Methods in Enzymology 477 - Guide to techniques in mouse development: Mouse molecular genetics

METHODS IN ENZYMOLOGY Editors-in-Chief

JOHN N. ABELSON AND MELVIN I. SIMON Division of Biology California Institute of Technology Pasadena, California, USA Founding Editors

SIDNEY P. COLOWICK AND NATHAN O. KAPLAN

Academic Press is an imprint of Elsevier 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 32 Jamestown Road, London NW1 7BY, UK First edition 2010 Copyright # 2010, Elsevier Inc. All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: permissions@ elsevier.com. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made For information on all Academic Press publications visit our website at elsevierdirect.com ISBN: 978-0-12-384882-6 (Paperback) ISBN: 978-0-12-384880-2 (Hardback) ISSN: 0076-6879 Printed and bound in United States of America 10 11 12 10 9 8 7 6 5 4 3 2 1

CONTRIBUTORS

Konstantinos Anastassiadis Center for Regenerative Therapy, Technische Universitaet Dresden, Dresden, Germany Richard R. Behringer Department of Genetics, University of Texas M.D. Anderson Cancer Center, Houston, Texas, USA David R. Beier Division of Genetics, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA Kaj Berhardt Genomics, BioInnovationsZentrum, Technische Universitaet Dresden, Dresden, Germany Allan Bradley Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, United Kingdom Joshua M. Brickman MRC Centre for Regenerative Medicine, Institute for Stem Cell Research, School of Biological Sciences, University of Edinburgh, Edinburgh, United Kingdom Hermann Bujard Zentrum fu¨r Molekulare Biologie Heidelberg (ZMBH), Im Neuenheimer Feld, Heidelberg, Germany Tamara Caspary Department of Human Genetics, Emory University School of Medicine, Atlanta, Georgia, USA Wilhelmine N. de Vries The Jackson Laboratory, Bar Harbor, Maine, USA Francesca E. Duncan Department of Obstetrics and Gynecology, Northwestern University, Chicago, Illinois, USA Adam J. Dupuy Department of Anatomy and Cell Biology, Carver College of Medicine, University of Iowa, Iowa City, Iowa, USA xiii

xiv

Contributors

Susan M. Dymecki Department of Genetics, Harvard Medical School, Boston, Massachusetts, USA Roland H. Friedel Department of Neurosurgery, and Department of Developmental and Regenerative Biology, Mount Sinai School of Medicine, New York, USA Jun Fu Genomics, BioInnovationsZentrum, Technische Universitaet Dresden, Dresden, Germany Stefan Glaser Genomics, BioInnovationsZentrum, Technische Universitaet Dresden, Dresden, Germany Manfred Gossen Berlin-Brandenburg Center for Regenerative Therapies (BCRT), Max Delbru¨ck Center for Molecular Medicine (MDC), Berlin, Germany

and

Joel H. Graber The Jackson Laboratory, Bar Harbor, Maine, USA Kyoji Horie Department of Social and Environmental Medicine, Graduate School of Medicine, Osaka University, Suita, Osaka, Japan Alexandra L. Joyner Memorial Sloan-Kettering Cancer Center, New York, USA Monica J. Justice Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas, USA ¨hn Ralf Ku Technical University Munich, Munich, Germany Mito Kanatsu-Shinohara Department of Molecular Genetics, Graduate School of Medicine, Kyoto University, Kyoto, Japan Jun C. Kim Department of Genetics, Harvard Medical School, Boston, Massachusetts, USA Aljoscha Kleinhammer Institute for Developmental Genetics, Helmholtz Center Munich, German Research Center for Environmental Health, Neuherberg, Munich, Germany Barbara B. Knowles Institute of Medical Biology, A*STAR, Immunos, Singapore Minoru S. H. Ko Developmental Genomics and Aging Section, Laboratory of Genetics, National Institute on Aging, NIH, Baltimore, Maryland, USA

Contributors

xv

Chikara Kokubu Center for Advanced Science and Innovation, Osaka University, Suita, Osaka, Japan Andrea Kranz Genomics, BioInnovationsZentrum, Technische Universitaet Dresden, Dresden, Germany Dung-Fang Lee Department of Gene and Cell Medicine, Black Family Stem Cell Institute, Mount Sinai School of Medicine, New York, USA Emilie Legue´ Memorial Sloan-Kettering Cancer Center, New York, USA Ihor R. Lemischka Department of Gene and Cell Medicine, Black Family Stem Cell Institute, Mount Sinai School of Medicine, New York, USA Meng Amy Li Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, United Kingdom Anne E. Peaston The Jackson Laboratory, Bar Harbor, Maine, USA Stephen J. Pettitt Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, United Kingdom Alexander Pfeifer Institute of Pharmacology and Toxicology, Biomedical Center (BMZ), University of Bonn, Germany Yulan Piao Developmental Genomics and Aging Section, Laboratory of Genetics, National Institute on Aging, NIH, Baltimore, Maryland, USA Frank J. Probst Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas, USA Roland Rad Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, United Kingdom Tiongti Lim Institute of Pharmacology and Toxicology, Life & Brain Center, University of Bonn, Germany Russell S. Ray Department of Genetics, Harvard Medical School, Boston, Massachusetts, USA

xvi

Contributors

Christopher S. Raymond Merck & Co., Inc., Rahway, New Jersey, USA Tetsuichiro Saito Department of Developmental Biology, Graduate School of Medicine, Chiba University, Chiba, Japan ¨nig Kai Scho Zentralinstitut fu¨r Seelische Gesundheit, Mannheim, Germany Christoph Schaniel Department of Gene and Cell Medicine, Black Family Stem Cell Institute, Mount Sinai School of Medicine, New York, USA Richard M. Schultz Department of Biology, University of Pennsylvania, Philadelphia, Pennsylvania, USA Frieder Schwenk TaconicArtemis GmbH, Cologne, and University of Applied Science Gelsenkirchen, Recklinghausen, Germany Jost Seibler TaconicArtemis GmbH, Cologne, Germany Alexei A. Sharov Developmental Genomics and Aging Section, Laboratory of Genetics, National Institute on Aging, NIH, Baltimore, Maryland, USA Myung K. Shin Merck & Co., Inc., Rahway, New Jersey, USA Takashi Shinohara Department of Molecular Genetics, Graduate School of Medicine, Kyoto University, Kyoto, Japan William Skarnes Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, England Philippe Soriano Department of Developmental and Regenerative Biology, Mount Sinai School of Medicine, New York, USA William L. Stanford Institute of Biomaterials and Biomedical Engineering, University of Toronto, Toronto, Ontario, Canada A. Francis Stewart Genomics, BioInnovationsZentrum, Technische Universitaet Dresden, Dresden, Germany

Contributors

xvii

M. David Stewart Department of Biology and Biochemistry, University of Houston, Houston, Texas, USA Rolf W. Stottmann Division of Genetics, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA Junji Takeda Center for Advanced Science and Innovation, Osaka University, Suita, Osaka, Japan Madeleine Teucher Genomics, BioInnovationsZentrum, Technische Universitaet Dresden, Dresden, Germany Christine To Institute of Biomaterials and Biomedical Engineering, University of Toronto, Toronto, Ontario, Canada Anestis Tsakiridis MRC Centre for Regenerative Medicine, Institute for Stem Cell Research, School of Biological Sciences, University of Edinburgh, Edinburgh, United Kingdom George Vassiliou Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, United Kingdom Thomas F. Vogt Merck & Co., Inc., Rahway, New Jersey, USA Wolfgang Wurst Technical University Munich, and Max-Planck-Institute of Psychiatry, Molecular Neurogenetics, Munich, Germany Kosuke Yusa Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, United Kingdom Lei Zhu Merck & Co., Inc., Rahway, New Jersey, USA Katrin Zimmermann Institute of Pharmacology and Toxicology, Biomedical Center (BMZ), University of Bonn, Germany

PREFACE

It has been 17 years since the first edition of the ‘‘Guide to Techniques in Mouse Development,’’, Volume 225 of Methods in Enzymology, was published by Academic Press. Needless to say, the development of technology used to investigate mouse development has not stood still during the interim. Enormous advances have occurred in genomics, transgenic and ES cell methodology, and reprogramming, culminating in the development of iPS cells. At both the cellular and molecular levels, there have been a great many technological advances that permit investigators to probe ever more deeply into all aspects of mouse development. Consequently, it appeared to be an appropriate time to publish a completely new version of the Guide, highlighting the technological advances used to study mouse development. As in the first edition of the Guide in 1993—‘‘Our purpose in assembling this volume is to create a source of state-of-the-art experimental approaches in mouse development useful at the laboratory bench to a diverse group of investigators. The aim is to provide investigators with reliable experimental protocols and recipes that are described in sufficient detail by leaders in the field.’’ We believe that these goals have been achieved with the publication of the new version of the Guide in 2010. It is notable that the new version of the Guide is divided into two volumes, whereas the 1993 version was published as a single volume. This change reflects the increased number of topics covered and the increased sophistication of the methodology. In addition, we have not shied away from including articles on the same topic written by authors from different laboratories. For example, multiple chapters cover the use of recombinases and ENU mutagenesis in Sections III and IV, respectively. Users of the first edition found this aspect of the Guide particularly helpful since they could compare protocols on the same or similar topics and choose the methodology just right for them. We sincerely hope that the Guide will find its way into many laboratories and proves to be useful at the bench. We are grateful to the authors for both their contributions and forbearance with the publication schedules. Also, Paul Wassarman thanks Philippe Soriano for agreeing to co-edit the volume, which made the job more enjoyable. PAUL M. WASSARMAN AND PHILIPPE M. SORIANO

xix

METHODS IN ENZYMOLOGY

VOLUME I. Preparation and Assay of Enzymes Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME II. Preparation and Assay of Enzymes Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME III. Preparation and Assay of Substrates Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME IV. Special Techniques for the Enzymologist Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME V. Preparation and Assay of Enzymes Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME VI. Preparation and Assay of Enzymes (Continued) Preparation and Assay of Substrates Special Techniques Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME VII. Cumulative Subject Index Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME VIII. Complex Carbohydrates Edited by ELIZABETH F. NEUFELD AND VICTOR GINSBURG VOLUME IX. Carbohydrate Metabolism Edited by WILLIS A. WOOD VOLUME X. Oxidation and Phosphorylation Edited by RONALD W. ESTABROOK AND MAYNARD E. PULLMAN VOLUME XI. Enzyme Structure Edited by C. H. W. HIRS VOLUME XII. Nucleic Acids (Parts A and B) Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME XIII. Citric Acid Cycle Edited by J. M. LOWENSTEIN VOLUME XIV. Lipids Edited by J. M. LOWENSTEIN VOLUME XV. Steroids and Terpenoids Edited by RAYMOND B. CLAYTON xxi

xxii

Methods in Enzymology

VOLUME XVI. Fast Reactions Edited by KENNETH KUSTIN VOLUME XVII. Metabolism of Amino Acids and Amines (Parts A and B) Edited by HERBERT TABOR AND CELIA WHITE TABOR VOLUME XVIII. Vitamins and Coenzymes (Parts A, B, and C) Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT VOLUME XIX. Proteolytic Enzymes Edited by GERTRUDE E. PERLMANN AND LASZLO LORAND VOLUME XX. Nucleic Acids and Protein Synthesis (Part C) Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME XXI. Nucleic Acids (Part D) Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME XXII. Enzyme Purification and Related Techniques Edited by WILLIAM B. JAKOBY VOLUME XXIII. Photosynthesis (Part A) Edited by ANTHONY SAN PIETRO VOLUME XXIV. Photosynthesis and Nitrogen Fixation (Part B) Edited by ANTHONY SAN PIETRO VOLUME XXV. Enzyme Structure (Part B) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XXVI. Enzyme Structure (Part C) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XXVII. Enzyme Structure (Part D) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XXVIII. Complex Carbohydrates (Part B) Edited by VICTOR GINSBURG VOLUME XXIX. Nucleic Acids and Protein Synthesis (Part E) Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME XXX. Nucleic Acids and Protein Synthesis (Part F) Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME XXXI. Biomembranes (Part A) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME XXXII. Biomembranes (Part B) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME XXXIII. Cumulative Subject Index Volumes I-XXX Edited by MARTHA G. DENNIS AND EDWARD A. DENNIS VOLUME XXXIV. Affinity Techniques (Enzyme Purification: Part B) Edited by WILLIAM B. JAKOBY AND MEIR WILCHEK

Methods in Enzymology

VOLUME XXXV. Lipids (Part B) Edited by JOHN M. LOWENSTEIN VOLUME XXXVI. Hormone Action (Part A: Steroid Hormones) Edited by BERT W. O’MALLEY AND JOEL G. HARDMAN VOLUME XXXVII. Hormone Action (Part B: Peptide Hormones) Edited by BERT W. O’MALLEY AND JOEL G. HARDMAN VOLUME XXXVIII. Hormone Action (Part C: Cyclic Nucleotides) Edited by JOEL G. HARDMAN AND BERT W. O’MALLEY VOLUME XXXIX. Hormone Action (Part D: Isolated Cells, Tissues, and Organ Systems) Edited by JOEL G. HARDMAN AND BERT W. O’MALLEY VOLUME XL. Hormone Action (Part E: Nuclear Structure and Function) Edited by BERT W. O’MALLEY AND JOEL G. HARDMAN VOLUME XLI. Carbohydrate Metabolism (Part B) Edited by W. A. WOOD VOLUME XLII. Carbohydrate Metabolism (Part C) Edited by W. A. WOOD VOLUME XLIII. Antibiotics Edited by JOHN H. HASH VOLUME XLIV. Immobilized Enzymes Edited by KLAUS MOSBACH VOLUME XLV. Proteolytic Enzymes (Part B) Edited by LASZLO LORAND VOLUME XLVI. Affinity Labeling Edited by WILLIAM B. JAKOBY AND MEIR WILCHEK VOLUME XLVII. Enzyme Structure (Part E) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XLVIII. Enzyme Structure (Part F) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XLIX. Enzyme Structure (Part G) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME L. Complex Carbohydrates (Part C) Edited by VICTOR GINSBURG VOLUME LI. Purine and Pyrimidine Nucleotide Metabolism Edited by PATRICIA A. HOFFEE AND MARY ELLEN JONES VOLUME LII. Biomembranes (Part C: Biological Oxidations) Edited by SIDNEY FLEISCHER AND LESTER PACKER

xxiii

xxiv

Methods in Enzymology

VOLUME LIII. Biomembranes (Part D: Biological Oxidations) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LIV. Biomembranes (Part E: Biological Oxidations) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LV. Biomembranes (Part F: Bioenergetics) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LVI. Biomembranes (Part G: Bioenergetics) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LVII. Bioluminescence and Chemiluminescence Edited by MARLENE A. DELUCA VOLUME LVIII. Cell Culture Edited by WILLIAM B. JAKOBY AND IRA PASTAN VOLUME LIX. Nucleic Acids and Protein Synthesis (Part G) Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME LX. Nucleic Acids and Protein Synthesis (Part H) Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME 61. Enzyme Structure (Part H) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 62. Vitamins and Coenzymes (Part D) Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT VOLUME 63. Enzyme Kinetics and Mechanism (Part A: Initial Rate and Inhibitor Methods) Edited by DANIEL L. PURICH VOLUME 64. Enzyme Kinetics and Mechanism (Part B: Isotopic Probes and Complex Enzyme Systems) Edited by DANIEL L. PURICH VOLUME 65. Nucleic Acids (Part I) Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME 66. Vitamins and Coenzymes (Part E) Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT VOLUME 67. Vitamins and Coenzymes (Part F) Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT VOLUME 68. Recombinant DNA Edited by RAY WU VOLUME 69. Photosynthesis and Nitrogen Fixation (Part C) Edited by ANTHONY SAN PIETRO VOLUME 70. Immunochemical Techniques (Part A) Edited by HELEN VAN VUNAKIS AND JOHN J. LANGONE

Methods in Enzymology

xxv

VOLUME 71. Lipids (Part C) Edited by JOHN M. LOWENSTEIN VOLUME 72. Lipids (Part D) Edited by JOHN M. LOWENSTEIN VOLUME 73. Immunochemical Techniques (Part B) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 74. Immunochemical Techniques (Part C) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 75. Cumulative Subject Index Volumes XXXI, XXXII, XXXIV–LX Edited by EDWARD A. DENNIS AND MARTHA G. DENNIS VOLUME 76. Hemoglobins Edited by ERALDO ANTONINI, LUIGI ROSSI-BERNARDI, AND EMILIA CHIANCONE VOLUME 77. Detoxication and Drug Metabolism Edited by WILLIAM B. JAKOBY VOLUME 78. Interferons (Part A) Edited by SIDNEY PESTKA VOLUME 79. Interferons (Part B) Edited by SIDNEY PESTKA VOLUME 80. Proteolytic Enzymes (Part C) Edited by LASZLO LORAND VOLUME 81. Biomembranes (Part H: Visual Pigments and Purple Membranes, I) Edited by LESTER PACKER VOLUME 82. Structural and Contractile Proteins (Part A: Extracellular Matrix) Edited by LEON W. CUNNINGHAM AND DIXIE W. FREDERIKSEN VOLUME 83. Complex Carbohydrates (Part D) Edited by VICTOR GINSBURG VOLUME 84. Immunochemical Techniques (Part D: Selected Immunoassays) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 85. Structural and Contractile Proteins (Part B: The Contractile Apparatus and the Cytoskeleton) Edited by DIXIE W. FREDERIKSEN AND LEON W. CUNNINGHAM VOLUME 86. Prostaglandins and Arachidonate Metabolites Edited by WILLIAM E. M. LANDS AND WILLIAM L. SMITH VOLUME 87. Enzyme Kinetics and Mechanism (Part C: Intermediates, Stereo-chemistry, and Rate Studies) Edited by DANIEL L. PURICH VOLUME 88. Biomembranes (Part I: Visual Pigments and Purple Membranes, II) Edited by LESTER PACKER

xxvi

Methods in Enzymology

VOLUME 89. Carbohydrate Metabolism (Part D) Edited by WILLIS A. WOOD VOLUME 90. Carbohydrate Metabolism (Part E) Edited by WILLIS A. WOOD VOLUME 91. Enzyme Structure (Part I) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 92. Immunochemical Techniques (Part E: Monoclonal Antibodies and General Immunoassay Methods) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 93. Immunochemical Techniques (Part F: Conventional Antibodies, Fc Receptors, and Cytotoxicity) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 94. Polyamines Edited by HERBERT TABOR AND CELIA WHITE TABOR VOLUME 95. Cumulative Subject Index Volumes 61–74, 76–80 Edited by EDWARD A. DENNIS AND MARTHA G. DENNIS VOLUME 96. Biomembranes [Part J: Membrane Biogenesis: Assembly and Targeting (General Methods; Eukaryotes)] Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 97. Biomembranes [Part K: Membrane Biogenesis: Assembly and Targeting (Prokaryotes, Mitochondria, and Chloroplasts)] Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 98. Biomembranes (Part L: Membrane Biogenesis: Processing and Recycling) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 99. Hormone Action (Part F: Protein Kinases) Edited by JACKIE D. CORBIN AND JOEL G. HARDMAN VOLUME 100. Recombinant DNA (Part B) Edited by RAY WU, LAWRENCE GROSSMAN, AND KIVIE MOLDAVE VOLUME 101. Recombinant DNA (Part C) Edited by RAY WU, LAWRENCE GROSSMAN, AND KIVIE MOLDAVE VOLUME 102. Hormone Action (Part G: Calmodulin and Calcium-Binding Proteins) Edited by ANTHONY R. MEANS AND BERT W. O’MALLEY VOLUME 103. Hormone Action (Part H: Neuroendocrine Peptides) Edited by P. MICHAEL CONN VOLUME 104. Enzyme Purification and Related Techniques (Part C) Edited by WILLIAM B. JAKOBY

Methods in Enzymology

xxvii

VOLUME 105. Oxygen Radicals in Biological Systems Edited by LESTER PACKER VOLUME 106. Posttranslational Modifications (Part A) Edited by FINN WOLD AND KIVIE MOLDAVE VOLUME 107. Posttranslational Modifications (Part B) Edited by FINN WOLD AND KIVIE MOLDAVE VOLUME 108. Immunochemical Techniques (Part G: Separation and Characterization of Lymphoid Cells) Edited by GIOVANNI DI SABATO, JOHN J. LANGONE, AND HELEN VAN VUNAKIS VOLUME 109. Hormone Action (Part I: Peptide Hormones) Edited by LUTZ BIRNBAUMER AND BERT W. O’MALLEY VOLUME 110. Steroids and Isoprenoids (Part A) Edited by JOHN H. LAW AND HANS C. RILLING VOLUME 111. Steroids and Isoprenoids (Part B) Edited by JOHN H. LAW AND HANS C. RILLING VOLUME 112. Drug and Enzyme Targeting (Part A) Edited by KENNETH J. WIDDER AND RALPH GREEN VOLUME 113. Glutamate, Glutamine, Glutathione, and Related Compounds Edited by ALTON MEISTER VOLUME 114. Diffraction Methods for Biological Macromolecules (Part A) Edited by HAROLD W. WYCKOFF, C. H. W. HIRS, AND SERGE N. TIMASHEFF VOLUME 115. Diffraction Methods for Biological Macromolecules (Part B) Edited by HAROLD W. WYCKOFF, C. H. W. HIRS, AND SERGE N. TIMASHEFF VOLUME 116. Immunochemical Techniques (Part H: Effectors and Mediators of Lymphoid Cell Functions) Edited by GIOVANNI DI SABATO, JOHN J. LANGONE, AND HELEN VAN VUNAKIS VOLUME 117. Enzyme Structure (Part J) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 118. Plant Molecular Biology Edited by ARTHUR WEISSBACH AND HERBERT WEISSBACH VOLUME 119. Interferons (Part C) Edited by SIDNEY PESTKA VOLUME 120. Cumulative Subject Index Volumes 81–94, 96–101 VOLUME 121. Immunochemical Techniques (Part I: Hybridoma Technology and Monoclonal Antibodies) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 122. Vitamins and Coenzymes (Part G) Edited by FRANK CHYTIL AND DONALD B. MCCORMICK

xxviii

Methods in Enzymology

VOLUME 123. Vitamins and Coenzymes (Part H) Edited by FRANK CHYTIL AND DONALD B. MCCORMICK VOLUME 124. Hormone Action (Part J: Neuroendocrine Peptides) Edited by P. MICHAEL CONN VOLUME 125. Biomembranes (Part M: Transport in Bacteria, Mitochondria, and Chloroplasts: General Approaches and Transport Systems) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 126. Biomembranes (Part N: Transport in Bacteria, Mitochondria, and Chloroplasts: Protonmotive Force) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 127. Biomembranes (Part O: Protons and Water: Structure and Translocation) Edited by LESTER PACKER VOLUME 128. Plasma Lipoproteins (Part A: Preparation, Structure, and Molecular Biology) Edited by JERE P. SEGREST AND JOHN J. ALBERS VOLUME 129. Plasma Lipoproteins (Part B: Characterization, Cell Biology, and Metabolism) Edited by JOHN J. ALBERS AND JERE P. SEGREST VOLUME 130. Enzyme Structure (Part K) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 131. Enzyme Structure (Part L) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 132. Immunochemical Techniques (Part J: Phagocytosis and Cell-Mediated Cytotoxicity) Edited by GIOVANNI DI SABATO AND JOHANNES EVERSE VOLUME 133. Bioluminescence and Chemiluminescence (Part B) Edited by MARLENE DELUCA AND WILLIAM D. MCELROY VOLUME 134. Structural and Contractile Proteins (Part C: The Contractile Apparatus and the Cytoskeleton) Edited by RICHARD B. VALLEE VOLUME 135. Immobilized Enzymes and Cells (Part B) Edited by KLAUS MOSBACH VOLUME 136. Immobilized Enzymes and Cells (Part C) Edited by KLAUS MOSBACH VOLUME 137. Immobilized Enzymes and Cells (Part D) Edited by KLAUS MOSBACH VOLUME 138. Complex Carbohydrates (Part E) Edited by VICTOR GINSBURG

