Subheading. 12. Use QIAprep Spin M13 kit solutions and columns (Qiagen, Hilden, Germany) to purify single-stranded DNA template, with a modiﬁed procedure described in Subheading 3.2.

2.3 Site-Directed 1. The protocol was optimized using the following Kunkel Mutagenesis DNA-modifying enzymes provided by New England Biolabs to Construct IL-2 (Ipswich, USA): Variants l T4 Polynucleotide kinase (10,000 units/mL). l T7 DNA polymerase (unmodified) (10,000 units/mL). l T4 DNA ligase (400,000 units/mL). 2. Use solutions of ATP (10 mmol/L) and dNTPs (10 mmol/L each) also provided by New England Biolabs. 3. Prepare 10Â TM buffer by dissolving Tris (500 mmol/L) and magnesium chloride (100 mmol/L) in water. Adjust pH to 7.5 with HCl. Sterilize by autoclaving during 20 min at 120 C and 1 atm. 4. Prepare 100 mmol/L dithiotreitol (DTT) solution in sterilized water. Filter through 0.2 μm disposable filters. Make individual aliquots in vials and keep at À20 C until use. 5. Use sterilized water to dilute mutagenic oligonucleotides and to prepare the mutagenesis reactions. 6. Use XL-1 Blue E. coli strain (see Subheading 2.1) to obtain plasmid DNA for sequencing of the mutated genes. 7. Use TG1 E. coli strain (see Subheading 2.1) to rescue phagemid-containing viral particles displaying IL-2 mutated variants. 8. Prepare the following media and additives as described in Sub- heading 2.1: l 2Â TY. l Solid 2Â TY. l 2Â TY/AG. 9. Use disposable gamma-irradiated plastic materials for site- directed mutagenesis, E. coli culture and DNA purification (pipette tips, vials, PCR tubes/plates, 50 mL conical-bottom tubes, and Petri dishes). 10. Use QIAprep Spin minikit (Qiagen) to purify plasmid DNA for sequencing and storage, according to the manufacturer’s instructions. 540 Gertrudis Rojas and Tania Carmenate

2.4 Quantifying 1. Use polyvinyl chloride microtitration plates as the solid the Display Levels support. of IL-2 Variants by 2. Use 9E10 (anti-c-myc tag) monoclonal antibody (mAb) to ELISA measure the levels of phage-displayed IL-2 variants. 3. Prepare PBS as described in Subheading 2.1. 4. Always use aerosol-resistant ﬁltered pipette tips to handle samples containing phages, as described in Subheading 2.1. 5. Use 4% (w/v) skim powder milk in PBS (M-PBS) to block the plates and to dilute both phage samples and the conjugated antibody. 6. Prepare washing solution by diluting Tween 20 in water at a ﬁnal concentration of 0.1% (v/v). 7. The protocols were optimized using the anti-M13/horseradish peroxidase conjugate (1/5000 working dilution) provided by GE-Healthcare (Pittsburgh, USA) to detect bound phages. 8. Prepare a solution of dibasic sodium phosphate (200 mmol/L) in water. Prepare a second solution of citric acid (100 mmol/L) in water. Adjust the pH of the phosphate solution to 5.0 by adding the citric acid solution to obtain peroxidase substrate buffer. Keep the buffer at 4 C until use. 9. Prepare peroxidase substrate solution immediately before use by adding 5 mg of o-phenylenediamine and 5 μL of hydrogen peroxide (30% v/v) to 10 mL of peroxidase substrate buffer. 10. Dilute HCl (10% v/v) in water to obtain the stop solution.

2.5 Screening 1. Recombinant extracellular domains of IL-2 receptor subunits Receptor Binding can be either purchased from commercial sources (see Note 1) Properties of Phage- or expressed at the laboratory (see Note 2). Displayed IL-2 2. Speciﬁc mAbs against human or mouse IL-2 can be used to Variants by ELISA conﬁrm the presence and proper folding of phage-displayed IL-2 molecules (see Note 3 for examples of antibody clones). 3. See the previous Subheading for the rest of materials and solutions required to perform the ELISA.

® 2.6 Inducing CTLL- 1. Use the mouse CTLL-2 cell line (ATCC TIB-214™) to test 2 Proliferation by in vitro biological activity of IL-2 and IL-2 mutated variants Phage-Displayed IL-2 displayed on phages. This cell line is derived from cytotoxic and IL-2-Derived CD8þ T-cells on a C57BL6 background and depends upon Muteins IL-2 for growth. CTLL-2 cells respond to both human and mouse IL-2. 2. Store the frozen cells on liquid nitrogen. Use fetal bovine serum with 10% (v/v) dimethylsufoxide (DMSO) as a freezing medium. Phage Display of Cytokines 541

3. Use RPMI culture medium, supplemented with 2 mmol/L L- glutamine, 1 mmol/L sodium pyruvate, 10% (v/v) fetal bovine serum, 50 U/mL penicillin, 50 μg/mL streptomycin, and 50 IU/mL of recombinant IL-2 (see Note 4). 4. Use only disposable cell-culture treated gamma-irradiated plastic materials for CTLL-2 culture (pipette tips, vials, cryovials, 50 mL conical-bottom tubes, multi-well plates, and ﬂasks).

3 Methods

3.1 Displaying 1. Synthesize the IL-2 genes, ﬂanked by ApaLI and NotI unique Human/Mouse IL-2 restriction sites. Exclude the codons coding for the ﬁrst two on Filamentous amino acids (Ala-Pro) from the genetic constructs and remove Phages the unpaired Cys residues (C125 in human IL-2 and C140 in mouse IL-2) by replacing the corresponding codons by a triplet encoding a Ser (see Note 5). Optimize the DNA sequences according to E. coli codon usage. 2. Clone the IL-2 genes between ApaLI and NotI restriction sites into pCSM phagemid vector (Fig. 1) using standard DNA cloning procedures (see Note 6). 3. Transform TG1 E. coli competent cells with the resulting genetic constructs and grow transformed cells on solid 2Â TY/AG during 16–20 h at 37 C. 4. Pick isolated colonies and inoculate them into 50 mL tubes containing 10 mL of 2Â TY/AG (see Note 7). Grow at 37 C with shaking at 250 rpm during 16–20 h. 5. Dilute the cell suspension obtained after step 4 (1/100) in a 50 mL tube containing 10 mL of fresh 2Â TY/AG and grow at

b ApaLI a NotI Amber pLac

RBS DsbA SP h/m IL-2 gene M13 gene 3 Term 3 TSSSTKKTQLQLEHLLLDLQMILNGINNYKNPKLTRMLTFKFYMPKKATELKHLQC 58

c-myc 6His 59 LEEELKPLEEVLNLAQSKNFHLRPRDLISNINVIVLELKGSETTFMCEYADETATI 114 115 VEFLNRWITFSQSIISTLT 133 C pCSM+ Coli E1 5.0 kb F1 ori 3 TSSSTSSSTAEAQQQQQQQQQQQQHLEQLLMDLQELLSRMENYRNLKLPRMLTFKF 58 59 YLPKQATELKDLQCLEDELGPLRHVLDLTQSKSFQLEDAENFISNIRVTVVKLKGS 114 115 DNTFECQFDDESATVVDFLRRWIAFSQSIISTSP 148 Ampr