Methods in Enzymology

xxix

VOLUME 139. Cellular Regulators (Part A: Calcium- and Calmodulin-Binding Proteins) Edited by ANTHONY R. MEANS AND P. MICHAEL CONN VOLUME 140. Cumulative Subject Index Volumes 102–119, 121–134 VOLUME 141. Cellular Regulators (Part B: Calcium and Lipids) Edited by P. MICHAEL CONN AND ANTHONY R. MEANS VOLUME 142. Metabolism of Aromatic Amino Acids and Amines Edited by SEYMOUR KAUFMAN VOLUME 143. Sulfur and Sulfur Amino Acids Edited by WILLIAM B. JAKOBY AND OWEN GRIFFITH VOLUME 144. Structural and Contractile Proteins (Part D: Extracellular Matrix) Edited by LEON W. CUNNINGHAM VOLUME 145. Structural and Contractile Proteins (Part E: Extracellular Matrix) Edited by LEON W. CUNNINGHAM VOLUME 146. Peptide Growth Factors (Part A) Edited by DAVID BARNES AND DAVID A. SIRBASKU VOLUME 147. Peptide Growth Factors (Part B) Edited by DAVID BARNES AND DAVID A. SIRBASKU VOLUME 148. Plant Cell Membranes Edited by LESTER PACKER AND ROLAND DOUCE VOLUME 149. Drug and Enzyme Targeting (Part B) Edited by RALPH GREEN AND KENNETH J. WIDDER VOLUME 150. Immunochemical Techniques (Part K: In Vitro Models of B and T Cell Functions and Lymphoid Cell Receptors) Edited by GIOVANNI DI SABATO VOLUME 151. Molecular Genetics of Mammalian Cells Edited by MICHAEL M. GOTTESMAN VOLUME 152. Guide to Molecular Cloning Techniques Edited by SHELBY L. BERGER AND ALAN R. KIMMEL VOLUME 153. Recombinant DNA (Part D) Edited by RAY WU AND LAWRENCE GROSSMAN VOLUME 154. Recombinant DNA (Part E) Edited by RAY WU AND LAWRENCE GROSSMAN VOLUME 155. Recombinant DNA (Part F) Edited by RAY WU VOLUME 156. Biomembranes (Part P: ATP-Driven Pumps and Related Transport: The Na, K-Pump) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER

xxx

Methods in Enzymology

VOLUME 157. Biomembranes (Part Q: ATP-Driven Pumps and Related Transport: Calcium, Proton, and Potassium Pumps) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 158. Metalloproteins (Part A) Edited by JAMES F. RIORDAN AND BERT L. VALLEE VOLUME 159. Initiation and Termination of Cyclic Nucleotide Action Edited by JACKIE D. CORBIN AND ROGER A. JOHNSON VOLUME 160. Biomass (Part A: Cellulose and Hemicellulose) Edited by WILLIS A. WOOD AND SCOTT T. KELLOGG VOLUME 161. Biomass (Part B: Lignin, Pectin, and Chitin) Edited by WILLIS A. WOOD AND SCOTT T. KELLOGG VOLUME 162. Immunochemical Techniques (Part L: Chemotaxis and Inflammation) Edited by GIOVANNI DI SABATO VOLUME 163. Immunochemical Techniques (Part M: Chemotaxis and Inflammation) Edited by GIOVANNI DI SABATO VOLUME 164. Ribosomes Edited by HARRY F. NOLLER, JR., AND KIVIE MOLDAVE VOLUME 165. Microbial Toxins: Tools for Enzymology Edited by SIDNEY HARSHMAN VOLUME 166. Branched-Chain Amino Acids Edited by ROBERT HARRIS AND JOHN R. SOKATCH VOLUME 167. Cyanobacteria Edited by LESTER PACKER AND ALEXANDER N. GLAZER VOLUME 168. Hormone Action (Part K: Neuroendocrine Peptides) Edited by P. MICHAEL CONN VOLUME 169. Platelets: Receptors, Adhesion, Secretion (Part A) Edited by JACEK HAWIGER VOLUME 170. Nucleosomes Edited by PAUL M. WASSARMAN AND ROGER D. KORNBERG VOLUME 171. Biomembranes (Part R: Transport Theory: Cells and Model Membranes) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 172. Biomembranes (Part S: Transport: Membrane Isolation and Characterization) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER

Methods in Enzymology

xxxi

VOLUME 173. Biomembranes [Part T: Cellular and Subcellular Transport: Eukaryotic (Nonepithelial) Cells] Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 174. Biomembranes [Part U: Cellular and Subcellular Transport: Eukaryotic (Nonepithelial) Cells] Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 175. Cumulative Subject Index Volumes 135–139, 141–167 VOLUME 176. Nuclear Magnetic Resonance (Part A: Spectral Techniques and Dynamics) Edited by NORMAN J. OPPENHEIMER AND THOMAS L. JAMES VOLUME 177. Nuclear Magnetic Resonance (Part B: Structure and Mechanism) Edited by NORMAN J. OPPENHEIMER AND THOMAS L. JAMES VOLUME 178. Antibodies, Antigens, and Molecular Mimicry Edited by JOHN J. LANGONE VOLUME 179. Complex Carbohydrates (Part F) Edited by VICTOR GINSBURG VOLUME 180. RNA Processing (Part A: General Methods) Edited by JAMES E. DAHLBERG AND JOHN N. ABELSON VOLUME 181. RNA Processing (Part B: Specific Methods) Edited by JAMES E. DAHLBERG AND JOHN N. ABELSON VOLUME 182. Guide to Protein Purification Edited by MURRAY P. DEUTSCHER VOLUME 183. Molecular Evolution: Computer Analysis of Protein and Nucleic Acid Sequences Edited by RUSSELL F. DOOLITTLE VOLUME 184. Avidin-Biotin Technology Edited by MEIR WILCHEK AND EDWARD A. BAYER VOLUME 185. Gene Expression Technology Edited by DAVID V. GOEDDEL VOLUME 186. Oxygen Radicals in Biological Systems (Part B: Oxygen Radicals and Antioxidants) Edited by LESTER PACKER AND ALEXANDER N. GLAZER VOLUME 187. Arachidonate Related Lipid Mediators Edited by ROBERT C. MURPHY AND FRANK A. FITZPATRICK VOLUME 188. Hydrocarbons and Methylotrophy Edited by MARY E. LIDSTROM VOLUME 189. Retinoids (Part A: Molecular and Metabolic Aspects) Edited by LESTER PACKER

xxxii

Methods in Enzymology

VOLUME 190. Retinoids (Part B: Cell Differentiation and Clinical Applications) Edited by LESTER PACKER VOLUME 191. Biomembranes (Part V: Cellular and Subcellular Transport: Epithelial Cells) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 192. Biomembranes (Part W: Cellular and Subcellular Transport: Epithelial Cells) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 193. Mass Spectrometry Edited by JAMES A. MCCLOSKEY VOLUME 194. Guide to Yeast Genetics and Molecular Biology Edited by CHRISTINE GUTHRIE AND GERALD R. FINK VOLUME 195. Adenylyl Cyclase, G Proteins, and Guanylyl Cyclase Edited by ROGER A. JOHNSON AND JACKIE D. CORBIN VOLUME 196. Molecular Motors and the Cytoskeleton Edited by RICHARD B. VALLEE VOLUME 197. Phospholipases Edited by EDWARD A. DENNIS VOLUME 198. Peptide Growth Factors (Part C) Edited by DAVID BARNES, J. P. MATHER, AND GORDON H. SATO VOLUME 199. Cumulative Subject Index Volumes 168–174, 176–194 VOLUME 200. Protein Phosphorylation (Part A: Protein Kinases: Assays, Purification, Antibodies, Functional Analysis, Cloning, and Expression) Edited by TONY HUNTER AND BARTHOLOMEW M. SEFTON VOLUME 201. Protein Phosphorylation (Part B: Analysis of Protein Phosphorylation, Protein Kinase Inhibitors, and Protein Phosphatases) Edited by TONY HUNTER AND BARTHOLOMEW M. SEFTON VOLUME 202. Molecular Design and Modeling: Concepts and Applications (Part A: Proteins, Peptides, and Enzymes) Edited by JOHN J. LANGONE VOLUME 203. Molecular Design and Modeling: Concepts and Applications (Part B: Antibodies and Antigens, Nucleic Acids, Polysaccharides, and Drugs) Edited by JOHN J. LANGONE VOLUME 204. Bacterial Genetic Systems Edited by JEFFREY H. MILLER VOLUME 205. Metallobiochemistry (Part B: Metallothionein and Related Molecules) Edited by JAMES F. RIORDAN AND BERT L. VALLEE

Methods in Enzymology

xxxiii

VOLUME 206. Cytochrome P450 Edited by MICHAEL R. WATERMAN AND ERIC F. JOHNSON VOLUME 207. Ion Channels Edited by BERNARDO RUDY AND LINDA E. IVERSON VOLUME 208. Protein–DNA Interactions Edited by ROBERT T. SAUER VOLUME 209. Phospholipid Biosynthesis Edited by EDWARD A. DENNIS AND DENNIS E. VANCE VOLUME 210. Numerical Computer Methods Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 211. DNA Structures (Part A: Synthesis and Physical Analysis of DNA) Edited by DAVID M. J. LILLEY AND JAMES E. DAHLBERG VOLUME 212. DNA Structures (Part B: Chemical and Electrophoretic Analysis of DNA) Edited by DAVID M. J. LILLEY AND JAMES E. DAHLBERG VOLUME 213. Carotenoids (Part A: Chemistry, Separation, Quantitation, and Antioxidation) Edited by LESTER PACKER VOLUME 214. Carotenoids (Part B: Metabolism, Genetics, and Biosynthesis) Edited by LESTER PACKER VOLUME 215. Platelets: Receptors, Adhesion, Secretion (Part B) Edited by JACEK J. HAWIGER VOLUME 216. Recombinant DNA (Part G) Edited by RAY WU VOLUME 217. Recombinant DNA (Part H) Edited by RAY WU VOLUME 218. Recombinant DNA (Part I) Edited by RAY WU VOLUME 219. Reconstitution of Intracellular Transport Edited by JAMES E. ROTHMAN VOLUME 220. Membrane Fusion Techniques (Part A) Edited by NEJAT DU¨ZGU¨NES, VOLUME 221. Membrane Fusion Techniques (Part B) Edited by NEJAT DU¨ZGU¨NES, VOLUME 222. Proteolytic Enzymes in Coagulation, Fibrinolysis, and Complement Activation (Part A: Mammalian Blood Coagulation Factors and Inhibitors) Edited by LASZLO LORAND AND KENNETH G. MANN

xxxiv

Methods in Enzymology

VOLUME 223. Proteolytic Enzymes in Coagulation, Fibrinolysis, and Complement Activation (Part B: Complement Activation, Fibrinolysis, and Nonmammalian Blood Coagulation Factors) Edited by LASZLO LORAND AND KENNETH G. MANN VOLUME 224. Molecular Evolution: Producing the Biochemical Data Edited by ELIZABETH ANNE ZIMMER, THOMAS J. WHITE, REBECCA L. CANN, AND ALLAN C. WILSON VOLUME 225. Guide to Techniques in Mouse Development Edited by PAUL M. WASSARMAN AND MELVIN L. DEPAMPHILIS VOLUME 226. Metallobiochemistry (Part C: Spectroscopic and Physical Methods for Probing Metal Ion Environments in Metalloenzymes and Metalloproteins) Edited by JAMES F. RIORDAN AND BERT L. VALLEE VOLUME 227. Metallobiochemistry (Part D: Physical and Spectroscopic Methods for Probing Metal Ion Environments in Metalloproteins) Edited by JAMES F. RIORDAN AND BERT L. VALLEE VOLUME 228. Aqueous Two-Phase Systems Edited by HARRY WALTER AND GO¨TE JOHANSSON VOLUME 229. Cumulative Subject Index Volumes 195–198, 200–227 VOLUME 230. Guide to Techniques in Glycobiology Edited by WILLIAM J. LENNARZ AND GERALD W. HART VOLUME 231. Hemoglobins (Part B: Biochemical and Analytical Methods) Edited by JOHANNES EVERSE, KIM D. VANDEGRIFF, AND ROBERT M. WINSLOW VOLUME 232. Hemoglobins (Part C: Biophysical Methods) Edited by JOHANNES EVERSE, KIM D. VANDEGRIFF, AND ROBERT M. WINSLOW VOLUME 233. Oxygen Radicals in Biological Systems (Part C) Edited by LESTER PACKER VOLUME 234. Oxygen Radicals in Biological Systems (Part D) Edited by LESTER PACKER VOLUME 235. Bacterial Pathogenesis (Part A: Identification and Regulation of Virulence Factors) Edited by VIRGINIA L. CLARK AND PATRIK M. BAVOIL VOLUME 236. Bacterial Pathogenesis (Part B: Integration of Pathogenic Bacteria with Host Cells) Edited by VIRGINIA L. CLARK AND PATRIK M. BAVOIL VOLUME 237. Heterotrimeric G Proteins Edited by RAVI IYENGAR VOLUME 238. Heterotrimeric G-Protein Effectors Edited by RAVI IYENGAR

Methods in Enzymology

xxxv

VOLUME 239. Nuclear Magnetic Resonance (Part C) Edited by THOMAS L. JAMES AND NORMAN J. OPPENHEIMER VOLUME 240. Numerical Computer Methods (Part B) Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND VOLUME 241. Retroviral Proteases Edited by LAWRENCE C. KUO AND JULES A. SHAFER VOLUME 242. Neoglycoconjugates (Part A) Edited by Y. C. LEE AND REIKO T. LEE VOLUME 243. Inorganic Microbial Sulfur Metabolism Edited by HARRY D. PECK, JR., AND JEAN LEGALL VOLUME 244. Proteolytic Enzymes: Serine and Cysteine Peptidases Edited by ALAN J. BARRETT VOLUME 245. Extracellular Matrix Components Edited by E. RUOSLAHTI AND E. ENGVALL VOLUME 246. Biochemical Spectroscopy Edited by KENNETH SAUER VOLUME 247. Neoglycoconjugates (Part B: Biomedical Applications) Edited by Y. C. LEE AND REIKO T. LEE VOLUME 248. Proteolytic Enzymes: Aspartic and Metallo Peptidases Edited by ALAN J. BARRETT VOLUME 249. Enzyme Kinetics and Mechanism (Part D: Developments in Enzyme Dynamics) Edited by DANIEL L. PURICH VOLUME 250. Lipid Modifications of Proteins Edited by PATRICK J. CASEY AND JANICE E. BUSS VOLUME 251. Biothiols (Part A: Monothiols and Dithiols, Protein Thiols, and Thiyl Radicals) Edited by LESTER PACKER VOLUME 252. Biothiols (Part B: Glutathione and Thioredoxin; Thiols in Signal Transduction and Gene Regulation) Edited by LESTER PACKER VOLUME 253. Adhesion of Microbial Pathogens Edited by RON J. DOYLE AND ITZHAK OFEK VOLUME 254. Oncogene Techniques Edited by PETER K. VOGT AND INDER M. VERMA VOLUME 255. Small GTPases and Their Regulators (Part A: Ras Family) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL

xxxvi

Methods in Enzymology

VOLUME 256. Small GTPases and Their Regulators (Part B: Rho Family) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 257. Small GTPases and Their Regulators (Part C: Proteins Involved in Transport) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 258. Redox-Active Amino Acids in Biology Edited by JUDITH P. KLINMAN VOLUME 259. Energetics of Biological Macromolecules Edited by MICHAEL L. JOHNSON AND GARY K. ACKERS VOLUME 260. Mitochondrial Biogenesis and Genetics (Part A) Edited by GIUSEPPE M. ATTARDI AND ANNE CHOMYN VOLUME 261. Nuclear Magnetic Resonance and Nucleic Acids Edited by THOMAS L. JAMES VOLUME 262. DNA Replication Edited by JUDITH L. CAMPBELL VOLUME 263. Plasma Lipoproteins (Part C: Quantitation) Edited by WILLIAM A. BRADLEY, SANDRA H. GIANTURCO, AND JERE P. SEGREST VOLUME 264. Mitochondrial Biogenesis and Genetics (Part B) Edited by GIUSEPPE M. ATTARDI AND ANNE CHOMYN VOLUME 265. Cumulative Subject Index Volumes 228, 230–262 VOLUME 266. Computer Methods for Macromolecular Sequence Analysis Edited by RUSSELL F. DOOLITTLE VOLUME 267. Combinatorial Chemistry Edited by JOHN N. ABELSON VOLUME 268. Nitric Oxide (Part A: Sources and Detection of NO; NO Synthase) Edited by LESTER PACKER VOLUME 269. Nitric Oxide (Part B: Physiological and Pathological Processes) Edited by LESTER PACKER VOLUME 270. High Resolution Separation and Analysis of Biological Macromolecules (Part A: Fundamentals) Edited by BARRY L. KARGER AND WILLIAM S. HANCOCK VOLUME 271. High Resolution Separation and Analysis of Biological Macromolecules (Part B: Applications) Edited by BARRY L. KARGER AND WILLIAM S. HANCOCK VOLUME 272. Cytochrome P450 (Part B) Edited by ERIC F. JOHNSON AND MICHAEL R. WATERMAN VOLUME 273. RNA Polymerase and Associated Factors (Part A) Edited by SANKAR ADHYA

Methods in Enzymology

xxxvii

VOLUME 274. RNA Polymerase and Associated Factors (Part B) Edited by SANKAR ADHYA VOLUME 275. Viral Polymerases and Related Proteins Edited by LAWRENCE C. KUO, DAVID B. OLSEN, AND STEVEN S. CARROLL VOLUME 276. Macromolecular Crystallography (Part A) Edited by CHARLES W. CARTER, JR., AND ROBERT M. SWEET VOLUME 277. Macromolecular Crystallography (Part B) Edited by CHARLES W. CARTER, JR., AND ROBERT M. SWEET VOLUME 278. Fluorescence Spectroscopy Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 279. Vitamins and Coenzymes (Part I) Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER VOLUME 280. Vitamins and Coenzymes (Part J) Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER VOLUME 281. Vitamins and Coenzymes (Part K) Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER VOLUME 282. Vitamins and Coenzymes (Part L) Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER VOLUME 283. Cell Cycle Control Edited by WILLIAM G. DUNPHY VOLUME 284. Lipases (Part A: Biotechnology) Edited by BYRON RUBIN AND EDWARD A. DENNIS VOLUME 285. Cumulative Subject Index Volumes 263, 264, 266–284, 286–289 VOLUME 286. Lipases (Part B: Enzyme Characterization and Utilization) Edited by BYRON RUBIN AND EDWARD A. DENNIS VOLUME 287. Chemokines Edited by RICHARD HORUK VOLUME 288. Chemokine Receptors Edited by RICHARD HORUK VOLUME 289. Solid Phase Peptide Synthesis Edited by GREGG B. FIELDS VOLUME 290. Molecular Chaperones Edited by GEORGE H. LORIMER AND THOMAS BALDWIN VOLUME 291. Caged Compounds Edited by GERARD MARRIOTT VOLUME 292. ABC Transporters: Biochemical, Cellular, and Molecular Aspects Edited by SURESH V. AMBUDKAR AND MICHAEL M. GOTTESMAN

xxxviii

Methods in Enzymology

VOLUME 293. Ion Channels (Part B) Edited by P. MICHAEL CONN VOLUME 294. Ion Channels (Part C) Edited by P. MICHAEL CONN VOLUME 295. Energetics of Biological Macromolecules (Part B) Edited by GARY K. ACKERS AND MICHAEL L. JOHNSON VOLUME 296. Neurotransmitter Transporters Edited by SUSAN G. AMARA VOLUME 297. Photosynthesis: Molecular Biology of Energy Capture Edited by LEE MCINTOSH VOLUME 298. Molecular Motors and the Cytoskeleton (Part B) Edited by RICHARD B. VALLEE VOLUME 299. Oxidants and Antioxidants (Part A) Edited by LESTER PACKER VOLUME 300. Oxidants and Antioxidants (Part B) Edited by LESTER PACKER VOLUME 301. Nitric Oxide: Biological and Antioxidant Activities (Part C) Edited by LESTER PACKER VOLUME 302. Green Fluorescent Protein Edited by P. MICHAEL CONN VOLUME 303. cDNA Preparation and Display Edited by SHERMAN M. WEISSMAN VOLUME 304. Chromatin Edited by PAUL M. WASSARMAN AND ALAN P. WOLFFE VOLUME 305. Bioluminescence and Chemiluminescence (Part C) Edited by THOMAS O. BALDWIN AND MIRIAM M. ZIEGLER VOLUME 306. Expression of Recombinant Genes in Eukaryotic Systems Edited by JOSEPH C. GLORIOSO AND MARTIN C. SCHMIDT VOLUME 307. Confocal Microscopy Edited by P. MICHAEL CONN VOLUME 308. Enzyme Kinetics and Mechanism (Part E: Energetics of Enzyme Catalysis) Edited by DANIEL L. PURICH AND VERN L. SCHRAMM VOLUME 309. Amyloid, Prions, and Other Protein Aggregates Edited by RONALD WETZEL VOLUME 310. Biofilms Edited by RON J. DOYLE

Methods in Enzymology

xxxix

VOLUME 311. Sphingolipid Metabolism and Cell Signaling (Part A) Edited by ALFRED H. MERRILL, JR., AND YUSUF A. HANNUN VOLUME 312. Sphingolipid Metabolism and Cell Signaling (Part B) Edited by ALFRED H. MERRILL, JR., AND YUSUF A. HANNUN VOLUME 313. Antisense Technology (Part A: General Methods, Methods of Delivery, and RNA Studies) Edited by M. IAN PHILLIPS VOLUME 314. Antisense Technology (Part B: Applications) Edited by M. IAN PHILLIPS VOLUME 315. Vertebrate Phototransduction and the Visual Cycle (Part A) Edited by KRZYSZTOF PALCZEWSKI VOLUME 316. Vertebrate Phototransduction and the Visual Cycle (Part B) Edited by KRZYSZTOF PALCZEWSKI VOLUME 317. RNA–Ligand Interactions (Part A: Structural Biology Methods) Edited by DANIEL W. CELANDER AND JOHN N. ABELSON VOLUME 318. RNA–Ligand Interactions (Part B: Molecular Biology Methods) Edited by DANIEL W. CELANDER AND JOHN N. ABELSON VOLUME 319. Singlet Oxygen, UV-A, and Ozone Edited by LESTER PACKER AND HELMUT SIES VOLUME 320. Cumulative Subject Index Volumes 290–319 VOLUME 321. Numerical Computer Methods (Part C) Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND VOLUME 322. Apoptosis Edited by JOHN C. REED VOLUME 323. Energetics of Biological Macromolecules (Part C) Edited by MICHAEL L. JOHNSON AND GARY K. ACKERS VOLUME 324. Branched-Chain Amino Acids (Part B) Edited by ROBERT A. HARRIS AND JOHN R. SOKATCH VOLUME 325. Regulators and Effectors of Small GTPases (Part D: Rho Family) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 326. Applications of Chimeric Genes and Hybrid Proteins (Part A: Gene Expression and Protein Purification) Edited by JEREMY THORNER, SCOTT D. EMR, AND JOHN N. ABELSON VOLUME 327. Applications of Chimeric Genes and Hybrid Proteins (Part B: Cell Biology and Physiology) Edited by JEREMY THORNER, SCOTT D. EMR, AND JOHN N. ABELSON

xl

Methods in Enzymology

VOLUME 328. Applications of Chimeric Genes and Hybrid Proteins (Part C: Protein–Protein Interactions and Genomics) Edited by JEREMY THORNER, SCOTT D. EMR, AND JOHN N. ABELSON VOLUME 329. Regulators and Effectors of Small GTPases (Part E: GTPases Involved in Vesicular Traffic) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 330. Hyperthermophilic Enzymes (Part A) Edited by MICHAEL W. W. ADAMS AND ROBERT M. KELLY VOLUME 331. Hyperthermophilic Enzymes (Part B) Edited by MICHAEL W. W. ADAMS AND ROBERT M. KELLY VOLUME 332. Regulators and Effectors of Small GTPases (Part F: Ras Family I) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 333. Regulators and Effectors of Small GTPases (Part G: Ras Family II) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 334. Hyperthermophilic Enzymes (Part C) Edited by MICHAEL W. W. ADAMS AND ROBERT M. KELLY VOLUME 335. Flavonoids and Other Polyphenols Edited by LESTER PACKER VOLUME 336. Microbial Growth in Biofilms (Part A: Developmental and Molecular Biological Aspects) Edited by RON J. DOYLE VOLUME 337. Microbial Growth in Biofilms (Part B: Special Environments and Physicochemical Aspects) Edited by RON J. DOYLE VOLUME 338. Nuclear Magnetic Resonance of Biological Macromolecules (Part A) Edited by THOMAS L. JAMES, VOLKER DO¨TSCH, AND ULI SCHMITZ VOLUME 339. Nuclear Magnetic Resonance of Biological Macromolecules (Part B) Edited by THOMAS L. JAMES, VOLKER DO¨TSCH, AND ULI SCHMITZ VOLUME 340. Drug–Nucleic Acid Interactions Edited by JONATHAN B. CHAIRES AND MICHAEL J. WARING VOLUME 341. Ribonucleases (Part A) Edited by ALLEN W. NICHOLSON VOLUME 342. Ribonucleases (Part B) Edited by ALLEN W. NICHOLSON VOLUME 343. G Protein Pathways (Part A: Receptors) Edited by RAVI IYENGAR AND JOHN D. HILDEBRANDT VOLUME 344. G Protein Pathways (Part B: G Proteins and Their Regulators) Edited by RAVI IYENGAR AND JOHN D. HILDEBRANDT

Methods in Enzymology

xli

VOLUME 345. G Protein Pathways (Part C: Effector Mechanisms) Edited by RAVI IYENGAR AND JOHN D. HILDEBRANDT VOLUME 346. Gene Therapy Methods Edited by M. IAN PHILLIPS VOLUME 347. Protein Sensors and Reactive Oxygen Species (Part A: Selenoproteins and Thioredoxin) Edited by HELMUT SIES AND LESTER PACKER VOLUME 348. Protein Sensors and Reactive Oxygen Species (Part B: Thiol Enzymes and Proteins) Edited by HELMUT SIES AND LESTER PACKER VOLUME 349. Superoxide Dismutase Edited by LESTER PACKER VOLUME 350. Guide to Yeast Genetics and Molecular and Cell Biology (Part B) Edited by CHRISTINE GUTHRIE AND GERALD R. FINK VOLUME 351. Guide to Yeast Genetics and Molecular and Cell Biology (Part C) Edited by CHRISTINE GUTHRIE AND GERALD R. FINK VOLUME 352. Redox Cell Biology and Genetics (Part A) Edited by CHANDAN K. SEN AND LESTER PACKER VOLUME 353. Redox Cell Biology and Genetics (Part B) Edited by CHANDAN K. SEN AND LESTER PACKER VOLUME 354. Enzyme Kinetics and Mechanisms (Part F: Detection and Characterization of Enzyme Reaction Intermediates) Edited by DANIEL L. PURICH VOLUME 355. Cumulative Subject Index Volumes 321–354 VOLUME 356. Laser Capture Microscopy and Microdissection Edited by P. MICHAEL CONN VOLUME 357. Cytochrome P450, Part C Edited by ERIC F. JOHNSON AND MICHAEL R. WATERMAN VOLUME 358. Bacterial Pathogenesis (Part C: Identification, Regulation, and Function of Virulence Factors) Edited by VIRGINIA L. CLARK AND PATRIK M. BAVOIL VOLUME 359. Nitric Oxide (Part D) Edited by ENRIQUE CADENAS AND LESTER PACKER VOLUME 360. Biophotonics (Part A) Edited by GERARD MARRIOTT AND IAN PARKER VOLUME 361. Biophotonics (Part B) Edited by GERARD MARRIOTT AND IAN PARKER

xlii

Methods in Enzymology

VOLUME 362. Recognition of Carbohydrates in Biological Systems (Part A) Edited by YUAN C. LEE AND REIKO T. LEE VOLUME 363. Recognition of Carbohydrates in Biological Systems (Part B) Edited by YUAN C. LEE AND REIKO T. LEE VOLUME 364. Nuclear Receptors Edited by DAVID W. RUSSELL AND DAVID J. MANGELSDORF VOLUME 365. Differentiation of Embryonic Stem Cells Edited by PAUL M. WASSAUMAN AND GORDON M. KELLER VOLUME 366. Protein Phosphatases Edited by SUSANNE KLUMPP AND JOSEF KRIEGLSTEIN VOLUME 367. Liposomes (Part A) Edited by NEJAT DU¨ZGU¨NES, VOLUME 368. Macromolecular Crystallography (Part C) Edited by CHARLES W. CARTER, JR., AND ROBERT M. SWEET VOLUME 369. Combinational Chemistry (Part B) Edited by GUILLERMO A. MORALES AND BARRY A. BUNIN VOLUME 370. RNA Polymerases and Associated Factors (Part C) Edited by SANKAR L. ADHYA AND SUSAN GARGES VOLUME 371. RNA Polymerases and Associated Factors (Part D) Edited by SANKAR L. ADHYA AND SUSAN GARGES VOLUME 372. Liposomes (Part B) Edited by NEJAT DU¨ZGU¨NES, VOLUME 373. Liposomes (Part C) Edited by NEJAT DU¨ZGU¨NES, VOLUME 374. Macromolecular Crystallography (Part D) Edited by CHARLES W. CARTER, JR., AND ROBERT W. SWEET VOLUME 375. Chromatin and Chromatin Remodeling Enzymes (Part A) Edited by C. DAVID ALLIS AND CARL WU VOLUME 376. Chromatin and Chromatin Remodeling Enzymes (Part B) Edited by C. DAVID ALLIS AND CARL WU VOLUME 377. Chromatin and Chromatin Remodeling Enzymes (Part C) Edited by C. DAVID ALLIS AND CARL WU VOLUME 378. Quinones and Quinone Enzymes (Part A) Edited by HELMUT SIES AND LESTER PACKER VOLUME 379. Energetics of Biological Macromolecules (Part D) Edited by JO M. HOLT, MICHAEL L. JOHNSON, AND GARY K. ACKERS VOLUME 380. Energetics of Biological Macromolecules (Part E) Edited by JO M. HOLT, MICHAEL L. JOHNSON, AND GARY K. ACKERS

Methods in Enzymology

xliii

VOLUME 381. Oxygen Sensing Edited by CHANDAN K. SEN AND GREGG L. SEMENZA VOLUME 382. Quinones and Quinone Enzymes (Part B) Edited by HELMUT SIES AND LESTER PACKER VOLUME 383. Numerical Computer Methods (Part D) Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 384. Numerical Computer Methods (Part E) Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 385. Imaging in Biological Research (Part A) Edited by P. MICHAEL CONN VOLUME 386. Imaging in Biological Research (Part B) Edited by P. MICHAEL CONN VOLUME 387. Liposomes (Part D) Edited by NEJAT DU¨ZGU¨NES, VOLUME 388. Protein Engineering Edited by DAN E. ROBERTSON AND JOSEPH P. NOEL VOLUME 389. Regulators of G-Protein Signaling (Part A) Edited by DAVID P. SIDEROVSKI VOLUME 390. Regulators of G-Protein Signaling (Part B) Edited by DAVID P. SIDEROVSKI VOLUME 391. Liposomes (Part E) Edited by NEJAT DU¨ZGU¨NES, VOLUME 392. RNA Interference Edited by ENGELKE ROSSI VOLUME 393. Circadian Rhythms Edited by MICHAEL W. YOUNG VOLUME 394. Nuclear Magnetic Resonance of Biological Macromolecules (Part C) Edited by THOMAS L. JAMES VOLUME 395. Producing the Biochemical Data (Part B) Edited by ELIZABETH A. ZIMMER AND ERIC H. ROALSON VOLUME 396. Nitric Oxide (Part E) Edited by LESTER PACKER AND ENRIQUE CADENAS VOLUME 397. Environmental Microbiology Edited by JARED R. LEADBETTER VOLUME 398. Ubiquitin and Protein Degradation (Part A) Edited by RAYMOND J. DESHAIES VOLUME 399. Ubiquitin and Protein Degradation (Part B) Edited by RAYMOND J. DESHAIES

xliv

Methods in Enzymology

VOLUME 400. Phase II Conjugation Enzymes and Transport Systems Edited by HELMUT SIES AND LESTER PACKER VOLUME 401. Glutathione Transferases and Gamma Glutamyl Transpeptidases Edited by HELMUT SIES AND LESTER PACKER VOLUME 402. Biological Mass Spectrometry Edited by A. L. BURLINGAME VOLUME 403. GTPases Regulating Membrane Targeting and Fusion Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 404. GTPases Regulating Membrane Dynamics Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 405. Mass Spectrometry: Modified Proteins and Glycoconjugates Edited by A. L. BURLINGAME VOLUME 406. Regulators and Effectors of Small GTPases: Rho Family Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 407. Regulators and Effectors of Small GTPases: Ras Family Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 408. DNA Repair (Part A) Edited by JUDITH L. CAMPBELL AND PAUL MODRICH VOLUME 409. DNA Repair (Part B) Edited by JUDITH L. CAMPBELL AND PAUL MODRICH VOLUME 410. DNA Microarrays (Part A: Array Platforms and Web-Bench Protocols) Edited by ALAN KIMMEL AND BRIAN OLIVER VOLUME 411. DNA Microarrays (Part B: Databases and Statistics) Edited by ALAN KIMMEL AND BRIAN OLIVER VOLUME 412. Amyloid, Prions, and Other Protein Aggregates (Part B) Edited by INDU KHETERPAL AND RONALD WETZEL VOLUME 413. Amyloid, Prions, and Other Protein Aggregates (Part C) Edited by INDU KHETERPAL AND RONALD WETZEL VOLUME 414. Measuring Biological Responses with Automated Microscopy Edited by JAMES INGLESE VOLUME 415. Glycobiology Edited by MINORU FUKUDA VOLUME 416. Glycomics Edited by MINORU FUKUDA VOLUME 417. Functional Glycomics Edited by MINORU FUKUDA

Methods in Enzymology

xlv

VOLUME 418. Embryonic Stem Cells Edited by IRINA KLIMANSKAYA AND ROBERT LANZA VOLUME 419. Adult Stem Cells Edited by IRINA KLIMANSKAYA AND ROBERT LANZA VOLUME 420. Stem Cell Tools and Other Experimental Protocols Edited by IRINA KLIMANSKAYA AND ROBERT LANZA VOLUME 421. Advanced Bacterial Genetics: Use of Transposons and Phage for Genomic Engineering Edited by KELLY T. HUGHES VOLUME 422. Two-Component Signaling Systems, Part A Edited by MELVIN I. SIMON, BRIAN R. CRANE, AND ALEXANDRINE CRANE VOLUME 423. Two-Component Signaling Systems, Part B Edited by MELVIN I. SIMON, BRIAN R. CRANE, AND ALEXANDRINE CRANE VOLUME 424. RNA Editing Edited by JONATHA M. GOTT VOLUME 425. RNA Modification Edited by JONATHA M. GOTT VOLUME 426. Integrins Edited by DAVID CHERESH VOLUME 427. MicroRNA Methods Edited by JOHN J. ROSSI VOLUME 428. Osmosensing and Osmosignaling Edited by HELMUT SIES AND DIETER HAUSSINGER VOLUME 429. Translation Initiation: Extract Systems and Molecular Genetics Edited by JON LORSCH VOLUME 430. Translation Initiation: Reconstituted Systems and Biophysical Methods Edited by JON LORSCH VOLUME 431. Translation Initiation: Cell Biology, High-Throughput and Chemical-Based Approaches Edited by JON LORSCH VOLUME 432. Lipidomics and Bioactive Lipids: Mass-Spectrometry–Based Lipid Analysis Edited by H. ALEX BROWN VOLUME 433. Lipidomics and Bioactive Lipids: Specialized Analytical Methods and Lipids in Disease Edited by H. ALEX BROWN

xlvi

Methods in Enzymology

VOLUME 434. Lipidomics and Bioactive Lipids: Lipids and Cell Signaling Edited by H. ALEX BROWN VOLUME 435. Oxygen Biology and Hypoxia Edited by HELMUT SIES AND BERNHARD BRU¨NE VOLUME 436. Globins and Other Nitric Oxide-Reactive Protiens (Part A) Edited by ROBERT K. POOLE VOLUME 437. Globins and Other Nitric Oxide-Reactive Protiens (Part B) Edited by ROBERT K. POOLE VOLUME 438. Small GTPases in Disease (Part A) Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 439. Small GTPases in Disease (Part B) Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 440. Nitric Oxide, Part F Oxidative and Nitrosative Stress in Redox Regulation of Cell Signaling Edited by ENRIQUE CADENAS AND LESTER PACKER VOLUME 441. Nitric Oxide, Part G Oxidative and Nitrosative Stress in Redox Regulation of Cell Signaling Edited by ENRIQUE CADENAS AND LESTER PACKER VOLUME 442. Programmed Cell Death, General Principles for Studying Cell Death (Part A) Edited by ROYA KHOSRAVI-FAR, ZAHRA ZAKERI, RICHARD A. LOCKSHIN, AND MAURO PIACENTINI VOLUME 443. Angiogenesis: In Vitro Systems Edited by DAVID A. CHERESH VOLUME 444. Angiogenesis: In Vivo Systems (Part A) Edited by DAVID A. CHERESH VOLUME 445. Angiogenesis: In Vivo Systems (Part B) Edited by DAVID A. CHERESH VOLUME 446. Programmed Cell Death, The Biology and Therapeutic Implications of Cell Death (Part B) Edited by ROYA KHOSRAVI-FAR, ZAHRA ZAKERI, RICHARD A. LOCKSHIN, AND MAURO PIACENTINI VOLUME 447. RNA Turnover in Bacteria, Archaea and Organelles Edited by LYNNE E. MAQUAT AND CECILIA M. ARRAIANO VOLUME 448. RNA Turnover in Eukaryotes: Nucleases, Pathways and Analysis of mRNA Decay Edited by LYNNE E. MAQUAT AND MEGERDITCH KILEDJIAN

Methods in Enzymology

xlvii

VOLUME 449. RNA Turnover in Eukaryotes: Analysis of Specialized and Quality Control RNA Decay Pathways Edited by LYNNE E. MAQUAT AND MEGERDITCH KILEDJIAN VOLUME 450. Fluorescence Spectroscopy Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 451. Autophagy: Lower Eukaryotes and Non-Mammalian Systems (Part A) Edited by DANIEL J. KLIONSKY VOLUME 452. Autophagy in Mammalian Systems (Part B) Edited by DANIEL J. KLIONSKY VOLUME 453. Autophagy in Disease and Clinical Applications (Part C) Edited by DANIEL J. KLIONSKY VOLUME 454. Computer Methods (Part A) Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND VOLUME 455. Biothermodynamics (Part A) Edited by MICHAEL L. JOHNSON, JO M. HOLT, AND GARY K. ACKERS (RETIRED) VOLUME 456. Mitochondrial Function, Part A: Mitochondrial Electron Transport Complexes and Reactive Oxygen Species Edited by WILLIAM S. ALLISON AND IMMO E. SCHEFFLER VOLUME 457. Mitochondrial Function, Part B: Mitochondrial Protein Kinases, Protein Phosphatases and Mitochondrial Diseases Edited by WILLIAM S. ALLISON AND ANNE N. MURPHY VOLUME 458. Complex Enzymes in Microbial Natural Product Biosynthesis, Part A: Overview Articles and Peptides Edited by DAVID A. HOPWOOD VOLUME 459. Complex Enzymes in Microbial Natural Product Biosynthesis, Part B: Polyketides, Aminocoumarins and Carbohydrates Edited by DAVID A. HOPWOOD VOLUME 460. Chemokines, Part A Edited by TRACY M. HANDEL AND DAMON J. HAMEL VOLUME 461. Chemokines, Part B Edited by TRACY M. HANDEL AND DAMON J. HAMEL VOLUME 462. Non-Natural Amino Acids Edited by TOM W. MUIR AND JOHN N. ABELSON VOLUME 463. Guide to Protein Purification, 2nd Edition Edited by RICHARD R. BURGESS AND MURRAY P. DEUTSCHER VOLUME 464. Liposomes, Part F Edited by NEJAT DU¨ZGU¨NES,

xlviii

Methods in Enzymology

VOLUME 465. Liposomes, Part G Edited by NEJAT DU¨ZGU¨NES, VOLUME 466. Biothermodynamics, Part B Edited by MICHAEL L. JOHNSON, GARY K. ACKERS, AND JO M. HOLT VOLUME 467. Computer Methods Part B Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND VOLUME 468. Biophysical, Chemical, and Functional Probes of RNA Structure, Interactions and Folding: Part A Edited by DANIEL HERSCHLAG VOLUME 469. Biophysical, Chemical, and Functional Probes of RNA Structure, Interactions and Folding: Part B Edited by DANIEL HERSCHLAG VOLUME 470. Guide to Yeast Genetics: Functional Genomics, Proteomics, and Other Systems Analysis, 2nd Edition Edited by GERALD FINK, JONATHAN WEISSMAN, AND CHRISTINE GUTHRIE VOLUME 471. Two-Component Signaling Systems, Part C Edited by MELVIN I. SIMON, BRIAN R. CRANE, AND ALEXANDRINE CRANE VOLUME 472. Single Molecule Tools, Part A: Fluorescence Based Approaches Edited by NILS G. WALTER VOLUME 473. Thiol Redox Transitions in Cell Signaling, Part A Chemistry and Biochemistry of Low Molecular Weight and Protein Thiols Edited by ENRIQUE CADENAS AND LESTER PACKER VOLUME 474. Thiol Redox Transitions in Cell Signaling, Part B Cellular Localization and Signaling Edited by ENRIQUE CADENAS AND LESTER PACKER VOLUME 475. Single Molecule Tools, Part B: Super-Resolution, Particle Tracking, Multiparameter, and Force Based Methods Edited by NILS G. WALTER VOLUME 476. Guide to Techniques in Mouse Development, Part A: Mice, Embryos, and Cells, 2nd Edition Edited by PAUL M. WASSARMAN AND PHILIPPE M. SORIANO VOLUME 477. Guide to Techniques in Mouse Development, Part B: Mouse Molecular Genetics, 2nd Edition Edited by PAUL M. WASSARMAN AND PHILIPPE M. SORIANO

C H A P T E R

O N E

Lentivirus Transgenesis Alexander Pfeifer,* Tiongti Lim,† and Katrin Zimmermann* Contents 1. Introduction 1.1. Short overview: Viral vectors—Focus on lentiviruses 1.2. Transgenic technologies—Why retroviruses did not work? 2. Generation of Lentiviral Vectors 2.1. Lentiviral vector system 2.2. Production of high-titer lentiviral vector preparations 3. Generation of Transgenic Animals with Lentiviral Vectors 3.1. Subzonal injection 3.2. Lentiviral infection of denuded embryos 4. Characterization of Transgenic Animals 4.1. Genotyping of transgenic mice by PCR and Southern blotting 4.2. Analysis by quantitative real-time PCR 5. Summary Acknowledgments References

4 4 5 5 5 6 8 10 10 11 11 12 13 13 13

Abstract Lentiviral transgenesis is a promising alternative to direct microinjection of DNA into pronuclei, which is by and large restricted to certain mouse strains. Lentiviruses are complex retroviruses that integrate their genome into the host chromosome. Vectors derived from lentiviruses can efficiently transfer transgenes in oocytes and early embryos, which is the basis for the use of these vectors in transgenesis. Lentivirus transgenesis has been used in many different species, including mouse, rat, pig, bovine, monkeys, and even birds. Here we present a protocol for generating transgenic animals by lentiviral transduction of early embryos as well as for analyzing viral integrants in transgenic animals.