Fig. 1 Genetic constructs for IL-2 phage display. Schematic representation of the pCSMþ phagemid vector is shown in panel a. The vector includes DNA sequences coding for DsbA signal peptide (MKKIWLALAGLVLAF- SASA), human or mouse (h/m) IL-2, and hexahistidine and c-myc tags. Additional elements are shown: lac promoter, ribosomal binding site, amber stop codon, ApaLI and NotI restriction sites, M13 gene 3, transcription terminator sequence, E. coli and phage replication origins, and ampicillin resistance gene. Panels b and c show protein sequences deduced from the cloned human and mouse IL-2 genes respectively. Unpaired C125 and C140 in these sequences are replaced by Ser (shaded in grey) 542 Gertrudis Rojas and Tania Carmenate

37 C with shaking at 250 rpm until the culture reaches an absorbance at 600 nm in the range 0.4–0.8. 6. Add 1011 plaque forming units (pfu) of M13KO7 helper phage (see Note 8) to grown E. coli and incubate at 37 C during 30 min without shaking. 7. Centrifuge at 2000 Â g during 15 min. Carefully remove the supernatant. 8. Resuspend the cell pellet in 40 mL of 2Â TY/AK contained in a 250 mL erlenmeyer flask. 9. Grow the cells at 28 C during 16–20 h with shaking at 250 rpm. 10. Centrifuge at 4000 Â g during 15 min at 4 C. 11. Collect the supernatant and mix it with 10 mL of PEG/NaCl solution. Incubate during 1 h on ice to precipitate the phages (see Note 9). 12. Centrifuge at 4000 Â g during 15 min at 4 C. Carefully remove the supernatant. 13. Resuspend the phage pellet in 1 mL of PBS and transfer to a vial. 14. Centrifuge at 10,000 Â g during 10 min at 4 C in a microcentrifuge to remove the remaining E. coli and cell debris. 15. Mix the supernatant with 250 μL of PEG/NaCl solution in a new vial (see Note 10). Incubate during 20 min on ice to precipitate the phages. 16. Centrifuge at 10,000 Â g during 10 min at 4 C in a microcentrifuge. Carefully remove the supernatant. 17. Gently resuspend the phage pellet in 1 mL of PBS (see Note 11). Add 500 μL of 60% glycerol solution to the vial and mix. 18. Filter purified phages through a 0.2 μm centrifugal filtering device (see Note 12) to sterilize them (see Note 13). 19. Keep purified phages at À20 C to characterize their binding and biological properties as described in Subheadings 3.4–3.6.

3.2 Producing 1. Transform CJ236 E. coli competent cells with the genetic con- Single-Stranded DNA structs obtained after step 2 of Subheading 3.1 and grow Templates for IL- transformed cells in solid 2Â TY/AG during 16–20 h at 2 Kunkel Mutagenesis 37 C(see Note 14). 2. Pick an isolated colony (see Note 15) and inoculate it into a 50 mL tube containing 10 mL of 2Â TY/AG supplemented with 5 μg/mL chloramphenicol (see Note 16). Grow at 37 C with shaking at 250 rpm during 16–20 h. 3. Dilute the cell suspension obtained after step 2 (1/100) in a 50 mL tube containing 10 mL of fresh 2Â TY/AG Phage Display of Cytokines 543

supplemented with 5 μg/mL chloramphenicol and grow at 37 C with shaking at 250 rpm until the culture reaches an absorbance at 600 nm in the range 0.4–0.8. 4. Add 1011 pfu of M13KO7 helper phage (see Note 8) and incubate at 37 C during 30 min without shaking. 5. Centrifuge at 2000 Â g during 15 min. Carefully remove the supernatant. 6. Resuspend the cell pellet in 40 mL of 2Â TY/AK supplemented with 2% glucose and 0.25 μg/mL uridine, contained in a 250 mL erlenmeyer flask (see Note 17). 7. Grow the cells at 37 C during 20 h with shaking at 250 rpm (see Note 18). 8. Centrifuge at 4000 Â g during 15 min at 4 C. 9. Collect the supernatant and mix it with 10 mL of PEG/NaCl solution. Incubate during 1 h on ice to precipitate phages (see Note 19). 10. Centrifuge at 4000 Â g during 15 min at 4 C. Carefully remove the supernatant. 11. Resuspend the phage pellet in 0.5 mL of PBS and transfer to a vial. 12. Centrifuge at 10,000 Â g during 10 min in a microcentrifuge to remove the remaining E. coli and cell debris. 13. Mix the supernatant in a new vial with 10 μL of MP solution from the Qiaprep Spin M13 kit. Incubate for 2 min at room temperature (RT) (see Note 20). 14. Add the mixture to a QIA column. Centrifuge for 1 min at 6000 Â g. Discard the flow-through. 15. Add 0.7 mL of MLB/PB solution from the Qiaprep Spin M13 kit to the column. Centrifuge 1 min at 6000 Â g. Discard the flow-through. 16. Add 0.7 mL of MLB/PB solution from the Qiaprep Spin M13 kit to the column. Incubate for 1 min at RT. 17. Centrifuge for 1 min at 6000 Â g. Discard the flow-through. 18. Add 0.75 mL of PE solution from the Qiaprep Spin M13 kit (already containing ethanol as recommended by the manufacturer) to the column. Centrifuge for 1 min at 6000 Â g. Dis- card the flow-through. 19. Repeat step 18. 20. Transfer the column to an empty vial and centrifuge again 1 min at 6000 Â g to remove residual PE. 544 Gertrudis Rojas and Tania Carmenate

Fig. 2 Typical results of agarose gel electrophoresis of puriﬁed single-stranded DNA templates and site- directed Kunkel mutagenesis products. Lanes a, c, e and g show the main band corresponding to single- stranded DNA, sometimes accompanied by less intense bands with lower electrophoretic mobility. Lanes b, d, f and h correspond to site-directed mutagenesis products and exhibit the band corresponding to double- stranded mutated DNA and a band with lower electrophoretic mobility which should represent DNA derived from aberrant strand displacement synthesis (polymerization does not stop after the emerging strand reaches the double strand formed between the template and mutagenic oligonucleotide). A very weak band between both presumably corresponds to double-stranded nicked DNA (the nick between end of the emerging strand and the beginning of the mutagenic oligonucleotide is not sealed). GeneRuler 1 kb Plus DNA ladder is included (lane i)

21. Transfer the column to a new vial. Add 100 μL of EB solution from the Qiaprep Spin M13 kit to the center of the column and incubate for 10 min at RT in order to elute DNA. 22. Centrifuge 1 min at 6000 Â g. Collect the eluted DNA in the vial. 23. Visualize the eluted single-stranded DNA (1 μL) in an agarose gel (1%) electrophoresis in the presence of ethidium bromide according to standard procedures (see Note 21). Typical electrophoresis results are shown in Fig. 2 (see lanes a, c, e, and g). 24. Determine the single-stranded DNA concentration in a Nano- drop quantitation machine. Yields vary widely between 50 and 500 ng/μL.