* Institute of Pharmacology and Toxicology, Biomedical Center (BMZ), University of Bonn, Germany Institute of Pharmacology and Toxicology, Life & Brain Center, University of Bonn, Germany

{

Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77001-4

#

2010 Elsevier Inc. All rights reserved.

3

4

Alexander Pfeifer et al.

1. Introduction 1.1. Short overview: Viral vectors—Focus on lentiviruses Lentiviruses belong to the large family of Retroviridae (Shaunak and Weber, 1992). They contain a complex viral RNA genome that is integrated into the host genome after reverse transcription (Desrosiers, 2001), thereby forming a so-called provirus (Goff, 2001). In contrast to prototypic oncoretroviruses like murine leukemia virus (MuLV), lentiviruses are able to transduce both dividing and nondividing cells (Naldini et al., 1996). The first lentivectors were derived from HIV-1 (Naldini et al., 1996; Poeschla et al., 1996). Since then, replication-defective lentivirus vector systems have been developed for many different lentiviruses such as simian and feline immunodeficiency viruses (Verma and Weitzman, 2005). The most widely used vector system is derived from HIV-1 and a 3rd generation of this lentivector system has been developed with optimized performance and biosafety features (for details see De Palma and Naldini, 2002). The lentiviral vector carries the promoter and transgene of interest flanked by the long terminal repeats (LTRs), whereas essential structural, packaging, and envelope plasmids are provided in trans by the packaging/producer cell (Dull et al., 1998; Follenzi and Naldini, 2002). Thus, the life cycle of produced lentiviral vectors is reduced to only one round of infection. Lentiviral vectors are an efficient tool for transferring transgenes to preimplantation embryos to generate transgenic mammals (Pfeifer, 2004). Viral transgenesis was initially developed by using simian virus 40 (SV40) DNA, which was injected into the cavity of mouse blastocysts to generate transgenic animals ( Jaenisch and Mintz, 1974). It was shown that the SV40 DNA persisted from the early embryonic stages to adulthood, indicating that viral infection might be a way to generate transgenic animals ( Jaenisch and Mintz, 1974). Meanwhile, lentiviral transgenesis has been successfully applied for mouse (Lois et al., 2002; Pfeifer et al., 2002), rat (Lois et al., 2002), pig (Hofmann et al., 2003; Whitelaw et al., 2004), cattle (Hofmann et al., 2004), chicken (McGrew et al., 2004), quail (Scott and Lois, 2005), and monkeys (Sasaki et al., 2009; Yang et al., 2008). Here, we present a protocol for infecting embryos with lentiviral vectors at very early stages of development in order to obtain animals that carry the transgene in every cell ( Jaenisch et al., 1975). Two alternative ways for generating transgenic mice are presented: subzonal injection of lentivector particles (Lois et al., 2002) or incubation of denuded embryos (generated by removal of the zona pellucida) with the lentiviral particles (Pfeifer et al., 2002).

Lentivirus Transgenesis

5

1.2. Transgenic technologies—Why retroviruses did not work? Pioneering studies with preimplantation embryos demonstrated that oncoretroviruses (e.g., MuLV) and retroviral vectors can be used to generate mosaic mice ( Jaenisch et al., 1975). Furthermore, germ line integration and transmission in a Mendelian manner of MuLV was demonstrated for mice ( Jaenisch, 1976). Although the viral genes were integrated into the murine genome, no expression was observed in the newborn mice. It turned out that the retroviral LTRs recruit host factors responsible for active repression of viral gene expression. A major mechanism for silencing of the retroviral genes is de novo methylation of viral promotor sequences and host DNA flanking the viral integration site ( Jahner and Jaenisch, 1985). Gene silencing was not only observed in mice but also in other species (Chan et al., 1998). Given the drawbacks of retroviral transgenesis, DNA microinjection in the pronucleus became the most widely employed method to create transgenic animals. Although large genomic DNA pieces can be transferred (Giraldo and Montoliu, 2001), the principal disadvantage of this method is that it is confined to certain mouse strains that can withstand piercing of the plasma membrane and have large, easy to visualize pronuclei. In other mouse strains, rats, and larger mammals, this technology results in extremely high production costs (Pfeifer, 2006). Lentiviral transgenesis presents itself as an ever more broadly used alternative method for creating transgenic animals.

2. Generation of Lentiviral Vectors 2.1. Lentiviral vector system A 3rd generation HIV-derived lentivector system is applied in our experiments (Dull et al., 1998) consisting of two major parts: the vector and the packaging constructs. The plasmids used were originally derived from the lab of Inder Verma (The Salk Institute for Biological Studies, Laboratory of Genetics, La Jolla, CA, USA). In addition, various HIV-derived lentivector plasmids can also be obtained from other labs like Didier Trono’s Lentiviral Vector Production Unit (http://tronolab.epfl.ch/page58115.html) or are commercially available (e.g., from Invitrogen or Sigma). The lentiviral vector contains the transgene and the promoter, as well as cis-acting sequences and regulatory elements (Fig. 1.1). One of these regulatory elements is the central polypurine tract (cPPT) that enhances the import of the viral preintegration complex into the nucleus of the infected cell (Zennou et al., 2001). The posttranscriptional regulatory element of woodchuck hepatitis virus (WPRE) enhances transgene expression (Zufferey et al., 1999). Furthermore, the promoter/enhancer sequences in the U3 region of the

6

Alexander Pfeifer et al.

5⬘LTR

5⬘SD

y

gag

3⬘SA cPPT RRE

CMV

eGFP

WPRE

3⬘LTR

Figure 1.1 Lentiviral vector construct of the 3rd generation. The lentiviral vector used carries the transgene (enhanced green fluorescent protein, eGFP) driven by an internal promoter (cytomegalovirus, CMV), the posttranscriptional regulatory element of woodchuck hepatitis virus (WPRE), the central polypurine tract (cPPT) as well as all necessary cis-acting sequences. The elements are flanked by long terminal repeats (LTRs), containing a self-inactivating mutation in the 30 LTR (SIN vector, black triangle).

30 LTR were deleted to generate self-inactivating (SIN) vectors (Dull et al., 1998; Miyoshi et al., 1998). During LTR conversion in the process of reverse transcription, the deletion of the promoter/enhancer elements of the 30 LTR are carried over to the 50 LTR leading to transcriptional inactivation of the provirus (Dull et al., 1998). Transgene expression is driven by incorporation of promoters in the vector construct (internal promoters). The packaging constructs encode the viral proteins Gag and Pol (e.g., on the plasmid pMDLg/pRRE), which are essential for production of lentiviral particles, and the Rev protein (e.g., encoded on plasmid RSV-rev) that is necessary to achieve sufficient expression of unspliced vector genomic RNA and of the gag and pol genes carrying a Rev-responsive element (RRE) (Dull et al., 1998). Furthermore, an envelope protein is provided in trans for the lentivirus particles (Pfeifer and Verma, 2001). We and others use the G protein of vesiculostomatitis virus (VSV-G) for pseudotyping (encoded, e.g., on pMD.G), because this results in a broad host range and stabilization of the viral particles produced. Concentration of the lentiviral vectors can thus be achieved by ultracentrifugation.

2.2. Production of high-titer lentiviral vector preparations In order to obtain high transgenic rates, high-titer lentivector preparations are essential. We routinely use HEK293T as packaging/producer cell line on 15 cm cell culture dishes that are transfected using calcium phosphate. If other cell culture dishes (e.g., 10 cm dishes or cell factories) are used, the components must be scaled proportionally. For preparing plasmids used for transfection, an endotoxin-free maxi-preparation kit (e.g., MachereyNagel, Qiagen) should be used. High-titer virus production is clearly dependent on transfection efficiency and testing of the 2 BBS (as described below) is of outmost importance. Preparation and testing of 2 BBS: 1. Dissolve the following components in a total of 400 ml water and autoclave the solution:

Lentivirus Transgenesis

7

4.26 g N,N-bis(hydroxyethyl)-2-aminoethanesulfonic acid 6.54 g NaCl 0.085 g Na2HPO4 pH 6.95 (with NaOH) 2. Culture HEK293T cells on a 15 cm cell culture dish and perform transfection according to the following protocol (calculate amounts of the components proportionally). The lentiviral vector used should contain a fluorescent transgene, for example, eGFP (enhanced green fluorescent protein). Sixteen to 18 h after transfection, fluorescence of the cells is determined by fluorescent microscopy and/or flowcytometry: The transfection efficiency should reach at least 80%. Standard lentiviral vector preparation: 1. Coat twelve 15 cm cell culture dishes with poly-L-lysine solution (50 ml of 0.01% solution, add 500 ml phosphate-buffered saline (PBS)) for at least 15 min at room temperature (the poly-L-lysine solution can be re-used, store at 4  C). 2. Plate HEK293T cells in 15 cm dishes covered with 20 ml DMEM (Dulbecco’s modified Eagle medium; supplemented with 10% fetal calf serum (FCS) and 100 U/ml Penicillin G/100
g/ml Streptomycin (PenStrep)). Incubate at 10% CO2 and 37  C. 3. After 24 h perform transfection of the cells: Apply dropwise 2.25 ml of the transfection mix, described below, per cell culture dish and mix carefully with the medium. Confluence of cells should not exceed 60%. Incubate the cells overnight at 3% CO2 and 37  C. Transfection mix: a. Mix 270
g lentivector, 175
g pMDLg/pRRE, 68
g RSV-rev, 95
g pMD.G, and 1.4 ml 2.5 M CaCl2 in a total volume of 14 ml. b. Add dropwise 14 ml of the transfection reagent 2 BBS (see above) and mix gently by inversion (5–10 times). Incubate the mixture for 15 min at room temperature. 4. Replace the medium 16–18 h after transfection by 16 ml fresh DMEM per plate and culture the cells at 10% CO2 and 37  C. 5. Perform the first virus collection 24 h after medium change: Process the supernatant through a 0.45
m surfactant-free cellulose acetate (SFCA) bottle-top filter (Nalgene) to remove cell debris. Culture the cells again at 10% CO2 and 37  C with 16 ml fresh DMEM per plate. 6. Preconcentrate the filtered supernatant for 2 h at 19,400 rpm (61,700  g) and 17  C in an ultracentrifuge (Optima L-100 XP, Beckman Coulter) with a SW32Ti rotor (Beckman Coulter) using 30 ml conical open-top thin-wall polyallomer centrifugation tubes. Resuspend the pellet in 50
l HBSS (Hanks’ balanced salt solution) each. Combine the suspensions and store the virus in a sterile screw-cap reaction tube at 4  C till the second virus-harvest has been finished.

8

Alexander Pfeifer et al.

7. Approximately 24 h after the first virus-harvest collect the supernatants and preconcentrate them again as described above. 8. Combine the suspensions of both virus-collections and perform the final ultraconcentration: Add the dissolved virus to 2 ml 20% (w/v) sucrose, provided in a 5 ml thin-wall polyallomer centrifugation tube. Centrifuge for 2 h at 21,000 rpm (53,500g) and 17  C in a SW55 rotor (Beckman Coulter). Resuspend the pellet in approximately 120
l HBSS. 9. Vortex the tube at 1400 rpm and 17  C for 45 min and centrifuge shortly (3 s, 16,000g) to pellet debris. Aliquot the opaque supernatant in sterile screw-caps and store the virus at 80  C. For analyzing the physical titer of the virus, produced commercially available enzyme-linked immunosorbent assays (ELISA) can be used that either determine the content of reverse transcriptase (e.g., reverse transcriptase assay, colorimetric; Roche) or the amount of lentiviral capsid protein p24 (e.g., RETRO-TEK HIV-1 p24 Antigen ELISA, Zeptometrix). If the lentiviral vector contains an eGFP-reporter, the biological titer can be determined additionally. In contrast to ELISA analyses, this method determines only the active and infectious particles. Measurement of biological titer: 1. Seed HEK293T cells (100,000 cells per 24-well). 2. Add 1
l of the concentrated virus to 330
l DMEM. Mix 30
l of the first dilution step with 300
l DMEM (dilution 1:10). Perform two further dilution steps in the same manner. After removal of the medium on the seeded HEK293T cells, add the different dilution mixtures and incubate cells overnight at 37  C and 10% CO2. Fill up medium on the next day. 3. 72 h after transduction, fix the cells with 4% paraformaldehyde (PFA) in PBS for 15 min on ice. Perform FACS analysis and use uninfected cells as negative control. Calculate the infective particles per ml (IP/ml) using the following equations: MOI ¼  lnðpercentage of eGFP  negative cells=100Þ Virus titer ðIP=mlÞ ¼ ðNo: of infected cellsÞ  ðMOIÞ ðdilution factorÞ  1000

3. Generation of Transgenic Animals with Lentiviral Vectors In principle, transgenesis is based on transfer of transgenes in embryos before implantation into the uterus. For an efficient infection of mammalian preimplantation embryos, the zona pellucida (ZP) must be overcome.

9

Lentivirus Transgenesis

This extracellular matrix surrounds the oocyte and the embryo and forms a physical barrier against viral infection (Pfeifer, 2006). Described here are two methods to infect the murine preimplantation embryo: first, subzonal virus injection into the perivitelline space (Lois et al., 2002) and second, incubation of denuded embryos with viral vectors after removal of the ZP (Pfeifer et al., 2002). Both methods require preimplantation embryos (zygotes) that are collected according to standard protocols (for details refer to Nagy et al., 2003). The zygotes are collected in KSOM–HEPES (Table 1.1) in 60 mm diameter dishes with center well (Falcon) or 4-wells for embryo culture (Falcon). Cultivation of embryos is performed at 37  C at 5% CO2 in the same dishes, but KSOM without HEPES is used (Table 1.1). The medium should be prepared freshly and used up within 2 weeks while storing at 4  C. Table 1.1 Embryo culture medium Component

KSOM medium

KSOM–HEPES

NaCl (mM) KCl (mM) KH2PO4 (mM) MgSO47H2O (mM) CaCl22H2O (mM) NaHCO3 (mM) Na lactate (mM) Na pyruvate (mM) D-Glucose (mM) Glutamine (Glutamax I) (mM) EDTA (tetrasodium salt) (mM) BSA (mg/ml) HEPES (mM) Phenol red (ml/ml) Penicillin G Na-salt (U/ml) Streptomycin sulfate (mg/ml) Amino acids (NEAA, Gibco #11140035) Amino acids (EAA, Gibco #11130036)

95 2.5 0.35 0.2 1.71 25 10 0.2 0.2 1

95 2.5 0.35 0.2 1.71 4 10 0.2 0.2 1

0.01

0.01

1 – 1 100 5 1

4 20 1 100 5 –

1

–

Pass through a sterile-filter (0.2
m). Incubate at 37  C and 5% CO2 overnight.

Pass through a sterile-filter (0.2
m). The pH should be between 7.2 and 7.4.

10

Alexander Pfeifer et al.

All substances used for manipulating and culture of mouse embryos must have highest purity grade and/or should be embryo-tested.

3.1. Subzonal injection Virus injection into the perivitelline space of the preimplantation embryo is a fast method but requires an inverted microscope with integrated modulation contrast (IMC) (e.g., DMIL, Leica; or inverted microscope with differential interference contrast (DIC), Zeiss) combined with a micromanipulation and injection system (e.g., Eppendorf ). One micromanipulator (TransferMan NK2, Eppendorf ) is used for the injection capillary, the other micromanipulator (Patchman NP2, Eppendorf ) for the holding pipette. 1. Transfer the embryos with a Transferpettor (Brand) with a 5
l glass capillary (Brand) to a microscope slide into a drop (ca. 200
l) of KSOM–HEPES. 2. Hold the embryos in place with a holding pipette (inner diameter ¼ 20– 25
m, angle 20 , fire polished ends; BioMedical Instruments), connected to a Cell tram Oil (Eppendorf ). 3. Fill the high-titer lentivirus (at least 108 IP/ml, higher titer is recommended) into an injection capillary using tip fillers (Eppendorf). Inject the viral particles under the ZP using a Transjector 5246 (Eppendorf ). 4. Transfer the embryos back into KSOM medium after the subzonal injection and culture them at 5% CO2 and 37  C until they reach the blastocyst stage. Transduction efficiency can easily be determined by fluorescence microscopy if a reporter gene (e.g., eGFP) was used. 5. Implant the blastocysts into female foster mice according to standard protocols for mouse manipulation (Nagy et al., 2003).

3.2. Lentiviral infection of denuded embryos Transgenesis by using denuded mouse embryos and incubating them with viral vectors does not need expensive equipment or well-trained workforce. The major disadvantage of this method is the impairment of embryonic viability due to the removal of the ZP. 1. Spot 50
l drops of acidic Tyrode’s solution (Sigma) on a cell culture dish and cover them with light mineral oil (Sigma). 2. Pipet the preimplantation embryos up and down in the Tyrode’s solution drops for a few seconds (up to 30 s). Observe the dissolving of the ZP under a stereo microscope. Avoid longer incubations in acidic Tyrode’s solution, because this is detrimental for the embryo. 3. Wash the embryos several times in drops of KSOM–HEPES.

Lentivirus Transgenesis

11

4. Dilute the lentivirus in KSOM (final concentration: at least 107 IP/ml). Spot 20–50
l virus containing medium on a center well cell culture dish and cover with mineral oil. Transfer one denuded embryo per drop and incubate at 37  C and 5% CO2 for 4–6 h. 5. Wash the embryos several times by transferring them to fresh drops of KSOM. Finally, incubate them in appropriate culture dishes with fresh KSOM at 37  C and 5% CO2 until they reach the blastocyst stage. 6. Implant the blastocysts into female foster mice according to standard protocols for mouse manipulation (Nagy et al., 2003).

4. Characterization of Transgenic Animals Analysis of transgenic animals can be performed by using different methods. For the procedures described below, genomic DNA can be isolated from the mouse tail. Applying PCR is the simplest way, but gives only information about the presence of the transgene. In contrast, Southern blot analysis displays the number of integrated lentiviral copies. Quantitative realtime PCR is another way to determine the amount of integrated provirus by using the TaqManÒ probe method. The method for analysis of HIV integrants developed by Butler et al. (2001) was further adapted by combining two sets of primers and probes (duplex PCR). This allows analysis of two different DNA sequences in parallel, the integrated lentiviral DNA as well as a single-copy housekeeping gene for determining total DNA loading. Described here are the principal ways for the analysis of lentivirus transgenic animals. All methods require genomic DNA extraction, performed by proteinase K treatment and classical phenol/chloroform extraction. Genomic DNA is then dissolved in an appropriate volume of H2O and stored at 4  C.

4.1. Genotyping of transgenic mice by PCR and Southern blotting For PCR analysis, genomic DNA is amplified by using specific primers for the transgene. 1. Mix 1
l of genomic DNA with 100 pM of each primer in a total volume of 50
l reaction buffer containing 200
M dNTPs and 2.5 U of Taq DNA polymerase. 2. Incubate the mixture in a thermocycler for 30 s at 95  C, 30 s at annealing temperature of the primers and the according time at 72  C (depending on size of transgene: approximately 1 min for 1000 bp) for a total of 40 cycles. Southern blot analysis is performed applying digested genomic DNA by using a restriction enzyme that cuts only once in the viral genome. Hence

12

Alexander Pfeifer et al.

DNA fragments of different sizes are generated depending on the integration site in the host genome. 1. Digest the genomic DNA with the appropriate restriction enzyme, for example, BamHI for the lentiviral vector shown here (Fig. 1.1; assuming that there is no additional BamHI site in the transgene). 2. Perform Southern blot analysis according to standard protocol (Southern, 1975): Separate the DNA via gel electrophoresis, transfer it to a hybridization membrane, and hybridize with a 32P-labeled cDNA probe, binding to the transgene or the vector. Visualization can be achieved by using a Bio-Imaging analyzer.

4.2. Analysis by quantitative real-time PCR Apart from Southern blotting, quantitative real-time PCR is a versatile tool for the determination of integrated provirus per genome. 1. Digest 5
g genomic DNA with two restriction enzymes (e.g., BamHI and EcoRI, if they are noncutters in the transgene). Precipitate the digested DNA with isopropanol and add an appropriate volume (ca. 100 ml) of H2O after drying. 2. Perform real-time PCR according to the manufacturer’s instructions with the IQ real-time PCR system and the IQ Multiplex Powermix (BioRad) in 25
l volumes. Use primers and probes for the transgene as well as for a single-copy housekeeping gene. For the transgene, the probe should be labeled with 6-FAM at the 50 end and with Eclipse Dark Quencher (EDQ) at the 30 end. For mice the single-copy (one copy per cell/genome) housekeeping gene Burkitt lymphoma receptor 1 (BLR1) can be used to determine total DNA amount applying the specific primers 50 -CGGAGCTCAACCGAGACCT-30 (forward) and 50 -TGCAAAAGGCAGGATGAAGA-30 (reverse) and the 50 -TexasRed-labeled probe 50 -TexasRed-CTGTTCCACCTCGCAGTAGCC GAC-EDQ-30 . Use primers and probes at a concentration of 5
M. 3. Dilute 1
l of genomic DNA with 5.6
l H2O. 4. Mix 12.5
l IQ Multiplex Powermix with 1.5
l of each forward and reverse primers and 1
l of each probe in a final volume of 23
l. For measurement in triplet, prepare a mastermix using 115% of the necessary amount of each component. 5. Use 75.9
l of the mastermix and add 6.6
l of the diluted DNA. Transfer the mixture to a 96-well plate with optical caps applying 25
l per well for triplicate measurement. 6. Amplify the transgene as well as the housekeeping gene according to following conditions: 3 min at 95  C, a two-step cycle of 95  C for 15 s and 60  C for 60 s for a total of 40 cycles.

13

Lentivirus Transgenesis

7. The amount of integrated provirus per genome is determined by the means of the cycle threshold (Ct) values: DCt ¼ CtðBLR1Þ  CtðtransgeneÞ Amount provirus=genome ¼ 2DCt

5. Summary Transgenic animals play an important role in basic and clinical research. Especially, transgenesis of large mammals is necessary for studying biological functions of genes of interest and for the development of new therapies and drugs to treat human diseases (Park, 2007). Lentivirus transgenesis is a versatile tool to achieve this goal. An important prerequisite for achieving high transgenesis rates are high-titer virus preparations. Therefore, not only the methods for successful lentiviral transgenesis and analysis of the transgenic animals but also the protocols to obtain high-titer lentivirus were described.

ACKNOWLEDGMENTS The authors would like to thank the house of experimental therapy at the University of Bonn for providing preimplantation embryos and implanting the blastocysts as well as Jutta Mu¨lich and Christina Stichnote for technical assistance. Our work is supported by the DFG (FOR 535).