3.3 Site-Directed 1. Design and order the synthesis of the mutagenic oligonucleo- Kunkel Mutagenesis tides required to introduce the desired replacements. These are to Construct IL-2 antisense oligonucleotides (see Note 22) annealing with IL-2 Variants genes from 15 nucleotides before the beginning of the target region to be modified to 15 nucleotides after its end, and containing the desired changes (see Note 23). The target region can include a single triplet to be modified (for individual mutations) or several contiguous triplets for multiple mutated variants, as well as various triplets separated by non-changed sequences. Distant target sequences can be modified by simultaneously using several mutagenic oligonucleotides (see Note 24). Figure 3 illustrates the diversity of mutagenic oligonucleotides’ designs. Phage Display of Cytokines 545 a 55 62 71 I) HLQCLEEELKPLEEVLN II) 5' CATCTGCAGTGCCTGGAAGAAGAACTGAAACCGCTGGAAGAAGTGCTGAAC 3'

III) 5' CACTTCTTCCAGCGGTTTCAGCGCTTCTTCCAGGCACTG 3'

IV) 3' GTCACGGACCTTCTTCGCGACTTTGGCGACCTTCTTCAC 5' ...... 5' CATCTGCAGTGCCTGGAAGAAGAACTGAAACCGCTGGAAGAAGTGCTGAAC 3'

55 62 71 V) HLQCLEEALKPLEEVLN VI) 5' CATCTGCAGTGCCTGGAAGAAGCGCTGAAACCGCTGGAAGAAGTGCTGAAC 3' b 73 80 81 85 86 92 99 I) AQSKNFHLRPRDLISNINVIVLELKGS II) 5' GCCCAGAGCAAAAACTTTCATCTGCGTCCGCGTGATCTGATTAGCAACATCAACGTGATTGTGCTGGAACTGAAAGGCAGC 3'

III) 5' TTTCAGTTCCAGCACAAACACGTTGATGTTGCTCACCACATCACGCGGATCAAAATGAAAGTTTTTGCT 3'

IV) 3' TCGTTTTTGAAAGTAAAACTAGGCGCACTACACCACTCGTTGTAGTTGCACAAACACGACCTTGACTTT 5' ...... 5' GCCCAGAGCAAAAACTTTCATCTGCGTCCGCGTGATCTGATTAGCAACATCAACGTGATTGTGCTGGAACTGAAAGGCAGC

73 80 81 85 86 92 99 V) AQSKNFHFDPRDVVSNINVFVLELKGS VI) 5' GCCCAGAGCAAAAACTTTCATTTTGATCCGCGTGATGTGGTGAGCAACATCAACGTGTTTGTGCTGGAACTGAAAGGCAGC 3'

Fig. 3 Examples of Kunkel mutagenic oligonucleotides’ design to introduce the single replacement E62A (panel a) and a set of ﬁve mutations comprising L80F, R81D, L85V, I86V and I92F (panel b) in human IL-2. The original protein (I) and DNA (II) IL-2 sequences contain the positions to be changed (shaded in grey). Mutagenic oligonucleotides are shown in (III), with the modiﬁed triplet(s) underlined. Annealing between each mutagenic oligonucleotide and the original template appears in (IV). Mutated protein (V) and DNA (VI) sequences exhibit the desired changes (underlined)

2. Prepare stocks of mutagenic oligonucleotides (dissolved in water) at 330 ng/μL. 3. Prepare a phosphorylation mix, containing the following components per each mutagenic oligonucleotide to be phosphorylated (see Note 25): l 2 μL10Â TM buffer. l 2 μL ATP (10 mmol/L). l 1 μL DTT (100 mmol/L). l 11 μL water. l 2 μL Polynucleotide kinase (10,000 units/mL). 4. Add 2 μL of each oligonucleotide stock to 18 μL of the mix described above. Incubate for 1 h at 37 C. 546 Gertrudis Rojas and Tania Carmenate

5. Calculate the volumes required to prepare the annealing reactions: l 2.5 μLof10Â TM buffer. l x μL of single-stranded template DNA (1 μg DNA). l 2 μL of each mutagenic oligonucleotide to be used in the same reaction. l y μL of water to complete a total volume of 25 μL. 6. Mix all the components described above, except the mutagenic oligonucleotides. 7. Add 2 μL of each phosphorylated oligonucleotide obtained after step 4 to the annealing mix in a PCR tube or PCR microplate (see Note 26). 8. Incubate all the annealing reactions 3 min at 90 C, 3 min at 50 C, and 5 min at 20 C on a thermocycler. 9. Prepare a ﬁll-in mix with the following components per each annealing reaction (see Note 27): l 1 μL ATP (10 mmol/L). l 2.5 μL dNTPs (10 mmol/L each). l 1.5 μL DTT (100 mmol/L). l 0.6 μL T4 DNA ligase (400,000 units/mL). l 0.4 μL T7 DNA polymerase (10,000 units/mL). 10. Add 6 μL of the above-described ﬁll-in mix to each annealing reaction. Incubate during 16–20 h at RT. 11. Visualize 8 μL of the products of each reaction in an agarose gel (1%) in the presence of ethidium bromide according to standard procedures. Use 1 μL of the single-stranded DNA template as a control. Figure 2 (lanes b, d, f, and h) shows the typical electrophoretic pattern of successful reactions (see Note 28). 12. Transform competent XL-1 Blue E. coli cells with 5 μL of the products from each reaction and grow transformed cells on solid 2Â TY/AG during 16–20 h at 37 C. 13. Pick isolated colonies (up to six from each mutagenesis reaction) and inoculate each one in 5 mL of 2Â TY/AG. 14. Grow the cultures during 16–20 h at 37 C with shaking at 250 rpm. 15. Purify plasmid DNA with QIAprep Spin minikit according to the manufacturer’s instructions. 16. Send the plasmids for automated sequencing of the mutated IL-2 genes. Phage Display of Cytokines 547

Fig. 4 Typical results of Kunkel mutagenesis products’ sequencing. Panel a shows an original segment of the human IL-2 gene, while panel b shows the same segment after introducing the replacements Q126Y, I129D and S130R. Changed triplets are underlined in both of them. The sequence in panel c, although apparently mutated, exhibits an overlapping between the mutated nucleotide sequence and the original one. Overlapping peaks are indicated by arrows

17. Select at least one genetic construct having the desired mutation (or set of mutations) in the IL-2 gene without additional undesired changes along it. Verify that there is no overlap between the mutated sequence and the non-mutated original sequence at the target sites (see Note 29). Examples of overlapped and non-overlapped sequences are shown in Fig. 4. 18. Transform TG1 E. coli competent cells with the selected mutated genetic constructs and grow transformed cells on solid 2Â TY/AG during 16–20 h at 37 C. 19. Produce phages displaying the mutated IL-2 variants following the procedure described in Subheading 3.1 (starting from step 4).