REFERENCES Butler, S. L., Hansen, M. S., and Bushman, F. D. (2001). A quantitative assay for HIV DNA integration in vivo. Nat. Med. 7, 631–634. Chan, A. W., Homan, E. J., Ballou, L. U., Burns, J. C., and Bremel, R. D. (1998). Transgenic cattle produced by reverse-transcribed gene transfer in oocytes. Proc. Natl. Acad. Sci. USA 95, 14028–14033. De Palma, M., and Naldini, L. (2002). Transduction of a gene expression cassette using advanced generation lentiviral vectors. Methods Enzymol. 346, 514–529. Desrosiers, R. C. (2001). Nonhuman lentiviruses. In ‘‘Fields Virology,’’ (D. M. Knipe, P. M. Howley, D. Griffin, A. Martin, R. A. Lamb, B. Roizman, and S. E. Straus, eds.), pp. 2095–2121. Lippincott-Raven Publishers, Philadelphia, PA. Dull, T., Zufferey, R., Kelly, M., et al. (1998). A third-generation lentivirus vector with a conditional packaging system. J. Virol. 72, 8463–8471. Follenzi, A., and Naldini, L. (2002). Generation of HIV-1 derived lentiviral vectors. Methods Enzymol. 346, 454–465. Giraldo, P., and Montoliu, L. (2001). Size matters: Use of YACs, BACs and PACs in transgenic animals. Transgenic Res. 10, 83–103.

14

Alexander Pfeifer et al.

Goff, S. P. (2001). Retroviridae: The retroviruses and their replication. In ‘‘Fields Virology, Vol. 2,’’ (D. M. Knipe, P. M. Howley, and D. Griffin, et al., eds.), 4th edn. pp. 1871–1939. Lippincott-Raven Publishers, Philadelphia, PA. Hofmann, A., Kessler, B., Ewerling, S., et al. (2003). Efficient transgenesis in farm animals by lentiviral vectors. EMBO Rep. 4, 1054–1060. Hofmann, A., Zakhartchenko, V., Weppert, M., et al. (2004). Generation of transgenic cattle by lentiviral gene transfer into oocytes. Biol. Reprod. 71, 405–409. Jaenisch, R. (1976). Germ line integration and Mendelian transmission of the exogenous Moloney leukemia virus. Proc. Natl. Acad. Sci. USA 73, 1260–1264. Jaenisch, R., and Mintz, B. (1974). Simian virus 40 DNA sequences in DNA of healthy adult mice derived from preimplantation blastocysts injected with viral DNA. Proc. Natl. Acad. Sci. USA 71, 1250–1254. Jaenisch, R., Fan, H., and Croker, B. (1975). Infection of preimplantation mouse embryos and of newborn mice with leukemia virus: Tissue distribution of viral DNA and RNA and leukemogenesis in the adult animal. Proc. Natl. Acad. Sci. USA 72, 4008–4012. Jahner, D., and Jaenisch, R. (1985). Retrovirus-induced de novo methylation of flanking host sequences correlates with gene inactivity. Nature 315, 594–597. Lois, C., Hong, E. J., Pease, S., Brown, E. J., and Baltimore, D. (2002). Germline transmission and tissue-specific expression of transgenes delivered by lentiviral vectors. Science 295, 868–872. McGrew, M. J., Sherman, A., Ellard, F. M., et al. (2004). Efficient production of germline transgenic chickens using lentiviral vectors. EMBO Rep. 5, 728–733. Miyoshi, H., Blomer, U., Takahashi, M., Gage, F. H., and Verma, I. M. (1998). Development of a self-inactivating lentivirus vector. J. Virol. 72, 8150–8157. Nagy, A., Gertsenstein, M., Vintersten, K., and Behringer, R. (2003). Manipulating the Mouse Embryo: A Laboratory Manual. 3rd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Naldini, L., Blomer, U., Gallay, P., et al. (1996). In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science 272, 263–267. Park, F. (2007). Lentiviral vectors: Are they the future of animal transgenesis? Physiol. Genomics 31, 159–173. Pfeifer, A. (2004). Lentiviral transgenesis. Transgenic Res. 13, 513–522. Pfeifer, A. (2006). Lentiviral transgenesis—A versatile tool for basic research and gene therapy. Curr. Gene Ther. 6, 535–542. Pfeifer, A., and Verma, I. M. (2001). Virus vectors and their application. In ‘‘Fields Virology,’’ (P. M. Howley, D. M. Knipe, D. Griffin, A. Martin, R. A. Lamb, B. Roizman, and S. E. Straus, eds.), pp. 469–491. Lippincott-Raven Publishers, Philadelphia, PA. Pfeifer, A., Ikawa, M., Dayn, Y., and Verma, I. M. (2002). Transgenesis by lentiviral vectors: Lack of gene silencing in mammalian embryonic stem cells and preimplantation embryos. Proc. Natl. Acad. Sci. USA 99, 2140–2145. Poeschla, E., Corbeau, P., and Wong-Staal, F. (1996). Development of HIV vectors for anti-HIV gene therapy. Proc. Natl. Acad. Sci. USA 93, 11395–11399. Sasaki, E., Suemizu, H., Shimada, A., et al. (2009). Generation of transgenic non-human primates with germline transmission. Nature 459, 523–527. Scott, B. B., and Lois, C. (2005). Generation of tissue-specific transgenic birds with lentiviral vectors. Proc. Natl. Acad. Sci. USA 102, 16443–16447. Shaunak, S., and Weber, J. N. (1992). The retroviruses: Classification and molecular biology. Baillie`res Clin. Neurol. 1, 1–21. Southern, E. M. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98, 503–517. Verma, I. M., and Weitzman, M. D. (2005). Gene therapy: Twenty-first century medicine. Annu. Rev. Biochem. 74, 711–738.

Lentivirus Transgenesis

15

Whitelaw, C. B., Radcliffe, P. A., Ritchie, W. A., et al. (2004). Efficient generation of transgenic pigs using equine infectious anaemia virus (EIAV) derived vector. FEBS Lett. 571, 233–236. Yang, S. H., Cheng, P. H., Banta, H., et al. (2008). Towards a transgenic model of Huntington’s disease in a non-human primate. Nature 453, 921–924. Zennou, V., Serguera, C., Sarkis, C., et al. (2001). The HIV-1 DNA flap stimulates HIV vector-mediated cell transduction in the brain. Nat. Biotechnol. 19, 446–450. Zufferey, R., Donello, J. E., Trono, D., and Hope, T. J. (1999). Woodchuck hepatitis virus posttranscriptional regulatory element enhances expression of transgenes delivered by retroviral vectors. J. Virol. 73, 2886–2892.

C H A P T E R

T W O

Germline Modification Using Mouse Spermatogonial Stem Cells Mito Kanatsu-Shinohara and Takashi Shinohara Contents 1. Introduction 2. Establishing and Maintaining a GS Cell Culture 2.1. GS cell culture medium 2.2. Step 1. Dissociation of testis cells 2.3. Step 2. Initiation of GS cell culture 2.4. Step 3. Maintenance 2.5. Trouble shooting 2.6. Optional: Feeder-free culture and GS cell suspension culture 2.7. Optional: Establishing GS cells from adult testes 2.8. Optional: Derivation of mGS cells and its application in gene targeting 3. Gene Transduction and Genetic Selection of GS Cells 3.1. Step 1. Gene transduction to GS cells 3.2. Step 2. Drug selection 3.3. Step 3. DNA isolation and detection of homologous recombination 4. Spermatogonial Transplantation and Offspring Production 4.1. Donor cell preparation 4.2. Recipient preparation 4.3. Transplantation 4.4. Optional: Measurement of SSC activity by analyzing the recipient testes 4.5. Offspring production from recipient mice References

18 19 20 20 23 23 25 25 26 26 27 28 28 29 29 29 30 31 33 33 34

Abstract Spermatogonial stem cells (SSCs) in the testes are a new target for germline modification. With the development of an in vitro culture system and spermatogonial transplantation technique, SSCs can now be manipulated and used as an

Department of Molecular Genetics, Graduate School of Medicine, Kyoto University, Kyoto, Japan Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77002-6

#

2010 Elsevier Inc. All rights reserved.

17

18

Mito Kanatsu-Shinohara and Takashi Shinohara

alternative to embryonic stem cells for knockout mice production. The genetic and epigenetic stability of SSCs provide new possibilities for the application of germline mutagenesis in a wide range of animals.

1. Introduction Spermatogonial stem cells (SSCs) provide a foundation for spermatogenesis. SSCs, which constitute a fraction of spermatogonia in the testes, self-renew and differentiate to produce sperm throughout adult life. While female germline cells stop proliferating during the fetal period and isolation of oocyte/eggs is limited, male germ cells can be isolated and expanded in number. These advantages suggest that SSCs are a valuable target for germline modification. Recently, several SSC manipulation techniques have been developed. First, a breakthrough was made in the establishment of a germ cell transplantation technique. When dissociated donor testicular cells are injected into seminiferous tubules of infertile recipient testes lacking endogenous spermatogenesis, they colonize, differentiate into sperm, and produce normal offspring (Brinster and Avarbock, 1994; Brinster and Zimmermann, 1994). Spermatogenesis also occurs with cryopreserved SSCs and with SSCs in xenogeneic host. Rat SSCs undergo spermatogenesis in immunodeficient nude mouse testis and produce normal offspring, indicating a significant flexibility of spermatogenesis (Clouthier et al., 1996; Shinohara et al., 2006). Second, a long-term SSC culture technique was established. In 2000, glial cell line-derived neurotrophic factor (GDNF) was found to induce spermatogonial proliferation (Meng et al., 2000). Homozygous GDNF knockout mice die perinatally, whereas heterozygous knockout mice exhibit reduced spermatogenesis and eventually become infertile due to germ cell depletion. In contrast, GDNF-transgenic mice possess clumps of undifferentiated spermatogonia, suggesting that GDNF stimulates the selfrenewal division of SSCs. Considering this finding, our group succeeded in the long-term culture of SSCs in 2003, and we designated these cells as germline stem (GS) cells (Kanatsu-Shinohara et al., 2003a). In the presence of GDNF, GS cells produce uniquely shaped germ cell colonies. Although GS cells were originally established from neonatal testis, similar cells were subsequently established from adult testis, demonstrating that GS cells can be derived from SSCs at various stages (Kanatsu-Shinohara et al., 2004a; Kubota et al., 2004; Ogawa et al., 2004). GS cells can be used to produce transgenic and knockout animals through genetic transduction and drug selection (Kanatsu-Shinohara et al., 2005c, 2006a). Because the transgene is transmitted to half of the haploid cells, the efficiency of transgenesis is about

Mutagenesis of Male Germ Line Cells

19

50%, and it is 5–10 times higher than that achieved by conventional methods using eggs or oocytes (Nagano et al., 2001). Moreover, the frequency of homologous recombination is comparable to that achieved in embryonic stem (ES) cells (Kanatsu-Shinohara et al., 2006a). Most importantly, GS cells possess a very stable germline potential, retain a normal karyotype and DNA methylation patterns, and produce normal fertile offspring even after 2 years of culture (Kanatsu-Shinohara et al., 2005b). This is in contrast to ES cells, which often change DNA methylation patterns and lose their germ cell potential due to trisomy (Liu et al., 1997; Longo et al., 1997). Thus, SSCs may serve as a new target for animal transgenesis, which may provide an alternative to ES cells. We also found that the developmental potential of SSCs is not limited to spermatogenesis. Although primordial germ cells (PGCs), the fetal precursors of SSCs, can give rise to ES-like pluripotent cells (Matsui et al., 1992; Resnick et al., 1992), germline cells were believed to be fully committed to the germline by the middle of gestation and that such ES-like potential was missing from postnatal germ cells (Labosky et al., 1994). Unexpectedly, however, ES-like cells rarely appear in GS cell cultures of neonatal testes during culture initiation (Kanatsu-Shinohara et al., 2004a). These cells, referred to as multipotent GS (mGS) cells, not only differentiate into somatic cells, but also differentiate into germ cells. Several groups have reported the derivation of similar pluripotent/multipotent cells from mouse and human postnatal testes, including adults (Golestaneh et al., 2009; Guan et al., 2006; Kossack et al., 2009). Although the origin of these pluripotent/ multipotent cells is unclear, we recently discovered that GS cells may be converted directly into mGS cells in vitro (Kanatsu-Shinohara et al., 2008a). mGS cells behave like ES cells and are capable of producing knockout animals in a manner similar to ES cells (Takehashi et al., 2007b). Although the efficiency of establishing mGS cells (1 of 30 testes) is low and needs improvement, these results suggest an alternative use for SSCs in germline modification. In this chapter, we describe the methods associated with the SSC culture technique. Although several groups have reported alternative SSC culture methods (Guan et al., 2006; Kubota et al., 2004; Seandel et al., 2007), our method allows for the genetic selection and production of knockout mice from SSCs.

2. Establishing and Maintaining a GS Cell Culture GS cell culture may be initiated from both neonatal/pup and adult testes (Kanatsu-Shinohara et al., 2003a, 2004a; Kubota et al., 2004; Ogawa et al., 2004), but establishment is quicker and more efficient using neonatal/

20

Mito Kanatsu-Shinohara and Takashi Shinohara

pup testes. Immature testes are useful because the germ cell/somatic cell ratio is relatively high, and germ cells are easily separated from somatic cells based on their differential ability to attach to a gelatin-coated dish; therefore, antibody-mediated purification is not necessary. Furthermore, immature germ cells proliferate more actively than adult germ cells. In contrast, one must purify SSCs from adult testes because only 0.02–0.03% of all germ cells in the testis are stem cells (Meistrich and van Beek, 1993; Tegelenbosch and de Rooij, 1993). In both cases, removing as many somatic cells as possible is advisable. The efficiency of establishing GS cells is affected by the mice strain. While DBA/2, ICR, and C57BL/6
DBA/2F1(B6D2F1) are efficient, C57BL/6 is less efficient. GS cells can also be established from ddy, C3H, A, and AKR with variable efficiencies. Here, we describe a protocol for establishing a GS cell culture from P0-3 testes of DBA/2 or the ICR strain.

2.1. GS cell culture medium GS cell culture medium is prepared by modifying commercial medium (StemProÒ-34 serum-free medium (SFM); Invitrogen, Carlsbad, CA). Although other conditions with defined medium can also support SSC proliferation, a modified StemProÒ-34 medium currently provides the most efficient proliferation of GS cells in our laboratory. SSC proliferation is maintained by a combination of several cytokines. GS cell culture medium contains GDNF, and fibroblast growth factor-2 (FGF2). Although epidermal growth factor (EGF) was included in the original protocol, it is dispensable, and GS cell proliferation can be maintained with GDNF þ FGF2 only. Leukemia inhibitory factor (LIF) enhances colony formation from gonocytes, while it is dispensable for spermatogonia culture (Kanatsu-Shinohara et al., 2007). It is also dispensable for the maintenance of GS cells. Complete GS cell culture medium is prepared by adding several factors to basal medium (Table 2.1). Basal medium is made by adding 16 components to StemProÒ-34 SFM, followed by filtration through a 0.22-mm bottle-top filter. It can be stored in a refrigerator for at least 3 weeks. The six additives listed in Table 2.2 should be added before use to make the complete medium. Complete medium with growth factors can be stored in a refrigerator for up to 3 days.

2.2. Step 1. Dissociation of testis cells Dissolve collagenase (#C5138; Sigma, St. Louis, MO) at 1 mg/ml and deoxyribonuclease (DNase, #DN25; Sigma) at 7 mg/ml in Hanks’ balanced salt solution (HBSS) and filter. Isolate the testes from the mice, and remove

Table 2.1 Composition of basal medium for GS cell culture

Component

Insulin

Catalogue # a

Final concentration

Volume

Aliquots

25 mg/ml

500 ml

100 mg/ml 60 mM 30 nM 6 mg/ml 200 mg/ml 1 ml/ml 5 mg/ml

1 ml 500 ml 500 ml 3g 100 mg 500 ml 2.5 g

Dissolve 100 mg/3.8 ml DDW þ 0.2 ml 1 N HCl; store at  20  C Dissolve 100 mg/2 ml DDW; store at  20  C Dissolve 96.7 mg/10 ml DDW; store at  20  C Dissolve 5.2 mg/1000 ml DDW; store at  20  C Dissolve in 10 ml DDW and add all

2 mM 5
10 5 M

5 ml 5 ml 5 ml

100
; store at  20  C 5
10 3 M Store at 4  C

5 ml

Store at 4  C Dissolve 17.6 mg/ml DMSO; use immediately Dissolve 10 mg/ml DMSO; use immediately Dissolve 1 mg/ml ethanol; add 49 ml sterile medium. Store at  20  C Dissolve 1 mg/ml ethanol;, add 49 ml sterile medium. Store at  20  C

2-Mercaptoethanol MEM vitamin solution Nonessential amino acids Ascorbic acid d-Biotin b-Estradiol

Nacalai Tesque #19251-24 Sigma #T1147 Sigma #P7505 Sigma #S1382 Sigma #G7021 Sigma #P2256 Sigma #L4263 MP Biomedicalsb #810661 Sigma #G7513 Sigma #M3148 Invitrogen #11120-052 Invitrogen #11140-050 Sigma #A4544 Sigma #B4501 Sigma #E2758

10 4 M 10 mg/ml 30 ng/ml

500 ml 500 ml 750 l

Progesterone

Sigma #P8783

60 ng/ml

1.5 ml

Transferrin Putrescine Sodium selenite D-(þ)-glucose Pyruvic acid DL-Lactic acid Bovine albmin L-Glutamine

StemProÒ-34 SFM (Invitrogen #10639) is modified by addition of the following components. The amounts to be added to 500 ml StemProÒ-34 SFM are shown. a Nacalai Tesque, Inc., Kyoto, Japan. b MP Biomedicals, Inc., Irvine, CA.

Table 2.2 Components added to basal medium immediately before use Component Ò

StemPro -34 supplement Mouse EGF Human FGF2 Rat GDNF FBS ESGRO (murine LIF)

Catalog number

Invitrogen #10639 BD Biosciences 354010 Peprotech Inc.a #100-18B Peprotech Inc.a #450-51 Hycloneb #SH30396.03 Milliporec #ESG1107

The amounts to be added to 1 ml basal medium are shown. a Peprotech Inc. Rocky Hill, NJ. b Hyclone Laboratories, Inc. South Logan, UT. c Millipore, Billerica, MA.

Final

Volume

Aliquot

20 ml

50
Supplement is supplied with StemProÒ-34SFM.

20 ng/ml

2 ml

10 ng/ml

5 ml

Optional; Dissolve 100 mg/10 ml PBSþBSA; store at  20  C Dissolve 10 mg/5 ml PBSþBSA; store at  20  C

15 ng/ml

15 ml

Dissolve 10 mg/10 ml PBSþBSA; store at  20  C

1%

10 ml

103 units/ml

10 ml

Optional; enhances GS cell establishment when added to the initiation of neonatal testis culture. Dissolve 106 units/10 ml PBSþBSA; store at  20  C

Mutagenesis of Male Germ Line Cells

23

the tunica with fine forceps in cold HBSS. Wash two to three times with HBSS and transfer the tissue to 1–2 ml of collagenase and incubate at 37  C for 15 min. Agitate the tube several times during the incubation. Wash twice with HBSS and add 0.8 ml of 0.25% trypsin þ 0.2 ml of DNase, shake the tube several times to dissociate the seminiferous tubules, and incubate at 37  C for 10 min. Add 5 ml of Iscove’s modified Dulbecco’s medium (IMDM) þ 2% fetal bovine serum (FBS) and repeat pipetting until the cells are dissociated. Centrifuge and remove the supernatant.

2.3. Step 2. Initiation of GS cell culture Dissolve 1 g gelatin in 500 ml phosphate-buffered saline (PBS) and autoclave to make a 0.2% gelatin solution. Coat a 12-well culture plate with the 0.2% gelatin/PBS and incubate at room temperature for more than 20 min. Remove the gelatin solution, suspend the cells in complete culture medium, and transfer them to a gelatin-coated culture plate. The density should be 2
105 cells/0.8 ml medium per well of a 12-well culture plate. Incubate in 5% CO2 at 37  C overnight. Many cells attach to the plate after the overnight incubation, but a significant number of germ cells, as distinguished by their large size and characteristic pseudopod, remain floating. The floating cells should be transferred to a second culture plate after vigorous pipetting (use P1000pipette tips, 10–15 times; Gilson, Middleton, WI). The second culture plate does not need to be treated with gelatin. Very few germ cells are left on the original gelatin-coated plate, and cells transferred to secondary plates are relatively germ cell-enriched (Fig. 2.1A). Three to four days later, remove half of the culture medium and add the same volume. Within 1 week, the transferred cells will proliferate and spread on the bottom of the well; round proliferating cells will form germ cell colonies on top of the flat cell layer. Most of these primary colonies consist of compact clusters of cells with unclear borders. The timing of the first passage depends on colony growth, but between 10 and 14 days after culture initiation (DIV, days in vitro) is recommended. Wash twice with PBS, add 0.25% trypsin, and incubate at 37  C for 4 min. Add IMDM þ 2% FBS to stop the reaction. Replate at 1
dilution. The colonies will grow to their original size in about 10 days, when the cells are passed again (1/2
dilution).

2.4. Step 3. Maintenance After the third or fourth passage, the cells should be transferred onto mitomycin-C treated mouse embryonic fibroblasts (MEFs) (Fig. 2.1B). MEF feeders should be prepared according to the conventional ES cell

24

Mito Kanatsu-Shinohara and Takashi Shinohara

A

B

C

D

Figure 2.1 Culture appearances. (A) Gonocytes at 2 days in vitro. Gonocytes have large cell bodies and attach loosely to the plate, whereas fibroblasts and other somatic cells attach strongly to the plate. (B) Established GS cells. GS cell colonies produce a grape-like cluster. (C) Epiblast-like cell sheet, which appears in GS cell cultures. An epiblast-like cell sheet (arrow) is often observed when mGS cells appear in the culture. While distinguishing clearly between GS cell and mGS cell colonies is difficult, this is a clear sign indicating that mGS cells are starting to appear in the culture. However, this structure disappears after mGS cells are established, and whether it is an mGS cell precursor or a differentiated mGS cell state that exists only transiently is unclear. (D) mGS cells have similar appearances to ES cells. Bar ¼ 100 mm.

culture method, except they should be plated at a lower density (7.5
104 per well in 6-well culture plates). Briefly, treat MEF with 10 mg/ml mitomycin-C (#M053; Sigma; dissolve in PBS at 2 mg/4 ml and filter) for 2 h at 37  C and dissociate with trypsin. Suspend the MEF cells in Dulbecco’s modified Eagle’s medium (DMEM) þ 10% FBS and plate onto a gelatin-coated culture dish. MEF feeders should be used within 10 days (it is not necessary to change medium). Immediately before transferring the GS cells remove the medium from the MEF feeders and wash once (or twice) with PBS. GS cell growth becomes stable after about 30 DIV. The established GS cells should be plated at a density of 3
105 cells/well in 6-well culture plates. Cultures should be passed every 4–6 days depending on proliferation. The medium should be changed every 3 days (half medium change). Established GS cells continue to proliferate for more than 2 years without losing stem cell activity.