3.4 Quantifying 1. Coat a polyvinyl chloride ELISA microplate with 100 μL/well the Display Levels of 10 μg/mL anti-c-myc tag 9E10 mAb diluted in PBS (see of IL-2 Variants by Note 30). Cover the plate with a lid and incubate during ELISA 16–20 h at 4 C. 2. Discard the coating solution by inverting the plate several times. 548 Gertrudis Rojas and Tania Carmenate

3. Add 200 μL/well of 4% M-PBS to block the plate. Cover the plate with a lid and incubate for 30 min at RT. 4. During the above-mentioned incubation, dilute phage standard (see Note 31) and the other purified and filtered phage samples (see Note 32) in M-PBS. 5. Discard the blocking solution by inverting the plate several times. 6. Add 100 μL/well of diluted phage standard and samples. Each dilution (of both standard and samples) should be applied at least twice on the plate to obtain independent replicates of each data point. Add M-PBS alone to some coated and blocked wells in order to assess the background level of the assay. Coat the plate with a lid and incubate during 1 h at RT. 7. Discard the samples by inverting the plates several times and wash the plate at least five times by filling the wells with washing solution with a washing bottle and discarding it (see Note 33). 8. Add 100 μL/well of the anti-M13/peroxidase conjugate (diluted 1/5000 in M-PBS). Coat the plate with a lid and incubate during 1 h at RT. 9. Wash the plate at least eight times as described in step 7 of the current Subheading. 10. Add 100 μL/well of peroxidase substrate solution. Incubate during 15 min at RT. A yellow/orange color should develop in the standard and sample wells, while the wells containing no phage samples should remain colorless (indicating low background levels). 11. Stop the reaction with 50 μL/well of the stop solution. 12. Read the absorbances at 490 nm with a microplate reader. 13. Calculate the mean absorbance values for replicated data points. 14. Use the absorbance values from diluted phage standard to obtain the best fitting curve (see Note 34). Assume that undi- luted standard phage preparation has a concentration of 100 arbitrary display units/mL of the heterologous protein. Figure 5 shows an example of a typical standard curve. 15. Interpolate the absorbance value of each diluted phage sample on the standard curve to estimate its relative display level (see Note 35). Take into account the dilution factors to calculate the original concentrations (in display units/mL) in the undi- luted samples (see Note 36). Table 1 shows an example of calculated concentrations for a set of phage-displayed IL-2 variants. Phage Display of Cytokines 549

1.2

0.8

0.6 y = 0.2428ln(x) + 0.9274 0.4 R² = 0.9917 Absorbance 492 nm 492 Absorbance 0.2

0 0 0.5 1 1.5 2 2.5 c (standard phage-displayed IL-2) (display units/mL)

Fig. 5 Example of standard curve to quantitate the levels of phage-displayed proteins by ELISA. Diluted phages displaying human IL-2 (100 arbitrary display units/mL in the original phage preparation) were incubated on a microplate coated with the anti-c-myc 9E10 mAb. Bound phages were detected with an anti-M13 mAb labeled with horseradish peroxidase. The best ﬁtting curve was determined by using a logarithmic regression

Table 1 Example of determination of relative concentrations of phage-displayed IL-2 variants by ELISA

Mean Concentration of diluted Dilution Calculated concentration in the concentration Sample sample (U/mL) factor original sample (U/mL) (U/mL)

1 0.86 100 86 81 0.19 400 76 2 0.93 200 186 177 0.21 800 168 Diluted phage samples were incubated on a microplate coated with the anti-c-myc 9E10 mAb. Bound phages were detected with an anti-M13 mAb labeled with horseradish peroxidase. A standard curve (obtained from a phage prep for which a display level of 100 arbitrary display units/mL was assumed), was simultaneously analyzed. Concentrations of the diluted samples were interpolated from the standard curve and the original concentrations in the samples were calculated taking into account the dilution factors

16. Use the calculated concentrations of the displayed proteins to normalize the amounts of phage-displayed IL-2 variants in subsequent characterization experiments.

3.5 Screening 1. Coat a polyvinyl chloride ELISA microplate with 100 μL/well Receptor Binding of either recombinant alpha or beta IL-2R subunits at 5 μg/mL Properties of Phage- in PBS(see Note 37). Anti-IL-2 monoclonal antibodies and Displayed IL-2 9E10 anti-c-myc mAb can be used to coat additional wells in Variants by ELISA the same conditions in order to work as controls of the 550 Gertrudis Rojas and Tania Carmenate

presence and correct folding of the displayed IL-2 variants; the use of an unrelated coating protein is also recommended to assess nonspecific binding of highly concentrated phage samples (see Note 38). Cover the plate with a lid and incubate during 16–20 h at 4 C. 2. Discard the coating solutions by inverting the plate several times. 3. Add 200 μL/well of 4% M-PBS to block the plate. Cover the plate with a lid and incubate for 30 min at RT. 4. Discard the blocking solution by inverting the plates several times. 5. Add 100 μL/well of phage samples, properly diluted in M-PBS, to wells coated with the different molecules (see Note 39). Dilutions should be selected to achieve the same concentrations (in display units/mL) for all the phage- displayed IL-2 variants to be compared (see Note 40). Coat the plates with a lid and incubate during 1 h at RT. 6. Discard the samples by inverting the plate several times and wash the plate at least five times by filling the wells with washing solution with a washing bottle and discarding it (see Note 33). 7. Add 100 μL/well of the anti-M13/peroxidase conjugate (diluted 1/5000 in M-PBS). Coat the plate with a lid and incubate during 1 h at RT. 8. Wash the plate at least eight times as described in step 6 of the current Subheading. 9. Add 100 μL/well of peroxidase substrate solution. Incubate during 15 min at RT. A yellow/orange color should develop in some wells, while the wells containing no phage samples or a non-related coating molecule should remain colorless (indicating low background levels). 10. Stop the reaction with 50 μL/well of the stop solution. 11. Read the absorbances at 490 nm with a microplate reader. 12. Calculate the mean absorbance values from replicates and analyze the results. Figure 6 shows examples of comparisons of binding properties of several phage-displayed IL-2-derived muteins.

3.6 Inducing CTLL- 1. Thaw one CTLL-2 cell vial and grow the cells on supplemented 2 Proliferation by IL-2-containing RPMI medium, at 37 C and 5% of CO2. Keep Phage-Displayed IL-2 cell concentration between 103 and 104 cells/mL. Use 6-well 2 and IL-2-Derived suspension culture plates or 25 cm ﬂasks for suspension cul- Muteins tures (see Note 41). Phage Display of Cytokines 551

Fig. 6 Comparison of the reactivity of different IL-2 variants against IL-2 receptor subunits in ELISA. Plates were coated with either human alpha IL-2R subunit fused to Fc (panel a) or human beta IL-2R chain fused to Fc (panel b). Phage samples (previously normalized according to the levels of displayed IL-2 variants) were incubated on coated plates. Bound phages were detected with an anti-M13 mAb labeled with horseradish peroxidase. NA (R38A, F42A, Y45A, E62A) and SA (V69A, Q74P) are mutated human IL-2 variants exhibiting loss [9] and gain [7] of binding ability to human alpha IL-2R subunit respectively. SB (L80F, R81D, L85V, I86V, I92F) shows an increased afﬁnity to human beta IL-2R chain [8]. Phage-displayed non-mutated IL-2 was used to assess the reference binding levels to receptor subunits

2. Subculture the cells every two days. Centrifuge the cells at 300 Â g during 5 min, discard the supernatant and suspend the cells in fresh supplemented medium with IL-2. Subculture the cells until 95% of viability is reached. 3. To perform the cell proliferation assay, centrifuge the culture at 300 Â g during 5 min, and wash the cells three times with non-supplemented RPMI medium to remove the remaining IL-2 (see Note 42). 4. Use supplemented RPMI medium without IL-2 to dilute the phage samples. 5. Dilute each puriﬁed and ﬁltered phage sample appropriately to obtain the same concentration of displayed IL-2 or its mutated variants (measured as display units/mL) (see Note 43). 6. Add 100 μL/well of diluted phage samples to the 96-well tissue culture plate, use a phage sample displaying an unrelated protein as negative control, and some wells with RPMI medium to assess the assay background. 7. Adjust the CTLL-2 cells to a concentration of 2 Â 105 cells/ mL and add 100 μL of the cell suspension to each well. 8. Incubate the plate at 37 C, 5% CO2, for 48 h. 9. Add 20 μL of Alamar Blue dye (Invitrogen) to each well and incubate the plate in the same conditions for another 12 h. A pink color should appear in those wells with IL-2 or IL-2- derived functional muteins, while the original blue color 552 Gertrudis Rojas and Tania Carmenate

0.7

0.6 hIL-2 0.5 R38A 0.4 F42A

0.3 Y45A

0.2 E62A

0.1 scFv Abs 540 nm –Abs 630 nm Abs

0 0.01 0.1 1 10 c(IL-2 variants) (display units/mL)

Fig. 7 Comparison of the ability of different phage-displayed mutated human IL-2 variants to induce proliferation of CTLL-2 cells. Cells were incubated with diluted phages (previously normalized according to the levels of displayed IL-2 variants) and their proliferation levels were assessed with Alamar blue dye. Non-mutated human phage-displayed IL-2 was used to assess maximal proliferation levels. Phages displaying an unrelated protein (single chain Fv antibody fragment, scFv) were used as the negative control

Table 2 Doses of mutated phage-displayed human IL-2 variants able to induce half-maximal cell proliferation of CTLL-2 cells

Concentrations required to achieve 50% Phage-displayed molecule maximal proliferation of CTLL-2 cells (arbitrary display units/mL)

IL-2 0.13 Æ 0.01 R38A 0.27 Æ 0.05 F42A 6.41 Æ 0.78 Y45A 1.83 Æ 0.18 E62A 9.97 Æ 1.36 Cells were incubated with diluted phages (previously normalized according to the levels of displayed IL-2 variants) and their proliferation levels were assessed with Alamar blue dye. Non-mutated human phage-displayed IL-2 was used to assess maximal proliferation levels. Calculated concentrations represent the average of the values obtained from three independent experiments

should be kept in wells with the non-related phage samples or medium. 10. Read the absorbance at 540 and 630 nm with a microplate reader. 11. Calculate the difference between readings at both wavelengths and the mean value for replicates. Figure 7 shows an example of cell proliferation curves. Table 2 illustrates the results of an experiment designed to calculate the doses of phage-displayed Phage Display of Cytokines 553

IL-2 variants required to induce 50% of maximal cell proliferation.

4 Notes

1. Human and mouse IL-2 receptor alpha subunit (223-2A-025/ CF and 2438-RM-050/CF), as well as human beta subunit (224-2B-025/CF) recombinant extracellular domains can be purchased from R&D (Minneapolis, USA). 2. Extracellular domains of both alpha (CD25) and beta (CD122) subunits can be home-made as recombinant proteins produced by HEK293 cells. Fusion of their coding genes to the gene encoding a human Fc fragment allows high-level expression in human cells and protein A-based affinity purification. 3. Anti-human IL-2 antibodies MAB202 (clone 5334) and MAB602 (clone 5355) can be purchased from R&D. Anti- mouse IL-2 JES6-1A12 (554424), JES-5H4 (554425) and S4B6 (554375) mAbs are supplied by BD Pharmingen™ (New Jersey, USA). Other antibodies, particularly those recog- nizing conformational epitopes, can be used to assess the presence and proper folding of phage-displayed IL-2 molecules. 4. Recombinant IL-2 (34-8029-85) for CTLL-2 cell line culture can be purchased from eBioscience (San Diego, USA). Other IL-2 sources can be used as well, as long as specific activity is equal to or higher than 106 IU/mg. 5. The presence of the first two residues of mature IL-2 (Ala-Pro) has been shown to interfere with periplasmic secretion and phage display of the cytokine [16]. Unpaired Cys can cause the formation of non-natural disulfide bonds resulting in mis- sfolding and/or inter-molecular aggregation of the recombinant proteins. 6. Even though most of the experience displaying IL-2 and IL-2- derived mutated variants has been obtained with pCSM phagemid, other similar vectors can be used instead. Genetic fusion of the displayed cytokine to a tag (like c-myc) recognized by an available monoclonal antibody is a strict requirement for cytokine screening applications, as quantitation of the displayed cytokine variants is necessary to compare their binding and biological properties. 7. The procedure described in steps 5–19 of Subheading 3.1 refers to the production and purification of phages starting from a single colony. It is highly recommendable to obtain simultaneously phages from several colonies (at least two) in order to use independent replicas in characterization experiments. 554 Gertrudis Rojas and Tania Carmenate

8. M13KO7 helper phage stocks are labeled with the phage titer (measured as pfu/mL). 1011 pfu are usually contained in 10 μL of a typical helper phage stock obtained at the lab (1013 pfu/ mL). Depending on the source of helper phage, the titer can vary widely and the required volume needs to be calculated. 9. During the incubation of phage-containing supernatants with PEG/NaCl solution on ice, the formation of a white precipitate (composed mainly by phages) that tends to accumulate at the bottom of the conical tube is usually observed. 10. Since in the second phage precipitation step (in 1.5 mL vials) phages are more concentrated than in the first precipitation step, the formation of a white precipitate is immediately observed after mixing PBS containing phages with the PEG/- NaCl solution. The mixture looks like milk. If it does not happen, implying a failure in phage production, the samples should be discarded and the procedure has to be started again (from step 4 of Subheading 3.1). 11. Purified phage pellets (after E. coli and cell debris have been removed) are very difficult to resuspend. Try to avoid vortexing. Resuspend the pellet by gentle pipetting and (if necessary) leave the pellet in contact with PBS at room temperature with slow shaking until it can be disrupted. 12. Even though purified phage preparations can be filtered using 0.2 μm syringe filters, volume loss during conventional filtration makes it more recommendable to use centrifugal filter devices for such a purpose. Centrifugal force allows recovering most of the phage volume in the flow-through. 13. Purified phage filtration is an optional step. It is not necessary to screen the binding properties of phage-displayed cytokines as contaminant E. coli cells do not interfere with ELISA, but it is strictly required to keep the sterility in cell-based biological assays. Although filtration can result in slightly decreased phage content (see Subheading 3.4), this moderate phage loss does not preclude the use of filtered phages for binding and cell- based assays. 14. CJ236 is the preferred E. coli strain to obtain single-stranded DNA template for site-directed Kunkel mutagenesis due to the ability of this strain to produce uracil-containing template DNA. This feature allows the mutated DNA strand to be preferentially replicated in a conventional E. coli host over the non-mutated uracil-containing template and increases the efficiency of the mutagenesis process (see Subheading 3.3). Despite this advantage, it should be taken into account that phage production is not robust when using CJ236 strain. If the procedure described in Subheading 3.2 fails, replace CJ236 by XL-1 Blue E. coli strain. This strain renders higher amounts of Phage Display of Cytokines 555