Mutagenesis of Male Germ Line Cells

25

2.5. Trouble shooting One of the problems in culture initiation is overgrowth of somatic cells. Somatic cells grow faster than SSCs in vitro and overwhelm the culture. This can be alleviated, in part, by reducing the serum concentration and/or enhancing the GDNF concentration. However, SSCs, at present, cannot grow in the complete absence of serum. Although two different ‘‘serumfree’’ culture systems have been reported (Kanatsu-Shinohara et al., 2005a; Kubota et al., 2004), one of these cultures were maintained using serum to stop the trypsin reaction at each passage in one study (Kubota and Brinster, 2008). Because the same medium cannot support feeder-free GS cell culture without serum, we also cannot exclude the possibility that residual serum remaining with the feeder cells promote GS cell propagation (KanatsuShinohara et al., 2005a). Besides somatic cells, loss or reduced GDNF activity often causes problems. Because GDNF is essential for SSC selfrenewal, low concentrations of GDNF are detrimental to GS cell culture. When the culture reaches confluency, c-Kit expression, a marker of differentiating spermatogonia, is occasionally upregulated, suggesting that some levels of differentiation occur in culture. At the same time, GS cell growth is suppressed. In this situation, remove the medium and replace with fresh medium. In most cases, a decrease in the number of differentiating cells is observed, and the stem cells revive by GDNF action. If the GS cells look unhealthy, change the medium or add GDNF. GS cells are very tough, and in most cases will recover.

2.6. Optional: Feeder-free culture and GS cell suspension culture GS cells can be maintained under a feeder-free condition or in suspension culture (Kanatsu-Shinohara et al., 2005a, 2006b). For feeder-free culture, the culture plates should be coated with 20 mg/ml laminin (#354232; BD Biosciences, Franklin Lakes, NJ) for 1–2 h at room temperature (0.8 ml laminin solution per well of a 6-well culture plate). Remove the laminin solution and immediately plate the GS cells at 3
105 per well. The culture can be passed with trypsin. For the suspension culture, plate 2
105 GS cells directly onto a Petri dish. GS cells aggregate to make clumps, but continue to proliferate slowly. These cultures can be passed without trypsin, and GS cell clumps can be disaggregated using simple pipetting with P1000 tips. Although the proliferation rates are lower than MEF culture (doubling time, 2.7 days for MEF culture, 5.6 days for feeder-free culture, and 4.7 days for suspension culture), GS cells maintained feeder-free or in suspension continue to proliferate for more than 6 months without losing SSC activity.

26

Mito Kanatsu-Shinohara and Takashi Shinohara

2.7. Optional: Establishing GS cells from adult testes GS cells can also be established from adult testes, but enrichment using gelatin-coated plates is not effective. Because the SSC frequency is low in adult testes (about 2–3 in 104 cells; Tegelenbosch and de Rooij, 1993; Meistrich and van Beek, 1993), purification using either magnetic beads or a fluorescence-activated cell sorter is necessary. Several SSC markers are useful for SSC enrichment. For example, a relatively high rate of enrichment has been achieved in our lab using magnetic beads sorted with a6- or b1-integrins and CD9 (Kanatsu-Shinohara et al., 2004b; Shinohara et al., 1999). Thy-1 is also expressed in SSCs, but seems to work better after Percoll separation (Ryu et al., 2004). A higher enrichment efficiency (about 166-fold) can be achieved with cell sorting using a combination of several markers (Ryu et al., 2004; Shinohara et al., 2000), but enrichment by magnetic beads is sufficient for culture initiation in many cases. Purified cells are transferred to gelatin- or laminin-coated plates and cultured in GS cell culture medium. In a manner similar to the initiation of GS cell culture from neonatal testes, germ cell colonies are formed on top of the flat somatic cell layer. However, because adult germ cells proliferate more slowly than neonatal or pup germ cells in the initiation of culture, more time is required to establish a GS cell culture from adult testes than from neonatal testes. Retrieving GS cell colonies is sometimes helpful using glass needles in cases of somatic cell overproliferation.

2.8. Optional: Derivation of mGS cells and its application in gene targeting mGS cells were initially discovered in our attempt to produce knockout GS cells. Unusual colonies may rarely occur during the initiation of a GS cell culture; typical GS cell colonies show grape-like clumps, whereas pluripotent cell colonies appear like ES cells, or as an epiblast cell sheet (Fig. 2.1C), a derivative of the inner cell mass. Once these colonies develop in culture, they outgrow GS cells after several passages because they proliferate faster. Although they can grow in GS cell culture medium, changing the medium (DMEM þ 15% FBS) and supplementing it with LIF is advisable to maintain their undifferentiated state. GDNF is no longer necessary after mGS cells are established, and established mGS cells can be maintained in the same manner as ES cells (Fig. 2.1D). Although both GS and mGS cells are derived from testicular germ cells, they have different characteristics. GS cells are unipotent and committed to spermatogenesis, whereas mGS cells are pluripotent and can differentiate into various cell types. They behave exactly like ES cells, except they have a partial androgenetic imprinting pattern and characteristic centromeric DNA hypomethylation, reflecting their postnatal male germ cell origin (KanatsuShinohara et al., 2004a; Yamagata et al., 2007). Because SSC activity is lost

27

Mutagenesis of Male Germ Line Cells

Mouse testis Unipotent Stable karyotype Stable imprinting pattern GS cell Stable germline potential Nontumorigenic Slow proliferation

mGS cell Introduction of transgene

Pluripotent Unstable karyotype Unstable imprinting pattern Unstable germline potential Tumorigenic Fast proliferation

Transplantation into testis Injection into blastocyst Mating

Transfer to host mother

Tg 100% Germline transmission mu Chimera

Figure 2.2 GS and mGS cell differences in their utility for animal transgenesis.

in mGS cells, they produce teratomas instead of sperm when transplanted into testes (Fig. 2.2). They produce germline chimeras following injection into blastocysts, in a manner similar to ES cells. The application of conventional methods used for ES cells allows for the production of knockout mice or double-knockout cells from mGS cells (Takehashi et al., 2007b).

3. Gene Transduction and Genetic Selection of GS Cells SSCs can be transduced with a retrovirus (Nagano et al., 2000), lentivirus (Nagano et al., 2002), adenovirus (Takehashi et al., 2007a), or adenoassociated virus vector (Honoramooz et al., 2008), and with plasmid vectors by various methods, including the calcium phosphate method, lipofection, and electroporation (Kanatsu-Shinohara et al., 2005c). Knockout mice were produced with either a virus-based gene trap vector or with a gene-targeting plasmid vector. However, because viral origin promoters are repressed in germ cells and early embryonic cells, choosing promoters that are not suppressed in germline cells is necessary. According to our experience, most expression vectors used in ES cells will express foreign genes in SSCs.

28

Mito Kanatsu-Shinohara and Takashi Shinohara

Genetic selection with drugs is relatively difficult with GS cells. While ES cells grow robustly like transformed cells and can proliferate from single cells, GS cell proliferation is modest and is influenced by cell density. Therefore, one should avoid an extremely low cell density during culture. When only a few cells are recovered after drug selection, we add nontransfected ‘‘carrier’’ cells to enhance the growth of transfected cells (KanatsuShinohara et al., 2005c). These carrier cells are removed by repeated drug selection, which usually takes about 2 months. Here, we describe the protocol for gene transduction of a gene targeting vector and the genetic selection of GS cells using a neomycin-resistant gene.

3.1. Step 1. Gene transduction to GS cells GS cells can be transduced using the Cell Line T Nucleofector Kit (#VCA1002; Lonza, Mu¨nster, Germany). Briefly, dissociate the GS cells with trypsin and suspend them in IMDM þ 2% FBS. Remove as much of the MEF as possible because MEF abrogates GS cell survival. Cells (4–8
106) should be transferred to a tube and centrifuged. Wash the cells twice with PBS, and suspend the pellet in 100 ml of Nucleofector solution (#VCA1002; Lonza). Then add 5 mg of the gene targeting vector. The gene targeting vector is constructed following the conventional method used for ES cells, and we are able to isolate clones with a homologous recombination using both circular and linearized vectors. Plasmid DNA is suspended in sterile water or Tris–EDTA (TE). Transfer the cell suspension to a cuvette and treat with an electroporator (Nucleofector II device; Lonza) using program A-23. GS cells can also be efficiently transduced using an electroporator (Bio-Rad, Hercules, CA); the cells are suspended in PBS and treated with a single 500 mF pulse. After electroporation, culture the cells with conventional GS cell medium on MEF. The cells are passed in fresh MEF (1:1) the next day.

3.2. Step 2. Drug selection Start the drug selection after the cells have recovered from the damage of electroporation (about 7–10 days). At 2–3 days after passage, add 40–120 mg/ml G418 (Geneticin, #83-5027; Invitrogen). If the culture reaches confluency, regardless of the G418, the cells should be passed (1:1). Puromycin (#P7255; Sigma; dissolve 1 mg/ml water) can also be used as a selection marker at 0.1–0.2 mg/ml. In contrast, if the drug efficiently reduces cell density, leave the culture plate for about 3 weeks and only change the medium, until transduced cell colonies have formed. Treat the culture with 0.25% trypsin, and pass to fresh MEF at a 1:1 dilution; add nontransfected GS cells at a density of about 3
104 cells/ cm2. Restart the drug selection 2–3 days after the passage. Repeat the same procedure until a sufficient number of transfected GS cells is obtained at 100% purity. Overall, about 2–3 months is needed to obtain about 106 cells of an established clone.

Mutagenesis of Male Germ Line Cells

29

3.3. Step 3. DNA isolation and detection of homologous recombination Genomic DNA is isolated from cultured cells. More than about 20 mg DNA can be obtained from 2
106 GS cells. The clones inserted with a homologous recombination are screened with the polymerase chain reaction (PCR) or Southern blotting.

4. Spermatogonial Transplantation and Offspring Production SSCs will migrate into the germline niche after microinjecting them into the seminiferous tubules and create germ cell colonies. This method is conceptually similar to hematopoietic stem cell transplantation (Till and McCulloch, 1961), in that donor stem cells are reintroduced into a microenvironment that lacks endogenous stem cells. This technique provided the first functional SSC assay. Furthermore, because the recipient animals eventually produce offspring from donor SSCs, this technique also allowed the possibility to manipulate SSCs to produce offspring.

4.1. Donor cell preparation Having a donor marker during germ cell transplantation is advisable. Because single germ cell colonies originate from single SSCs (KanatsuShinohara et al., 2006c; Nagano et al., 1999), a marker allows for the quantitative assessment of donor SSC colonization. Moreover, it also helps identify abnormal germ cell development. Several transgenes, including LacZ or enhanced green fluorescent protein (EGFP), are used. Each has its advantage. For example, EGFP allows for offspring production using microinsemination or serial transplantation, whereas LacZ allows for better histological presentation of donor cell colonization. Single cell suspensions are prepared using an enzymatic digestion as described above. The reaction can be stopped by adding DMEM þ 10– 20% FBS. Cells must be filtered though a 40-mm nylon mesh (#BD352340; BD Biosciences) before transplanting, so that cell aggregates do not clog the injection needle. Cells are usually suspended in DMEM þ 10% FBS. The concentration of cells can be as high as 108 cells/ml. Although fresh testicular cells can be transplanted relatively easily, freeze–thawed cells or cultured cells often clog the injection needle, which may occasionally be resolved by adding a small amount of DNase (about 0.7 mg/ml; #DN25; Sigma) to the cell suspension. Although the cells can be stored on ice for several hours, injecting the cells as soon as the cell suspension is made is better.

30

Mito Kanatsu-Shinohara and Takashi Shinohara

4.2. Recipient preparation Donor SSCs must be histocompatible with the recipient animal. Although testes are immune-privileged organs, allogeneic SSCs can be rejected despite successful transplantation (Kanatsu-Shinohara et al., 2003b). Although several cases of successful transplantation have been reported, using immunodeficient nude/scid mice is advisable when the donor cells are not histocompatible with the recipients. This allows for a xenogeneic transplantation and produces fertile spermatozoa (Clouthier et al., 1996; Shinohara et al., 2006). Furthermore, elimination of endogenous SSCs is prerequisite. Transplanted SSCs are believed to compete for available niches with endogenous SSCs, and reducing the number of endogenous SSCs is thought to increase the transplantation efficiency (Shinohara et al., 2002). This can be achieved by busulfan or radiation treatment. Additionally, one can use congenitally infertile mutants such as WBB6F1-W/Wv (designated as W) mice (Brinster and Zimmermann, 1994). While local radiation of testis specifically eliminates germ cells (Creemers et al., 2002), busulfan treatment sometimes induces anemia caused by its systemic side effect on bone marrow. This can be a problem with nude mice that are relatively weak to busulfaninduced damage. However, most germ cell transplantation experiments depend on busulfan-treated recipients as described in the following protocol. 4.2.1. Step 1. Busulfan preparation Busulfan (#B2635; Sigma) is first dissolved in dimethyl sulfoxide (DMSO) at 8 mg/ml. Because of its strong toxicity, one must wear gloves and a mask for protection. Once it is dissolved in DMSO, an equal amount of distilled water is added to the busulfan solution at 4 mg/ml. Upon mixing, the solution will generate heat. It is advisable to use the busulfan solution before it starts to form precipitates. 4.2.2. Step 2. Injection of the busulfan solution into recipient animals Busulfan solution is injected intraperitoneally into animals. The amount of busulfan varies according to the animal background. However, we generally use doses ranging from 44 to 50 mg/kg. Inbred mice are more sensitive to busulfan than F1 hybrids or animals in closed colonies. Although spermatogenesis can be suppressed at a lower dose, it may regenerate from endogenous SSCs after several weeks, which interferes with SSC colonization and the subsequent development of germ cell colonies. In addition, one must consider that the complete elimination of endogenous spermatogenesis is undesirable to successfully produce offspring in both mice and rats (Brinster and Avarbock, 1994; Ryu et al., 2007).

Mutagenesis of Male Germ Line Cells

31

4.3. Transplantation A single donor cell suspension can be introduced into the seminiferous tubules via three routes: the seminiferous tubules, rete testis, or efferent duct (Ogawa et al., 1997). In the original report (Brinster and Zimmermann, 1994), the cells were injected directly into the seminiferous tubules by exposing the tubules to air via tunica albuginea removal. However, this method is time-consuming and is the most difficult. The most popular method is efferent duct injection. The efferent duct provides passage for mature spermatozoa from the testis to the epididymis. After microinjecting the donor cell suspension, the seminiferous tubules are filled in a retrograde manner. Therefore, transplanted cells migrate in the opposite direction to normal sperm transport. The injection can be performed manually with a 1-ml syringe equipped with a glass needle. This allows for easier guidance of the injection pipette and is recommended for the novice. Alternatively, automatic microinjection equipment used commonly for producing transgenic offspring (model 5242; Eppendorf, Hamburg, Germany) can be used. This has an advantage in that it provides a more accurate quantitative injection and better pressure control. Here, we describe the method used to inject into the efferent duct. 4.3.1. Step1. Glass needle preparation We regularly use glass needles (1 mm in diameter using a Sutter puller; model P-97 and #BF200-156-10; Sutter Instruments, Novato, CA) for transplantation. However, the needle edge does not need to be sharpened, as is often necessary for pronuclear DNA injection. The glass needle can be of any size or shape as long as cells can go pass through the needle tip into the seminiferous tubules. 4.3.2. Step 2. Efferent duct exposure After anesthetizing the animal, make a straight midline incision (1.5–2 cm) in the center of the abdomen. Using blunt forceps, expose the testis by pulling out the fat attached to the epididymis. Dissect out the efferent duct using fine forceps. Finding the efferent duct is sometimes difficult in aged, fat animals. Excessive dissection will damage the duct and is not advisable. 4.3.3. Step 3. Insertion of the donor cell suspension into the glass needle Insert the donor cell suspension from the rear end of the glass needle. Optionally, add trypan blue to the donor cell suspension to help visualize the fate of the transplanted cells. However, trypan blue may cause inflammation in some cases (Kanatsu-Shinohara et al., 2008b), and including the dye in the suspension solution is not advisable for regular experiments.

32

Mito Kanatsu-Shinohara and Takashi Shinohara

The transplantation volume varies depending on the recipient type. In busulfan-treated testis, as much as 10 ml of donor cell suspension can be injected to fill 80–90% of the tubules. For a W testis, we regularly inject 4 ml because it is smaller than a busulfan-treated testis. 4.3.4. Step 4. Injection into the efferent duct Insert the glass needle into the efferent duct (Fig. 2.3). Placing the epididymis on the medial side of the testis generally helps expose the duct. If the needle is injected too deeply along the efferent duct through the rete testis, cells will leak out into the interstitial tissue and will not colonize the seminiferous tubules. If clogging occurs during the injection, pull out the injection needle, and after removing the clog by breaking the pipette tip, reinsert it into the efferent duct multiple times as long as the efferent duct maintains its original shape and rigidity. It is advisable not to fill 100% of the tubules because this can limit blood supply or induce inflammation in the recipient testis. High pressure in the testis can be relieved by making a small incision in the tunica using a 26–30-gauge needle. The accidental injection of air does not seem to cause a problem, but should be avoided. Pull out the needle and move the testis back into the abdominal cavity. One need not place the testis back in the scrotum. Pull out another testis for transplantation. After the injection, return the testis to the body cavity. Muscle layers as well as skin must be sutured. Typically, transplantation into

Figure 2.3 Spermatogonial transplantation. A glass needle is inserted into the efferent duct of a W mouse. The donor cell suspension is visualized by adding trypan blue solution, which gradually fills up the seminiferous tubules by increasing the injection pressure. Arrow indicates the glass needle tip.

Mutagenesis of Male Germ Line Cells

33

one animal (with two testes) takes about 5–10 min for an experienced manipulator. Researchers are encouraged to refer to a detailed protocol if necessary (Ogawa et al., 1997).

4.4. Optional: Measurement of SSC activity by analyzing the recipient testes According to the donor cell marker type, colonization can be visualized by X-gal staining (LacZ) or under a UV light (EGFP). Within 2–3 weeks following transplantation, donor SSCs form a network or chains of spermatogonia on the basement membrane (Nagano et al., 1999). By 1 month, the cells start to differentiate vertically toward the tubule center. By 3 months posttransplantation, donor-derived spermatozoa appear in the seminiferous tubules. Although the distribution of SSCs in the colony remains unknown, SSCs are believed to preferentially undergo self-renewal divisions at both ends of the colony and start differentiation in the center (Nagano et al., 1999). During the course of colony development, the length of the colony gradually increases (1.73 mm/month), and colonies tend to merge after the long term. Colonies can be safely determined at least 6 weeks after transplantation. An important issue here is the definition of germ cell colony because numerous donor cell clusters are present in the seminiferous tubules and one must set a criteria to define a ‘‘germ cell colony.’’ We slightly modified the criterium of Nagano et al. (1999) and define a colony as when the donor cells occupy the entire seminiferous tubule and the colony is larger than 0.1 mm. Note that not all transplanted SSCs make germ cell colonies: 5–10% of the SSCs are thought to have a repopulating ability (Nagano et al., 1999; Ogawa et al., 2003). In the case of GS cells, about 1–2% of the transplanted cells repopulate seminiferous tubules (Kanatsu-Shinohara et al., 2005b).

4.5. Offspring production from recipient mice Offspring are produced from donor-derived cells either by crossing the recipient mice with wild-type females via natural mating or microinsemination. Both busulfan-treated recipients and W mice will produce offspring from donor SSCs. Transplantation into pup testes results in a more rapid fertility recovery, possibly due to the absence of tight junctions between Sertoli cells in immature testes (Shinohara et al., 2001). Theoretically, about 50% of the F1 offspring from founders are heterozygous because the transgene is transmitted to half of the haploid germ cells after meiosis. Unlike ES and mGS cells, the F1 offspring from the founder are not chimeras, and offspring are directly produced from the transplanted donor cells. However, it is necessary to confirm donor cell origin by PCR or Southern blotting because endogenous SSCs may regenerate to produce fertile sperm. Offspring were produced using primary testis cells by transplanting a few

34

Mito Kanatsu-Shinohara and Takashi Shinohara

hundred SSCs (Ogawa et al., 2000; Shinohara et al., 2001). Single GS cells can also produce offspring after drug selection and in vitro expansion (Kanatsu-Shinohara et al., 2005c, 2006a).