single-stranded DNA, and mutagenesis efficiency on such template is acceptable. 15. Phage production from CJ236 E. coli is less reproducible than the same procedure when using other E. coli strains, resulting in a failure to isolate single-stranded DNA from some colonies without any evident reason. That is why, it is highly recommendable to start the procedure described in Subheading 3.2 with at least three or four colonies, in order to be successful with some of them. 16. When using XL-1 Blue E. coli strain instead of CJ236 for single-stranded DNA production, chloramphenicol must be replaced by tetracycline at 10 μg/mL. 17. When using XL-1 Blue E. coli strain instead of CJ236 for single-stranded DNA isolation, uridine is not included during phage production. 18. When using XL-1 Blue E. coli strain instead of CJ236 for single-stranded DNA isolation, phage production is performed at 28 C instead of 37 C. 19. Low phage production by the CJ236 E. coli strain usually results in the absence of any visible precipitate during the incubation of phage-containing supernatants with PEG/NaCl solution on ice, in contrast to what is observed for other E. coli strains like TG1 or XL-1 Blue. 20. During phage precipitation with MP solution, the formation of a white precipitate is sometimes visible. While it can be observed or not when using CJ236 E. coli strain for phage production (and this is not a useful indicator to predict the subsequent success of single-stranded DNA isolation), precipitate formation resulting in the solution becoming completely white must be seen when using XL-1 Blue strain. 21. Successful single-stranded DNA purification results in a main band that can be clearly observed on an agarose gel electrophoresis. Sometimes, it is accompanied by several minor bands with less electrophoretic mobility (see Fig. 2). 22. As pCSM+ phagemid vector is used in the current protocol, the phage replication origin is in the sense strand of the plasmid, resulting in packaging of this strand into filamentous phages and isolation of sense single-stranded DNA template. Anti- sense mutagenic oligonucleotides are thus required to anneal with such a template. When using other phagemid vectors, it is À important to know whether they are + or versions. In the second case, the antisense template DNA would be obtained through the protocol described in Subheading 3.2, and sense mutagenic oligonucleotides directly including the triplet(s) to 556 Gertrudis Rojas and Tania Carmenate

be introduced and flanking IL-2 gene sequences should be designed. 23. When the ends of the mutagenic oligonucleotide coincide with repeated triplets in the template, annealing can occur at the wrong position resulting in undesired insertions or deletions. This can be avoided by extending the mutagenic oligonucleotides a few triplets beyond the repetitive region to guarantee specific annealing. See Fig. 3a for an example of extended oligonucleotide. 24. Distant target sequences can be modified with two or more mutagenic oligonucleotides (each one targeting a given region) as long as the space between such target sequences is above 30 nucleotides, allowing the simultaneous annealing of the mutagenic oligonucleotides to the same template molecule. 25. Even though individual phosphorylation mixes for each mutagenic oligonucleotide could be prepared, it is advantageous to make a master mix with all the common components, and distribute it in different vials to be mixed with each oligonucleotide. This is particularly useful when doing multiple simultaneous mutagenesis reactions. If n phosphorylation reactions are going to be performed, prepare a master mix for an excess of reactions (>n) in order to make sure that the mix is not exhausted before preparing all of them. 26. The use of a PCR microplate instead of individual tubes is particularly advantageous to perform multiple simultaneous mutagenesis reactions. 27. Although individual fill-in reactions could be prepared, it is advantageous to make a master mix to be distributed among all the annealing reactions. This is particularly useful when doing multiple simultaneous mutagenesis reactions. If n fill-in reactions are going to be performed, add enough components to the mix for an excess of reactions (>n) in order to make sure that the mix will be enough for all of them. 28. Typically, at least two bands are observed after the mutagenesis reaction: one with slightly less electrophoretic mobility than the template DNA, which corresponds to the hybrid double- stranded DNA formed by the template strand and the newly synthesized mutated strand (the desired mutagenesis product), and a second one with even lower electrophoretic mobility. The latter comes from strand displacement, an artifact arising during the fill-in reaction, when the mutagenic oligonucleotide primes the synthesis of a new strand as expected, but the synthesis is not stopped when the emerging chain reaches the double strand formed between the template and the mutagenic oligonucleotide. Synthesis thus continues, displacing the Phage Display of Cytokines 557

mutagenic oligonucleotide and the newly formed strand and giving rise to high molecular weight aberrant DNA species (see Fig. 2). Sometimes, a third weak band is observed between the two previously described bands. 29. As the product of a successful mutagenesis reaction is a hybrid molecule composed by the original template strand and a new mutated strand, both strands can in principle be replicated in E. coli. Usually, one of them is preferentially replicated after transformation of E. coli with mutagenesis products, resulting in sequences indicative of either successful mutation at the target position or no mutation at all. In some cases, however, sequencing reveals the coexistence of mutated and non-mutated DNA in the same bacteria (shown by overlapping sequencing peaks at the target position(s) (see Fig. 4). 30. Polyvinyl chloride ELISA microplates can be replaced by polystyrene plates. 31. Take one preparation of purified and filtered phages (usually displaying unmodified IL-2) as the standard to assess the relative display levels of the other phages. Assume that the display level in the standard is 100 arbitrary display units/ml. Make serial two-fold dilutions of this preparation (from 1/25 to 1/6400) in M-PBS to obtain a standard curve. 32. Use several dilutions of samples with unknown levels of displayed IL-2 variants (from 1/100 to 1/1600). If these dilutions do not produce absorbances corresponding to the linear range of the standard curve, select other suitable dilutions to be prepared. If the samples are going to be filtered before use, it is important to quantify their display levels after filtration and not before, as this process can result in a decrease of phage content. 33. When filling the ELISA plate wells with washing solution, make sure that all the wells are totally filled and no air burbles are preventing their surface from being washed. Exhaustive washing of every well guarantees low background levels and reproducibility of the results. 34. Multiple softwares are available to obtain the best fitting curve from a series of standard dilutions. Linear regression is often preferred for simplicity, but sometimes four parameter logistic (4-PL) or logarithmic regression produces an optimal fit of the data. You can also remove the highest and lowest standard concentrations to focus on the linear range of the standard curve. Fitting quality is assessed by the R2 value, which must always be higher than 0.99. See Fig. 5 for an example of standard curve. 35. Those sample dilutions producing absorbances corresponding to the linear range of the standard curve are optimal to calculate the concentrations of the displayed proteins. 558 Gertrudis Rojas and Tania Carmenate

36. If several dilutions of the same phage sample have been assayed, its display level is calculated as the average of the calculated display levels for each dilution (see Table 1 for an example). Sometimes, the values obtained from the extreme dilutions (either highest or lowest) strongly deviate from the rest. In these cases, such dilutions must be excluded from the analysis. 37. Polyvinyl chloride plates can be replaced by polystyrene microplates. Coating concentrations can vary from 1 to 10 μg/mL, depending on the availability of the coating molecules and the absorbances obtained at the end of the assay. 38. While anti-c-myc 9E10 mAb can be used to assess the presence in equivalent amounts of every mutated IL-2 variant, anti-IL- 2 mAbs, particularly those against conformational epitopes, can be used to guarantee the proper folding of the mutated variants on the surface of filamentous phages. It is important to analyze such results carefully (taking into account the fine specificity of each antibody), as mutations affecting the epitope recognized by a given mAb can result in a drastic reduction of reactivity without implying folding problems. The use of an unrelated coating protein provides a negative control, in order to assess the specificity of positive reactions. Highly concentrated phage samples can produce nonspecific background on such non-related coating molecules. If this happens, such dilutions should not be considered when analyzing results. 39. Dilutions of phage samples must be prepared in a volume high enough to be applied on wells coated with each coating molecule (IL-2R subunits, anti-IL-2 antibodies, anti-c-myc mAb, and the unrelated protein). The number of different coating proteins depends upon the experimental design. 40. Comparison of the binding properties of different IL-2 variants strictly depends on the use of equivalent amounts of all of them. Therefore, normalization of the diverse phage preps according to the content of the displayed protein(s) as described in Subheading 3.4 allows the results to be comparable. Calculate the required dilution for each phage prep in order to reach the same final concentration (in display units/ mL). 41. CTLL-2 cells grow in suspension. When the cells are actively growing they tend to form clusters. It is important to keep cell concentration below 2 Â 105 cells/ml. Otherwise, IL-2 in the medium would be rapidly depleted and the cells would lose viability very quickly. It could take up to ten days to obtain a properly growing culture, don’t give up. 42. To determine the activity of phage-displayed IL-2-derived molecules it is important to remove all the IL-2 present in Phage Display of Cytokines 559