REFERENCES Brinster, R. L., and Avarbock, M. R. (1994). Germline transmission of donor haplotype following spermatogonial transplantation. Proc. Natl. Acad. Sci. USA 91, 11303–11307. Brinster, R. L., and Zimmermann, J. W. (1994). Spermatogenesis following male germ-cell transplantation. Proc. Natl. Acad. Sci. USA 91, 11298–11302. Clouthier, D. E., Avarbock, M. R., Maika, S. D., Hammer, R. E., and Brinster, R. L. (1996). Rat spermatogenesis in mouse testis. Nature 381, 418–421. Creemers, L. B., Meng, X., den Ouden, K., van Pelt, A. M., Izadyar, F., Santoro, M., Sariola, H., and de Rooij, D. G. (2002). Transplantation of germ cells from glial cell linederived neurotrophic factor-overexpressing mice to host testes depleted of endogenous spermatogenesis by fractionated irradiation. Biol. Reprod. 66, 1579–1584. Golestaneh, N., Kokkinaki, M., Pant, D., Jiang, J., Destefano, D., Fernandez-Bueno, C., Rone, J. D., Haddad, B. R., Gallicano, G. I., and Dym, M. (2009). Pluripotent stem cells derived from adult human testes. Stem Cells Dev. 18, 1115–1126. Guan, K., Nayernia, K., Maier, L. S., Wagner, S., Dressel, R., Lee, J. H., Nolte, J., Wolf, F., Li, M., Engel, W., and Hasenfuss, G. (2006). Pluripotency of spermatogonial stem cells from adult mouse testes. Nature 440, 1199–1203. Honoramooz, A., Megee, S., Zeng, W., Destrempes, M. M., Overton, S. A., Luo, J., Galantino-Homer, H., Modelski, M., Chen, F., Blash, S., Melican, D. T., Gavin, W. G., et al. (2008). Adeno-associated virus (AAV)-mediated transduction of male germ line stem cells results in transgene transmission after germ cell transplantation. FASEB J. 22, 374–382. Kanatsu-Shinohara, M., Ogonuki, N., Inoue, K., Miki, H., Ogura, A., Toyokuni, S., and Shinohara, T. (2003a). Long-term proliferation in culture and germline transmission of mouse male germline stem cells. Biol. Reprod. 69, 612–616. Kanatsu-Shinohara, M., Ogonuki, N., Inoue, K., Ogura, A., Toyokuni, S., Honjo, T., and Shinohara, T. (2003b). Allogeneic offspring produced by male germ line stem cell transplantation into infertile mouse testis. Biol. Reprod. 68, 167–173. Kanatsu-Shinohara, M., Inoue, K., Lee, J., Yoshimoto, M., Ogonuki, N., Miki, H., Baba, S., Kato, T., Kazuki, Y., Toyokuni, S., Toyoshima, M., Niwa, O., et al. (2004a). Generation of pluripotent stem cells from neonatal mouse testis. Cell 119, 1001–1012. Kanatsu-Shinohara, M., Toyokuni, S., and Shinohara, T. (2004b). CD9 is a surface marker on mouse and rat male germline stem cells. Biol. Reprod. 70, 70–75. Kanatsu-Shinohara, M., Miki, H., Inoue, K., Ogonuki, N., Toyokuni, S., Ogura, A., and Shinohara, T. (2005a). Long-term culture of mouse male germline stem cells under serum- or feeder-free conditions. Biol. Reprod. 72, 985–991. Kanatsu-Shinohara, M., Ogonuki, N., Iwano, T., Lee, J., Kazuki, Y., Inoue, K., Miki, H., Takehashi, M., Toyokuni, S., Oshimura, M., Ishino, F., Ogura, A., et al. (2005b). Genetic and epigenetic properties of mouse male germline stem cells during long-term culture. Development 132, 4155–4163. Kanatsu-Shinohara, M., Toyokuni, S., and Shinohara, T. (2005c). Genetic selection of mouse male germline stem cells in vitro: Offspring from single stem cells. Biol. Reprod. 72, 236–240. Kanatsu-Shinohara, M., Ikawa, M., Takehashi, M., Ogonuki, N., Miki, H., Inoue, K., Kazuki, Y., Lee, J., Toyokuni, S., Oshimura, M., Ogura, A., and Shinohara, T. (2006a).

Mutagenesis of Male Germ Line Cells

35

Production of knockout mice by random and targeted mutagenesis in spermatogonial stem cells. Proc. Natl. Acad. Sci. USA 103, 8018–8023. Kanatsu-Shinohara, M., Inoue, K., Lee, J., Miki, H., Ogonuki, N., Toyokuni, S., Ogura, A., and Shinohara, T. (2006b). Anchorage-independent growth of mouse male germline stem cells in vitro. Biol. Reprod. 74, 522–529. Kanatsu-Shinohara, M., Inoue, K., Miki, H., Ogonuki, N., Takehashi, M., Morimoto, T., Ogura, A., and Shinohara, T. (2006c). Clonal origin of germ cell colonies after spermatogonial transplantation in mice. Biol. Reprod. 75, 68–74. Kanatsu-Shinohara, M., Inoue, K., Ogonuki, N., Miki, H., Yoshida, S., Toyokuni, S., Lee, J., Ogura, A., and Shinohara, T. (2007). Leukemia inhibitory factor enhances formation of germ cell colonies in neonatal mouse testis culture. Kanatsu-Shinohara, M., Lee, J., Inoue, K., Ogonuki, N., Miki, H., Toyokuni, S., Ikawa, M., Nakamura, T., Ogura, A., and Shinohara, T. (2008a). Pluripotency of a single spermatogonial stem cell in mice. Biol. Reprod. 78, 681–687. Kanatsu-Shinohara, M., Takehashi, M., and Shinohara, T. (2008b). Brief history, pitfalls, and prospects of mammalian spermatogonial stem cell research. Cold Spring Harb. Symp. Quant. Biol. 73, 17–23. Kossack, N., Meneses, J., Shefi, S., Nguyen, H. N., Chavez, S., Nicholas, C., Gromoll, J., Turek, P. J., and Reijo-Pera, R. A. (2009). Isolation and characterization of pluripotent human spermatogonial stem cell-derived cells. Stem Cells 27, 138–149. Kubota, H., and Brinster, R. L. (2008). Culture of rodent spermatogonial stem cells, male germline stem cells of the postnatal animal. In ‘‘Stem Cell Culture,’’ ( J. P. Mather, ed.), pp. 59–84. Academic Press, New York. Kubota, H., Avarbock, M. R., and Brinster, R. L. (2004). Growth factors essential for selfrenewal and expansion of mouse spermatogonial stem cells. Proc. Natl. Acad. Sci. USA 101, 16489–16494. Labosky, P. A., Barlow, D. P., and Hogan, B. L. M. (1994). Mouse embryonic germ (EG) cell lines: Transmission through the germline and differences in the methylation imprint of insulin-like growth factor 2 receptor (Igf2r) gene compared with cmbryonic stem (ES) cell lines. Development 120, 3197–3204. Liu, X., Wu, H., Loring, J., Hormuzdi, S., Disteche, C. M., Bornstein, P., and Jaenisch, R. (1997). Trisomy eight in ES cells is a common potential problem in gene targeting and interferes with germ line transmission. Dev. Dyn. 209, 85–91. Longo, L., Bygrave, A., Grosveld, F. G., and Pandolfi, P. P. (1997). The chromosome makeup of mouse embryonic stem cells is predictive of somatic and germ cell chimaerism. Transgenic Res. 6, 321–328. Matsui, Y., Zsebo, K., and Hogan, B. L. (1992). Derivation of pluripotential embryonic stem cells from murine primordial germ cells in culture. Cell 70, 841–847. Meistrich, M. L., and van Beek, M. E. A. B. (1993). Spermatogonial stem cells. In ‘‘Cell and Molecular Biology of the Testis,’’ (C. Desjardins and L. L. Ewing, eds.), pp. 266–295. Oxford University Press, New York. Meng, X., Lindahl, M., Hyvo¨nen, M. E., Parvinen, M., de Rooij, D. G., Hess, M. W., Raatikainen-Ahokas, A., Sainio, K., Rauvala, H., Lakso, M., Pichel, J. G., Westphal, H., et al. (2000). Regulation of cell fate decision of undifferentiated spermatogonia by GDNF. Science 287, 1489–1493. Nagano, M., Avarbock, M. R., and Brinster, R. L. (1999). Pattern and kinetics of mouse donor spermatogonial stem cell colonization in recipient testes. Biol. Reprod. 60, 1429–1436. Nagano, M., Shinohara, T., Avarbock, M. R., and Brinster, R. L. (2000). Retrovirusmediated gene delivery into male germ line stem cells. FEBS Lett. 475, 7–10. Nagano, M., Brinster, C. J., Orwig, K. E., Ryu, B. Y., Avarbock, M. R., and Brinster, R. L. (2001). Transgenic mice produced by retroviral transduction of male germ-line stem cells. Proc. Natl. Acad. Sci. USA 98, 13090–13095.

36

Mito Kanatsu-Shinohara and Takashi Shinohara

Nagano, M., Watson, D. J., Ryu, B. Y., Wolfe, J. H., and Brinster, R. L. (2002). Lentiviral vector transduction of male germ line stem cells in mice. FEBS Lett. 524, 111–115. Ogawa, T., Are´chaga, J. M., Avarbock, M. R., and Brinster, R. L. (1997). Transplantation of testis germinal cells into mouse seminiferous tubules. Int. J. Dev. Biol. 41, 111–122. Ogawa, T., Dobrinski, I., Avarbock, M. R., and Brinster, R. L. (2000). Transplantation of male germ line stem cells restores fertility in infertile mice. Nat. Med. 6, 29–34. Ogawa, T., Ohmura, M., Yumura, Y., Sawada, H., and Kubota, Y. (2003). Expansion of murine spermatogonial stem cells through serial transplantation. Biol. Reprod. 68, 316–322. Ogawa, T., Ohmura, M., Tamura, Y., Kita, K., Ohbo, K., Suda, T., and Kubota, Y. (2004). Derivation and morphological characterization of mouse spermatogonial stem cell lines. Arch. Histol. Cytol. 67, 297–306. Resnick, J. L., Bixler, L. S., Cheng, L., and Donovan, P. J. (1992). Long-term proliferation of mouse primordial germ cells in culture. Nature 359, 550–551. Ryu, B. Y., Orwig, K. E., Kubota, H., Avarbock, M. R., and Brinster, R. L. (2004). Phenotypic and functional characteristics of spermatogonial stem cells in rats. Dev. Biol. 274, 158–170. Ryu, B. Y., Orwig, K. E., Oatley, J. M., Lin, C. C., Chang, L. J., Avarbock, M. R., and Brinster, R. L. (2007). Efficient generation of transgenic rats through the male germline using lentiviral transduction and transplantation of spermatogonial stem cells. J. Androl. 28, 353–360. Seandel, M., James, D., Shmelkov, S. V., Falciatori, I., Kim, J., Chavala, S., Scherr, D. S., Zhang, F., Torres, R., Gale, N. W., Yancopoulos, G. D., Murphy, A., et al. (2007). Generation of functional multipotent adult stem cells from GPR125þ germline progenitors. Nature 449, 346–350. Shinohara, T., Avarbock, M. R., and Brinster, R. L. (1999). b1- and a6-integrin are surface markers on mouse spermatogonial stem cells. Proc. Natl. Acad. Sci. USA 96, 5504–5509. Shinohara, T., Orwig, K. E., Avarbock, M. R., and Brinster, R. L. (2000). Spermatogonial stem cell enrichment by multiparameter selection of mouse testis cells. Proc. Natl. Acad. Sci. USA 97, 8346–8351. Shinohara, T., Orwig, K. E., Avarbock, M. R., and Brinster, R. L. (2001). Remodeling of the postnatal mouse testis is accompanied by dramatic changes in stem cell number and niche accessibility. Proc. Natl. Acad. Sci. USA 98, 6186–6191. Shinohara, T., Orwig, K. E., Avarbock, M. R., and Brinster, R. L. (2002). Germ line stem cell competition in postnatal mouse testes. Biol. Reprod. 66, 1491–1497. Shinohara, T., Kato, M., Takehashi, M., Lee, J., Chuma, S., Nakatsuji, N., KanatsuShinohara, M., and Hirabayashi, M. (2006). Rats produced by interspecies spermatogonial transplantation in mice and in vitro microinsemination. Proc. Natl. Acad. Sci. USA 103, 13624–13628. Takehashi, M., Kanatsu-Shinohara, M., Inoue, K., Ogonuki, N., Miki, H., Toyokuni, S., Ogura, A., and Shinohara, T. (2007a). Adenovirus-mediated gene delivery into mouse spermatogonial stem cells. Proc. Natl. Acad. Sci. USA 104, 2596–2601. Takehashi, M., Kanatsu-Shinohara, M., Miki, H., Lee, J., Kazuki, Y., Inoue, K., Ogonuki, N., Toyokuni, S., Oshimura, M., Ogura, A., and Shinohara, T. (2007b). Production of knockout mice by gene targeting in multipotent germline stem cells. Dev. Biol. 312, 344–352. Tegelenbosch, R. A., and de Rooij, D. G. (1993). A quantitative study of spermatogonial multiplication and stem cell renewal in the C3H/101 F1 hybrid mouse. Mutat. Res. 290, 193–200. Till, J. E., and McCulloch, E. A. (1961). A direct measurement of the radiation sensitivity of normal mouse bone marrow cells. Radiat. Res. 14, 213–222. Yamagata, K., Yamazaki, T., Miki, H., Ogonuki, N., Inoue, K., Ogura, A., and Baba, T. (2007). Dev. Biol. 312, 419–426.

C H A P T E R

T H R E E

Embryonic In Vivo Electroporation in the Mouse Tetsuichiro Saito Contents 38 38 39 40 40 42 42 43 43 45 45 47 47 48 48 48 49 49 49

1. Introduction 1.1. Overview of in vivo electroporation 1.2. In vivo electroporation in the CNS 2. Materials 2.1. Equipments 2.2. Mouse 2.3. Plasmid and siRNA 3. In Utero Electroporation 3.1. Pulling out of the uterus 3.2. Microinjection 3.3. Electroporation 3.4. Repositioning of the uterus 4. Exo Utero Electroporation 4.1. Exposure of the embryo 4.2. Microinjection and electroporation 4.3. Repositioning of the embryo 5. Analysis of Electroporated Mice Acknowledgments References

Abstract Electroporation combined with surgery is a quick and highly efficient method to transfect nucleic acids into various embryonic tissues in a spatiotemporally restricted manner. Forceps-type electrodes facilitate transfection by delivering electric pulses from outside of the embryo. Many electroporated embryos survive in the pregnant mouse, are born, and are reared. The developing central nervous system (CNS) is a good target for transfection, because there are many neural progenitors adjacent to the ventricle, into which nucleic acids are relatively easily injected. The expression of transfected genes persists in neurons for months. Department of Developmental Biology, Graduate School of Medicine, Chiba University, Chiba, Japan Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77003-8

#

2010 Elsevier Inc. All rights reserved.

37

38

Tetsuichiro Saito

1. Introduction Electroporation is an efficient method to transfect DNA and RNA into a variety of cell types. In vivo electroporation, which uses living animals, is quickly performed with relatively simple equipment, compared with generation of transgenic and gene knock-in mice and construction of recombinant viruses, which require longer time and special facilities. Moreover, transfection can be spatiotemporally limited to areas that receive electric pulses. In vivo electroporation facilitates the functional analysis of genes through their misexpression and the transfection of their small interfering RNAs (siRNAs). Several genes can be simultaneously transfected into the same cells by electroporation of a mixture of plasmids (Saito and Nakatsuji, 2001). Transcriptional regulatory elements and transcriptional activation are also examined in vivo by transfecting reporter plasmids (Miyagi et al., 2004; Saba et al., 2005).

1.1. Overview of in vivo electroporation Based on membrane breakdown studies by electric pulses, electroporation techniques were initially established in the 1980s to introduce DNA into cells (Neumann et al., 1982; reviewed in Chen et al., 2006). In early studies, which adopted a single short electric pulse with a high-voltage exponential decay, sufficient transfection efficiency was accompanied by the death of many cells. The use of repetitive, long-duration, low-voltage, square pulses dramatically improved cell survival (Takahashi et al., 1991). Such repetitive square pulses were successfully applied to chick embryos (Muramatsu et al., 1997) and mouse embryos (Saito, 1999), and they are generally used for in vivo electroporation. Increased cell death has not been observed after in vivo electroporation performed under optimal conditions (Saito, 2006). Electroporation with repetitive pulses appears to be mediated by a complex mechanism that involves membrane permeabilization and electrophoresis rather than the simple transport of DNA through membrane pores (reviewed in Chen et al., 2006; Escoffre et al., 2009). DNA and RNA are injected into the embryo and then introduced into adjacent cells by electroporation. In contrast to chick embryos, it is not easy to manipulate mouse embryos in the uterus. Mouse embryos are maintained in vitro only for a limited time. To circumvent these difficulties, forceps-type electrodes were devised, and electric pulses were applied to embryonic cells inside the uterus by holding it with the electrodes (Saito, 1999; Saito and Nakatsuji, 2001). The site and size of transfected areas depend on the diffusion of the injected solution and the shape of the electrodes. Smaller electrodes are often used to restrict transfected areas (Saba et al., 2003). Transfection is

In Vivo Electroporation

39

unidirectional. Among cells exposed to the injected solution, only cells on the side of the anode are transfected (Saito and Nakatsuji, 2001). The same area can be transfected twice at different stages by repeating in vivo electroporation (Mizutani and Saito, 2005). The main cause of embryonic death appears to be related to the heart. Embryonic survival is greatly increased by placing the electrodes away from the heart (Saba et al., 2003). In vivo electroporation of mouse embryos can be performed in utero or exo utero. During in utero electroporation, the embryo in the uterus is electroporated. The electroporated embryo survives in the pregnant mouse and is born. However, it is difficult to see younger embryos, for example, at embryonic day (E) 11.5, through the uterine wall. To clearly visualize young embryos and/or to perform precise injection into a specific site, the uterine wall is cut, and exo utero electroporation is performed on an embryo that is still connected to its mother. The exo utero electroporated embryo survives in the yolk sac inside the pregnant mouse and can be reared by a foster mother after cesarean section. Transfected genes are limited to the tissues of electroporated mice and are not transferred to their offspring, with the exception of electroporation of the testis. Transgenic spermatozoa obtained after electroporation of the testis can be collected and intracytoplasmically injected into oocytes to generate transgenic mice (Huang et al., 2000).

1.2. In vivo electroporation in the CNS Although it appears that any types of embryonic cells are transfectable, the developing central nervous system (CNS) is preferable for in vivo electroporation, because there are many dividing neural progenitors adjacent to the ventricle, which is a relatively easy target for injection and confines the injected solution. Dividing cells are known to be more transfectable (Goldstein et al., 1989), possibly because DNA transferred into the cytoplasm is efficiently incorporated into the nucleus when the nuclear membrane is disrupted in mitosis. Neurons generated from electroporated progenitors express the transfected gene for a long time, probably because the transfected gene is not diluted owing to the lack of neuronal cell division. In contrast, copy numbers of the transfected gene appear to be decreased in dividing progenitors. Expression of transfected genes disappears in highly multiplying cells, such as cerebellar granule cell precursors, even after efficient electroporation (Kawauchi and Saito, 2008). The developing CNS is suitable for the functional analysis of genes, because the CNS has bilateral symmetry surrounding the ventricle; the transfected side is easily compared with the untransfected side as a negative control on the same section owing to the unidirectional nature of transfection. Genes have been successfully transfected into the telencephalon, diencephalon, midbrain,

40

Tetsuichiro Saito

hindbrain, cerebellum, and spinal cord (Kawauchi and Saito, 2008; Kawauchi et al., 2006; Saba et al., 2003; Saito and Nakatsuji, 2001). Because many types of neurons are generated at a specific embryonic stage, the expression of transfected genes can be restricted to some types of neurons, by choosing the stage at which in vivo electroporation is performed. Transfected genes are specifically expressed in neurons of distinct layers of the neocortex after electroporation at different stages (Mizutani and Saito, 2005; Saito and Nakatsuji, 2001). Whereas both Purkinje and granule cells in the cerebellum express transfected genes after electroporation at E11.5, only granule cells express them after electroporation at E14.5 (Kawauchi and Saito, 2008). This chapter describes protocols for electroporation in the developing CNS. The protocols will also be adapted to other tissues. The determination of the optimal conditions and the setup of the experimental system have been described previously (Saito, 2006; Saito and Nakatsuji, 2001).

2. Materials All experiments should be performed in accordance with protocols approved by the institutional animal care and use committee. The reagents and surgical instruments used during the procedures should be sterile.

2.1. Equipments The experimental setup is shown in Figs. 3.1 and 3.2. The following materials are required:  

     



Square pulse electroporator, for example, CUY21Edit (Bex, Japan) Forceps-type electrodes. Disk electrodes (LF650P3, LF650P5, and LF650P10, with diameters of 3, 5, and 10 mm, respectively), half-ring electrodes (LF651), and customized electrodes are available from Bex. During in vivo electroporation, electrodes should be kept wet by soaking them in saline. Peristaltic pump and tube to deliver warm saline Incubator to keep saline warm at 38  C Fiber optic light source Slide warmer to keep anesthetized pregnant mice warm Operating board: a plastic board with four holes (see Fig. 3.3) Glass capillary pulled by a micropipette puller. The capillary should be cut to an 60 m diameter, and its tip should be labeled with a waterresistant magic marker so that it can be easily found. The volume of injected solution can be measured by lines drawn every 5 l on the capillary. Mouth-controlled pipette (Drummond Scientific)

41

In Vivo Electroporation

A

B

E

C

D

Figure 3.1 A standard setup for in vivo electroporation. Square pulses are generated by an electroporator (A). Warm saline is delivered by a peristaltic pump (B). The electroporator and pump are regulated with on/off foot switches (C, D). The operated mouse is warmed on a slide warmer (E) to recover from the anesthetic.

D

F

E

B

A

C

Figure 3.2 Tools for in vivo electroporation. Electrodes (A) are soaked in saline in a Petri dish. The pregnant mouse is operated on the operating board (B) on a stack of paper towels (C). Fiber optic light source (D) is used to illuminate the embryo. Saline is warmed in an incubator (E) and pumped (F).

42

Tetsuichiro Saito

45 mm

75 mm

Figure 3.3 An anesthetized pregnant mouse is fixed on the operating board (85 mm  135 mm plastic board with four small holes, through which rubber bands are passed).     

Curved forceps, for example, A14 (Natsume Seisakusyo, Japan) Ring forceps, for example, A26 (Natsume Seisakusyo) Fine forceps, such as watchmaker #5 Scissors, gauze, and paper towels Needle and suture, for example, F17-50 braided silk (Natsume Seisakusyo)

For further details, see http://www.m.chiba-u.ac.jp/class/dev/protocol/ apparatus.html.

2.2. Mouse ICR mice (Clea, Japan) are often used, because they bear many pups. Other mouse strains can also be used. The noon of a day when a vaginal plug is found is designated E0.5.

2.3. Plasmid and siRNA 0.1–2 g/l (30–600 nM) of endotoxin-free plasmids in PBS are generally used. If the injection is successful, the expression level of transfected genes increases with increasing concentrations of plasmids, reaching a plateau

In Vivo Electroporation

43

at 150 nM. Covalently closed circular plasmids have a higher transfection efficiency than linearized plasmids. Relatively large (14 kb) plasmids can be successfully transfected (Mizutani and Saito, 2005). siRNA, such as 0.2 g/ l of Stealth RNA (Invitrogen), can repress gene expression specifically and effectively (Kawauchi and Saito, 2008; Muroyama and Saito, 2009). Plasmids carrying a fluorescent protein gene, such as EYFP (enhanced yellow fluorescent protein), are useful to monitor transfection efficiency. Although many cells simultaneously express two genes on different plasmids after electroporation of the mixture of the plasmids, not all of the transfected cells coexpress the two genes, especially when transfection efficiency is low. To coexpress two genes, an internal ribosomal entry site can be used to link them, but the expression level of the downstream gene is usually much lower than that of the upstream gene. Double promoter vectors, such as pCAG–EYFP–CAG, which drives expression of EYFP and an inserted gene by two CAG promoters, have been shown to provide equivalent expression levels of EYFP and the inserted gene in the same cells (Saba et al., 2003; Saito and Nakatsuji, 2001).

3. In Utero Electroporation In utero electroporation is mainly used for embryos that are E13.5 and older, because these embryos are easily observed through the uterine wall. After pulling the uterus out of the pregnant mouse, DNA and/or RNA are injected into the ventricle of the embryo inside the uterus, and electric pulses are delivered with forceps-type electrodes. The procedures from incision to closure of the skin should be finished within 45 min.