the culture. Cells can be maintained without IL-2 up to five hours before performing the cell proliferation assay. 43. Comparison of the biological activity of different IL-2 variants strictly depends on the use of equivalent amounts of all of them. Therefore, normalization of the diverse phage preps according to the content of the displayed protein(s) as described in Subheading 3.4 allows the results to be comparable. Calculate the required dilution for each phage prep in order to reach the same final concentration (in display units/ mL). Do not use dilutions of purified phage samples below 1/10, as too highly concentrated phages can interfere with the cell proliferation assay.

Acknowledgment

G. Rojas is partially supported by a Georg Forster research fellowship awarded by the Alexander von Humboldt Foundation, Germany.

References

1. Smith GP (1985) Filamentous fusion page: Interleukin-2 and regulatory T cells in graft- novel expression vectors that display cloned versus-host disease. N Engl J Med antigens on the virion surface. Science 365:2055–2066 228:1315–1317 7. Rao BM, Driver L, Lauffenburger DA, Wittrup 2. Marks JD, Hoogenboom HR, Bonnert TP, KD (2005) High-affinity CD25-binding IL-2 McCafferty J, Griffiths AD, Winter G (1991) mutants potently stimulate persistent T cell By-passing immunization-human antibodies growth. Biochemistry 44:10696–10701 from V-gene libraries displayed on phage. J 8. Levin AM, Bates DL, Ring AM, Krieg C, Lin Mol Biol 222:581–597 JT, Su L, Moraga I, Raeber ME, Bowman GR, 3. Felici F, Castagnoli L, Musacchio A, Japelli R, Novick P, Pande VS, Fathman CG, Boyman O, Cesareni G (1991) Selection of antibody Garcia KC (2012) Exploiting a natural confor- ligands from a large library of oligopeptides mational switch to engineer an interleukin-2- expressed on a multivalent exposition vector. J ‘superkine’. Nature 484:529–533 Mol Biol 222:301–310 9. Carmenate T, Pacios A, Enamorado M, 4. Smith GP, Petrenko VA (1997) Phage display. Moreno E, Garcia-Martinez K, Fuente D, Chem Rev 97:391–410 Leon K (2013) Human Interleukin-2 mutein 5. Klapper JA, Downey SG, Smith FO, Yang JC, with higher antitumor efficacy than wild type Hughes MS, Kammula US, Sherry RM, Royal IL-2. J Immunol 190:6230–6238 RE, Steinberg SM, Rosenberg S (2008) High- 10. Mitra S, Ring AM, Amarnath S, Spangler JB, dose interleukin-2 for the treatment of meta- Li P, Ju W, Fischer S, Oh J, Spolski R, static renal cell carcinoma: a retrospective anal- Weiskopf K, Kohrt H, Foley JE, ysis of response and survival in patients treated Rajagopalan S, Long EO, Fowler DH, Wald- in the surgery branch at the National Cancer mann TA, Garcia KC, Leonard WJ (2015) Institute between 1986 and 2006. Cancer Interleukin-2 activity can be fine tuned with 113:293–301 engineered receptor signaling clamps. Immu- 6. Koreth J, Phil D, Matsuoka K, Kim HT, Mc nity 42:826–838 Donough SM, Bindra B, Alyea EP, Armand P, 11. Hartley O, Dorgham K, Perez-Bercoff D, Cutler C, Ho VT, Treister NS, Bienfang DC, Cerini F, Heimann A, Gaertner H, Offord Prasad S, Tzachanis D, Joyce RM, Avigan DE, RE, Pancino G, Debre P, Gorochov G (2003) Antin JH, Ritz J, Soiffer RJ (2011) Human Immunodeficiency virus type 1 entry 560 Gertrudis Rojas and Tania Carmenate

inhibitors selected on living cells from a library 16. Vispo NS, Callejo M, Ojalvo AG, Santos A, of phage chemokines. J Virol 77:6637–6644 Chinea G, Gavilondo JV, Aranã MJ (1997) 12. Gram H, Strittmatter U, Lorenz M, Gluck D, Displaying human interleukin-2 on the surface Zenke G (1993) Phage display as a rapid gene of bacteriophage. Immunotechnology expression system: production of bioactive 3:185–193 cytokine-phage and generation of neutralizing 17. Kunkel TA (1985) Rapid and efficient site- monoclonal antibodies. J Immunol Methods specific mutagenesis without phenotypic selec- 161:169–176 tion. Proc Natl Acad Sci U S A 82:488–492 13. Kalie E, Jaitin DA, Abramovich R, Schreiber G 18. Rojas G, Pupo A, Leon K, Avellanet J, (2007) An interferon alpha-2 mutant opti- Carmenate T, Sidhu S (2013) Deciphering mized by phage display for IFNAR1 binding the molecular bases of the biological effects of confers specifically enhanced antitumor activ- antibodies against Interleukin-2: a versatile ities. J Biol Chem 282:11602–11611 platform for fine epitope mapping. Immuno- 14. Shibata H, Yoshioka Y, Ikemizu S, biology 218:105–113 Kobayashi K, Yamamoto Y, Mukai Y, 19. Rojas G, Infante YC, Pupo A, Carmenate T Okamoto T, Taniai M, Kawamura M, Abe Y, (2014) Fine specificity of antibodies against Nakagawa S, Hayakawa T, Nagata S, Interleukin-2 explains their paradoxical immu- Yamagata Y, Mayumi T, Kamada H, Tsutsumi nomodulatory effects. mAbs 6:273–285 Y (2004) Functionalization of tumor necrosis 20. Rojas G, Carmenate T, Leon K (2015) Molec- factor-alpha using phage display technique and ular dissection of the interactions of an antitu- PEGylation improves its antitumor therapeutic mor Interleukin-2-derived mutein on a phage window. Clin Cancer Res 10:8293–8300 display-based platform. J Mol Recognit 15. Buchli PJ, Wu Z, Ciardelli TL (1997) The 28:261–268 functional display of interleukin-2 on filamentous phage. Arch Biochem Biophys 339:79–84 INDEX

A F

® Affilin ...... 205–237 Filamentous phage ...... 4, Affinity maturation...... 21, 67, 84, 149, 50, 240, 251, 349, 358, 521, 536, 538, 541, 542, 160–162, 165, 207, 209, 212–219, 222, 555, 558 227–234, 394, 426 Fragment antigen binding (Fab)...... 4, 20, Antibody...... 3, 25, 45, 61, 83, 113, 25, 29–31, 47, 48, 61, 110, 113, 134, 139–143, 133, 148, 169, 189, 209, 239, 255, 273, 285, 145, 169, 240, 250, 256, 275, 280, 297, 308, 301, 321, 331, 349, 365, 381, 393, 411, 425, 350, 367, 426, 460 447, 463, 479, 497, 535 Fragment crystallizable/constant (Fc) ...... 48, engineering...... 535 260, 263, 264, 266, 514, 515, 551 fragments ...... 4, 9, 47, 134, Framework...... 10, 46, 47, 148, 159, 240, 256, 273–275, 278–283, 285, 50, 84, 151, 152, 154, 240, 350, 371, 394 290, 295, 296, 308, 313–315, 317, 322, 345, 349, 366–368, 381, 395, 397, 401–406, 448, 535 H gene libraries...... 6–10, 12, Heavy chain ...... 4, 9, 10, 26, 18, 20, 84, 85, 107, 239–252, 278, 282, 322 27, 34, 35, 42, 46, 48, 61–80, 85, 99, 114–117, B 119, 123, 128, 140, 147, 170, 176, 183, 185, 190, 192, 240, 426, 430, 431, 435, 448, 451, Biomarker ...... 26, 498, 499 454, 455, 458, 465 Biopanning ...... 134, 143, High throughput selection...... 274, 301 285, 411, 413–420, 422, 463–472, 522 Human libraries...... 10 Bovine libraries...... 113–129 I C IgG reformatting...... 447 Camel libraries...... 169–186, 435 Immune libraries ...... 9, 10, 20, Cell panning ...... 198, 273, 322, 355, 382 21, 26, 27, 33, 34, 40, 42, 83, 88–91, 93, 134, Chain shuffling ...... 62, 67, 79 191, 200, 201, 256, 274, 302, 350 Chicken libraries...... 189–201 Immunoglobulin...... 46, 113–129, Combinatorial libraries ...... 349–362, 414, 416, 463 134, 147, 170, 185, 189, 194, 321, 360, 425, Competition ...... 21, 236, 278, 402 426, 434, 435, 464, 479 Complementarity determining region (CDR) ...... 10, Immunoglobulin G (IgG) ...... 9, 12, 37, 46–48, 50, 51, 57, 72, 84, 115, 148, 186, 190, 19, 26, 29, 30, 34, 40, 42, 48, 87, 92, 94, 108, 240, 242, 251, 351, 360, 394, 498 109, 183, 256, 268, 274, 275, 362, 367, 372, 376, 397, 447–461, 465, 467, 472, 515 D Immunoglobulin new antigen receptor Diagnostic...... 3, 26, 113, (IgNAR)...... 147–165 189, 191, 301, 381, 382, 447, 479, 498, 499 Intrabodies ...... 239, 242, 246 In vitro selection ...... 46, 134, 315, 370 E L Epitope mapping...... 422, 497–516, 536 Error-prone PCR (epPCR) ...... 207, 214, Laser capture microdissection (LCM) ...... 331–345 217–219, 227, 229, 232, 393 Library screening...... 450

Michael Hust and Theam Soon Lim (eds.), Phage Display: Methods and Protocols, Methods in Molecular Biology, vol. 1701, DOI 10.1007/978-1-4939-7447-4, © Springer Science+Business Media LLC 2018 561 PHAGE DISPLAY:METHODS AND PROTOCOLS 562 Index Light chain (LC) ...... 9–11, 20, R 21, 26, 27, 34, 46, 48, 50, 61, 64, 65, 68, 72, 85, 99, 110, 114–119, 122, 123, 128, 139–141, 147, Rabbit libraries ...... 133–145 170, 190, 192, 240, 425, 448, 449, 451, 452, S 454, 455, 458–460 scFv-Fc...... 256, 263, 274, 283 M Screening ...... 58, 76, 84, M13 phage ...... 4, 478 133, 160, 274, 289, 315, 322, 340, 355, 365, Macaque libraries ...... 83–110 394, 412, 447, 464, 483, 501, 520, 535 Metasecretome ...... 479, 519–532 Shark libraries ...... 147–165 Microtiter plate (MTP)...... 75, 170, Single-chain fragment variable (scFv) ...... 3, 6–8, 259, 273–283, 286, 290, 293, 295, 296, 307, 26, 27, 29, 40, 47, 51, 61, 62, 66, 67, 71–73, 79, 308, 310–315, 317, 323, 327, 328, 340, 344, 85, 90, 92, 107, 108, 113, 116–119, 134, 159, 397, 401, 403, 404 169, 189–191, 195–201, 240, 244, 250, 256, Mutagenesis...... 48, 50, 51, 57, 257, 262–265, 268, 274, 275, 277, 280, 282, 207, 218, 219, 240, 242, 244, 246, 247, 251, 297, 308, 315, 322, 327, 332–334, 336, 339, 394, 395, 398, 498, 500, 536, 537, 539, 341–345, 349, 350, 352–360, 362, 366, 367, 542–547, 554, 556, 557 370, 372, 376, 383, 385, 395, 412, 426, 447, 448, 463, 464, 472 N Stability maturation...... 21, 393–406 Streptavidin...... 66, 67, 74, 76, Naive libraries...... 3–22, 25–43, 77, 227, 237, 257, 259, 260, 263, 273, 277, 278, 85, 191, 201, 239, 274, 302, 322, 350, 412, 413, 280, 281, 285–298, 305, 310, 312, 315, 416, 420 334–336, 340, 342, 344, 352, 354, 365, Nanoparticles...... 301–318 371–373, 377, 397, 402 Next-generation sequencing (NGS) ...... 361, Synthetic libraries ...... 9, 10, 26, 370, 411–422, 425–444, 463, 479, 522, 523, 46, 201, 240, 256, 302, 350, 439, 441 525, 526, 528, 530 Nonhuman primates (NHP) ...... 83–85 T

O Therapeutic antibodies ...... 3, 25, 50, 83, 274 Therapy...... 3, 83, 113, 331, Open reading frame selection...... 478, 479 381, 498, 499, 536 ORFeome phage display...... 477–494, 500 Tissue slides panning ...... 382, 383, 387, 390 P Tool antibodies ...... 285, 349 V Panning...... 20, 32, 110, 161, 170, 198, 257, 273, 286, 302, 322, 337, 353, 365, Variable fragment of the heavy chain (VH)...... 5, 6, 382, 394, 413, 429, 463, 478, 500, 522 8–18, 20, 21, 30, 34–37, 42, 43, 61, 62, 64–69, Phage ...... 3, 25, 46, 61, 84, 71–73, 78, 79, 84, 88–91, 93, 99–102, 104–107, 113, 134, 149, 170, 189, 206, 239, 256, 273, 110, 136, 139–141, 169, 170, 183, 240, 244, 285, 302, 321, 332, 349, 365, 381, 393, 411, 246, 251, 367, 448, 449, 451, 452, 455, 458–460 426, 447, 463, 477, 499, 520, 535 Variable fragment of the light chain (VL) ...... 5, Phage-display...... 3, 25, 46, 61, 9, 10, 15–18, 20, 21, 34, 36, 61, 62, 67, 70–73, 84, 113, 134, 149, 170, 193, 206, 247, 256, 273, 78, 79, 84, 89, 91, 99, 101–106, 139, 169, 170, 285, 302, 321, 332, 349, 365, 381, 393, 411, 367 426, 447, 463, 477, 499, 520, 535 VHH ...... 4, 147, 170–184, Phagekines ...... 535–559 186, 240, 250, 435 pMHC-binding antibodies...... 255–268 Protein array screening ...... 365–378