3.1. Pulling out of the uterus 1. Anesthetize a pregnant mouse. We use an intraperitoneal injection of 10% Nembutal (Abbott Laboratories) diluted with saline. 2. Place the mouse on its back on the operating board. Put each limb through a rubber band and fix it by pulling the rubber band downwards (Fig. 3.3). Place a stack of paper towels under the operating board to absorb spilled saline (Fig. 3.2). 3. Create an 30-mm-long slit in a piece of gauze (70 mm   150 mm) and center it over the abdomen. Moisten the gauze with 70% ethanol. 4. Through the slit of the gauze, make an 30-mm-long vertical incision along the midline of the skin and then the peritoneum with scissors (Fig. 3.4). 5. Use ring forceps to hold the uterus at the gap between embryos. Pull the uterus out of the abdominal cavity (Fig. 3.5) and gently place it on the

44

Tetsuichiro Saito

Figure 3.4 Vertical incision of the skin. The skin is pinched with forceps and cut with scissors through the slit of the gauze.

Figure 3.5 The uterus is pulled out of the abdominal cavity with ring forceps.

gauze. The embryo and mesometrial blood vessels should not be pinched with the forceps to avoid damage. The pulled-out uterus should be kept wet with warm saline.

In Vivo Electroporation

45

Figure 3.6 Microinjection into the embryo. The embryo is gently held with ring forceps, and DNA and/or RNA is injected into the ventricle.

3.2. Microinjection Inject 1–2 l of DNA and/or RNA solution into the ventricle by using the mouth-controlled micropipette (Fig. 3.6). The embryo is visible under the illumination of a fiber optic light source. Failure of microinjection is the main cause of unsuccessful transfection. If expression is detected in the pia mater and embryonic skin but not in the target site, the microinjection is thought to be unsuccessful. It is important to learn the position of the ventricle by studying textbooks, such as The Atlas of Mouse Development (Kaufman, 1992), and to practice the injection technique by using dyes, such as Indigocarmine (Daiichi Pharmaceutical, Japan). Dyes should be used only for practice to avoid any adverse effects.

3.3. Electroporation Hold the injected embryo through the uterus with forceps-type electrodes (Fig. 3.7) and deliver five square electric pulses of 50 ms duration with a 950 ms interval. The electrodes should always be wet with saline (Fig. 3.2A). Optimal voltages are 22 V for E11.5, 30 V for E12.5, 40 V for E13.5, and 45 V for E15.5 (Saito, 2006). It is important to choose appropriate electrodes that cover the region where transfection is targeted (Fig. 3.8). The transfected area can be more limited by using smaller electrodes. The heart should not be covered with the electrodes to minimize embryonic death.

46

Tetsuichiro Saito

Figure 3.7 Electroporation of the embryo. The embryo is held with forceps-type electrodes, and electric pulses are delivered.

A

D

B

E +

-

C

-

F

+

Figure 3.8 Localized transfection using small and specialized electrodes. Transfection is targeted to the telencephalon (A–C), hindbrain (D), and spinal cord (E, F). The embryo in the yolk sac and electrodes are schematically shown. An area of the dorsal or ventral telencephalon (as shown by hatched boxes) is selectively transfected owing to the unidirectional transfection toward the anode side (B, C). A smaller portion of the spinal cord is transfected by using small electrodes (F).

47

In Vivo Electroporation

3.4. Repositioning of the uterus 1. Pour warm saline on the uterus and gauze. Hold the uterus between embryos with ring forceps, lift it gently, and reposition it into the abdominal cavity. The uterus tends to stick to the gauze, if it gets dry. Warm saline helps the uterus to detach from the gauze and slip into the abdominal cavity. 2. Pour warm saline into the abdominal cavity. 3. Close the peritoneum and then the skin with a needled suture (Fig. 3.9). The peritoneum should be closed tightly to prevent leakage of saline. The skin can be closed with wound clips, if it is not necessary to let the pregnant mouse give birth to and nurse its pups. 4. Warm the mouse at 38  C, until the mouse recovers from the anesthetic.

4. Exo Utero Electroporation Exo utero electroporation is used to see embryos more clearly, especially for younger embryos. Embryos that are E11.5 and older can be used for exo utero electroporation. Electroporation of the developing spinal cord requires exo utero procedures. Embryos remaining in the yolk sac are exposed by an incision in the uterine wall of the pulled-out uterus, microinjected, and electroporated. The electroporated embryos are repositioned into the abdominal cavity without suture of the uterine wall and maintained in the pregnant mouse. It is important to perform exo utero electroporation very carefully to avoid damaging the placenta and mesometrial blood vessels. Exo utero electroporation procedures should be also finished within 45 min. If it is necessary to rear electroporated mice until postnatal stages,

Figure 3.9 Suture of the peritoneum.

48

Tetsuichiro Saito

Figure 3.10 Incision of the uterine wall.

cesarean section is performed at E19.5, and the recovered pups are transferred into a cage with a foster mother.

4.1. Exposure of the embryo The uterus is pulled out as described above. Hold an edge of the uterine wall with fine forceps and cut the uterine wall along the antiplacental side with scissors (Fig. 3.10). The yolk sac and placenta should not be broken. The uterine wall of both uterine horns can be cut. The yolk sac, placenta, uterine wall, and mesometrial blood vessels must be kept wet with warm saline.

4.2. Microinjection and electroporation Inject 1–2 l of DNA and/or RNA solution into the ventricle or the spinal cord central canal as described above. Hold the injected embryo through the yolk sac with forceps-type electrodes and deliver electric pulses as described above. Pulse numbers and optimal voltages are the same as for in utero electroporation. Half-ring and small electrodes are used to transfect genes into large and restricted parts of the spinal cord, respectively (Fig. 3.8E and F).

4.3. Repositioning of the embryo 1. Hold the uterine wall with ring forceps, lift it gently, and reposition it into the abdominal cavity as described above.

In Vivo Electroporation

49

2. Pour warm saline into the abdominal cavity, close the peritoneum and then the skin, as described above.

5. Analysis of Electroporated Mice The expression of transfected genes is detected at least 12 h after electroporation, when the CAG promoter, which is ubiquitous and strong, is used (Saito and Nakatsuji, 2001). After neural progenitors are transfected, the expression can persist in their daughter neurons for months (Saito, 2006). Many embryos survive and express transfected genes under optimal conditions (Saito, 2006; Saito and Nakatsuji, 2001). More than 90% of in utero electroporated embryos survive, and transfected genes are expressed in more than 90% of the surviving embryos. The survival rate of exo utero electroporated embryos is low but usually high enough for gene analyses: more than 70% and 80% for E11.5 and E12.5, respectively. Gene expression can be further restricted by using the Cre recombinase– loxP system (Mizutani and Saito, 2005) and cell-type- or region-specific enhancers, as shown by progenitor-specific expression with a Sox2 enhancer (Miyagi et al., 2004).

ACKNOWLEDGMENTS I thank C. Mamiya for creating the illustrations. This work was supported by JSPS KAKENHI 20240029 and the Mitsubishi Foundation.

REFERENCES Chen, C., Smye, S. W., Robinson, M. P., and Evans, J. A. (2006). Membrane electroporation theories: A review. Med. Biol. Eng. Comput. 44, 5–14. Escoffre, J.-M., Portet, T., Wasungu, L., Teissie, J., Dean, D., and Rols, M.-P. (2009). What is (still not) known of the mechanism by which electroporation mediates gene transfer and expression in cells and tissues. Mol. Biotechnol. 41, 286–295. Goldstein, S., Fordis, C. M., and Howard, B. H. (1989). Enhanced transfection efficiency and improved cell survival after electroporation of G2/M-synchronized cells and treatment with sodium butyrate. Nucleic Acids Res. 17, 3959–3971. Huang, Z., Tamura, M., Sakurai, T., Chuma, S., Saito, T., and Nakatsuji, N. (2000). In vivo transfection of testicular germ cells and transgenesis by using the mitochondrially localized jellyfish fluorescent protein gene. FEBS Lett. 487, 248–251. Kaufman, M. H. (1992). The Atlas of Mouse Development Academic Press. Kawauchi, D., and Saito, T. (2008). Transcriptional cascade from Math1 to Mbh1 and Mbh2 is required for cerebellar granule cell differentiation. Dev. Biol. 322, 345–354.

50

Tetsuichiro Saito

Kawauchi, D., Taniguchi, H., Watanabe, H., Saito, T., and Murakami, F. (2006). Direct visualization of nucleogenesis by precerebellar neurons: Involvement of ventricledirected, radial fibre-associated migration. Development 133, 1113–1123. Miyagi, S., Saito, T., Mizutani, K., Masuyama, N., Gotoh, Y., Iwama, A., Nakauchi, H., Masui, S., Niwa, H., Nishimoto, M., Muramatsu, M., and Okuda, A. (2004). The Sox-2 regulatory regions display their activities in two distinct types of multipotent stem cells. Mol. Cell. Biol. 24, 4207–4220. Mizutani, K., and Saito, T. (2005). Progenitors resume generating neurons after temporary inhibition of neurogenesis by Notch activation in the mammalian cerebral cortex. Development 132, 1295–1304. Muramatsu, T., Mizutani, Y., Ohmori, Y., and Okumura, J. (1997). Comparison of three nonviral transfection methods for foreign gene expression in early chicken embryos in ovo. Biochem. Biophys. Res. Commun. 230, 376–380. Muroyama, Y., and Saito, T. (2009). Identification of Nepro, a gene required for the maintenance of neocortex neural progenitor cells downstream of Notch. Development 136, 3889–3893. Neumann, E., Schaefer-Ridder, M., Wang, Y., and Hofschneider, P. H. (1982). Gene transfer into mouse lyoma cells by electroporation in high electric fields. EMBO J. 1, 841–845. Saba, R., Nakatsuji, N., and Saito, T. (2003). Mammalian BarH1 confers commissural neuron identity on dorsal cells in the spinal cord. J. Neurosci. 23, 1987–1991. Saba, R., Johnson, J. E., and Saito, T. (2005). Commissural neuron identity is specified by a homeodomain protein, Mbh1, that is directly downstream of Math1. Development 132, 2147–2155. Saito, T. (1999). Analysis of mammalian neuronal diversity using in vivo electroporation. The 607th National Institute of Genetics Colloquium, Mishima, Japan, p. 1. Saito, T. (2006). In vivo electroporation in the embryonic mouse central nervous system. Nat. Protoc. 1, 1552–1558. Saito, T., and Nakatsuji, N. (2001). Efficient gene transfer into the embryonic mouse brain using in vivo electroporation. Dev. Biol. 240, 237–246. Takahashi, M., Furukawa, T., Saitoh, H., Aoki, A., Koike, T., Moriyama, Y., and Shibata, A. (1991). Gene transfer into human leukemia cell lines by electroporation: Experience with exponentially decaying and square wave pulse. Leuk. Res. 15, 507–513.

C H A P T E R

F O U R

Current Applications of Transposons in Mouse Genetics Adam J. Dupuy Contents 54 54 55 55 61 61 63 64 65 65 67 67

1. Introduction 2. Molecular Characteristics of TEs with Activity in Mice 2.1. Class I elements—Retrotransposons 2.2. Class II elements—DNA transposons 3. Applications of TEs in Mouse Genetics 3.1. Germline mutagenesis 3.2. Transgenesis 3.3. Gene therapy 3.4. Induced pluripotent stem cells 3.5. Cancer genetics 4. The Future of TEs in Mouse Genetics References

Abstract Transposable elements (TEs) have been used to study the genetics of a wide variety of species, including prokaryotes, plants, yeast, and Drosophila. The use of TEs to study mouse genetics has previously not been possible as mice do not have endogenous, highly active TEs like other organisms. Over the past decade, however, two retrotransposons (class I TEs) and four DNA transposons (class II TEs) have been developed that are active when in mice. These elements have been used for a variety of applications including germline mutagenesis, transgenesis, gene therapy, the production of induced pluripotent stem cells, and cancer genetics. The molecular characteristics are summarized for each TE currently used in the mouse. In addition, the current applications of these TEs are discussed.

Department of Anatomy and Cell Biology, Carver College of Medicine, University of Iowa, Iowa City, Iowa, USA Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77004-X

#

2010 Elsevier Inc. All rights reserved.

53

54

Adam J. Dupuy

1. Introduction In 1950, Barbara McClintock published her initial work describing the Activator (Ac) and Dissociator (Ds) transposable elements (TEs) in maize (McClintock, 1950). Her work uncovered a new mechanism of non-Mendelian inheritance and established a new field of genetics that has impacted many other fields. In the decades following McClintock’s pioneering work, a wide array of TEs was identified in a variety of organisms (Roberts et al., 2008; Seberg and Petersen, 2009). Although there is significant diversity among TEs with regard to structure and molecular characteristics, one shared feature is their significant contribution to the evolutionary process—diverse biological processes such as antibiotic resistance in bacteria and adaptive immunity in vertebrate organisms are attributed to the activity of TEs (Flajnik and Kasahara, 2010; Siguier et al., 2006). Another significant result of McClintock’s work was the practical application of TEs to the genetic analysis of many species. Searles et al. (1982) were the first to make use of an endogenous TE in eukaryotes to isolate mutants in Drosophila. This approach has been especially useful in the functional genomic analysis of yeast, Caenorhabditis elegans, and D. melanogaster (Bazopoulou and Tavernarakis, 2009; Cooley et al., 1988; RossMacdonald et al., 1999). The development of TEs as genetic tools certainly contributed to the establishment of these species as model organisms. Unfortunately, the field of mouse genetics did not immediately benefit from the use of TEs, since mice do not have endogenous highly active TEs that would be suitable for mutagenesis experiments, and the more active TEs that are commonly used in invertebrate organisms do not function efficiently in mice. Despite these early setbacks, a number of groups continued to search for TEs that would function in mice for the purpose of performing genetic screens. This chapter describes recent progress in the description and application of TEs in mice.

2. Molecular Characteristics of TEs with Activity in Mice The application of TEs to the study of mouse genetics has become possible only in the past decade. In this time, five different TEs have been shown to have activity in vivo, and two of these (Sleeping Beauty, piggyBac) are currently in wide use for a variety of applications. The following is a summary of these TEs, including information about their transposition efficiency, molecular characteristics, and mutagenic potential.

Transposons in Mouse Genetics

55

2.1. Class I elements—Retrotransposons TEs are grouped into two classes according to transposition mechanism and transposon structure. Class I elements are often referred to as retrotransposons because their mechanism of transposition resembles that of retroviruses: an RNA intermediate is produced and is subsequently reverse transcribed to generate a new copy of the transposon at another genomic location (Ostertag and Kazazian, 2001). This family of TEs includes both long and short nuclear interspersed elements (LINEs, SINEs). Another feature that makes ***class I TEs—both LINEs and SINEs— unique is their active transposition in mice. This fact spurred several groups to investigate the possibility of using engineered retrotransposons to perform mutagenesis screens in mice. Ostertag et al. (2002) were the first to demonstrate transposition of an engineered human L1 retrotransposon in the mouse germline. However, the rate of retrotransposition (two events in 135 offspring) was too low to be useful. Subsequent work by An et al. (2006) produced an optimized L1 element, called ORFeus, which was later shown to have much higher activity in the mouse germline (0.33 de novo insertions per offspring) than the original L1. In addition, retrotransposition of the ORFeus element was readily detected in the somatic cells of mice. Finally, analysis of 200 ORFeus insertion sites suggested that this element has little insertion site bias, and thus may be useful as an insertional mutagen. Currently, there are several obstacles to the practical application of engineered retrotransposons, such as ORFeus. First, the ability to modify the structure of these elements is limited due to the somewhat stringent sequence requirements for this type of TE; this raises challenges in the construction of retrotransposons that include mutagenic cassettes (e.g., a gene trap) is more challenging. Another major limitation of the current engineered L1 elements is their relatively low rate of transposition. Although ORFeus shows a > 20-fold increase in transposition rate over endogenous L1 elements, the estimated mutagenesis rate remains too low to make ORFeus useful for mouse germline mutagenesis. Nevertheless, in somatic cells this lower mutagenesis frequency could be sufficient to induce or accelerate tumors in mice. In fact, the study of mouse cancer genetics has already benefited from the application of class II TEs (see below), and it is possible that a similar application of retrotransposons would also be successful.

2.2. Class II elements—DNA transposons The TEs initially identified by Barbara McClintock belong to a class of elements referred to as DNA transposons. As the name implies, transposition of these elements (i.e., class II transposons) does not involve an RNA intermediate. Instead, endogenous DNA transposons encode an enzyme that mediates the transposition reaction. Referred to as a transposase, this

56

Adam J. Dupuy

enzyme binds to sequences near the ends of the transposon and then mediates cleavage of the transposon from the donor site, and its integration at a new site within the host cell genome. The transposase binding sites are located within two repeat sequences that define the 50 and 30 boundaries of the DNA transposon. These repeat sequences typically have an inverted orientation relative to each other, and are thus often referred to as inverted repeats (IRs) or inverted terminal repeats (ITRs). An autonomous DNA transposon consists of a transposase gene flanked by its own IRs. This functional unit is quite literally a mobile genetic element capable of changing its location within the host cell genome. Active autonomous DNA transposons have been identified in a variety of species. In some cases, such as the P element in Drosophila, mobilization of endogenous DNA transposons has been used to perform mutagenesis experiments (Searles et al., 1982). However, this approach is not viable for the study of mouse genetics as no functional members of this group of TEs have been identified in the mouse genome. Another feature of DNA transposons is that nonautonomous elements can be mobilized along with them when the transposase enzyme is provided in trans. The ability to generate nonautonomous transposons makes this suitable for a number of practical applications, since any DNA fragment flanked by IRs can be introduced into a host cell genome in a transposasedependent manner. For example, DNA transposons can be engineered to deliver a transgene or mutagenic cassette such as a gene trap. These transposon vectors can then be used in a variety of applications by simply altering the delivery method for the transposon vector and the transposase. To date, four distinct DNA transposons have been shown to function in mice. From a practical perspective, a number of molecular characteristics are typically examined with each DNA transposon, as these features will determine the type of applications for which each type of DNA transposon can be used. These features often include consideration of DNA cargo capacity, transposition efficiency, and integration site bias. Local hopping is another significant feature of a subset of DNA transposons. This refers to the propensity of DNA transposons to integrate into a site physically linked to the donor site. Local hopping leads to an increased rate of transposon insertion in a region flanking the transposon donor site, and thus reduces the efficiency of whole-genome mutagenesis. The following is a summary of the four DNA transposons known to be active in mice. Current applications of each transposon type are described, as are its molecular characteristics (Table 4.1). 2.2.1. Tol2 The Tol2 transposon belongs to the hAT family (hobo/Ac/Tam3) of DNA transposons and was the first autonomous DNA transposon identified in a vertebrate organism (Kawakami et al., 1998). Several experiments have

Table 4.1 Properties of cut-and-paste transposons that function in mice Type

Family

Target sequence

Excision footprint

Cargo capacity

Integration site bias

Local hopping (avg. freq.)

Sleeping Beauty piggyBac Tol2 Minos

Tc1/mariner hAT hAT Tc1/mariner

TA TTAA Nonspecific TA

TACWGTA None Imprecise Imprecise

1.8–7 kb At least 12 kb At least 66 kb Unknown

Weak Modest Unknown Weak

25–75% < 10%  85%  60%

58

Adam J. Dupuy

shown that the Tol2 transposon has activity in a broad range of species with transposition demonstrated in Drosophila, zebrafish, frog, chicken, human, and mouse embryonic stem cells (Balciunas et al., 2006; Hamlet et al., 2006; Kawakami and Noda, 2004; Kawakami et al., 2004; Sato et al., 2007; Urasaki et al., 2008). Unlike most DNA transposons, Tol2 does not appear to prefer a specific target sequence for insertion. However, Tol2-induced transposition does lead to the duplication of an 8-base pair sequence flanking the newly integrated transposon (Kawakami et al., 2000). A final feature that distinguishes Tol2 is a high cargo capacity compared to that of most DNA transposons (Balciunas et al., 2006; Suster et al., 2009). The efficiency of Tol2-induced transposition has been studied in both mouse embryonic stem cells and in the mouse germline. Using a transposon bearing a neomycin resistance marker to select for Tol2-induced transposon insertions in mouse ES cells, Kawakami and Noda (2004) showed a dosedependent increase in colony number ( ). While the transfection conditions showing the highest efficiency in this experiment resulted in only a approximately sevenfold increase in colony number over background, the relationship between the number of drug resistant colonies and the amount of Tol2 transposase delivered was nearly linear. Thus, it is likely that a higher rate of transposition could be obtained. Keng et al. (2009a) recently characterized a Tol2 mutagenesis system in the mouse germline. Analysis of the transposon insertion sites indicated that most of the gene disruption events were the result of local hopping, with 85% of Tol2-induced insertions occurring within a 4-megabase (mb) region flanking the transposon donor site. Nevertheless, the transposition efficiency was reasonable with each offspring inheriting an average of three de novo insertion events. Thus, Tol2 may be useful for mutagenesis screens in mice, particularly for those that seek to saturate a specific region of the genome. The available Tol2 integration site data from mouse mutagenesis experiments was generated using gene trap transposons. As the integration sites in this case were selected, this study was not useful in identifying any integration site bias for Tol2. However, a recent Tol2 mutagenesis screen performed in zebrafish cloned 287 unselected Tol2 transposon insertion sites (Kondrychyn et al., 2009). Analysis of these integration sites showed that 32% and 7% of Tol2 insertion events occurred within introns and exons, respectively. These frequencies likely reflect a bias for insertion within genes. However, this conclusion will not be definitive until the complete annotated zebrafish genome is available. 2.2.2. Minos The Minos TE belongs to the Tc1/mariner class of DNA transposons and was originally identified in D. hydei (Franz and Savakis, 1991). Subsequent work showed that Minos has reasonable activity in cultured human cells,

Transposons in Mouse Genetics

59

and this observation spurred investigation into the use of Minos in mice (de Wit et al., 2010; Zagoraiou et al., 2001). The efficiency of Minosmediated transposition appears comparable to that of other DNA transposons tested in the mouse germline (2–3 de novo insertions per gamete). Minos also displays a local hopping tendency, with 60% of insertions occurring within a 50–90-mb region flanking the initial donor site. However, Minos does not show a significant integration site bias since only 20% of the 125 mapped insertion events occurred within a gene (de Wit et al., 2010). Other molecular features of the Minos transposon (e.g., cargo capacity) are yet to be determined. However, the published data suggest that this transposon functions much like the more widely used Tc1/mariner family member, Sleeping Beauty. 2.2.3. piggyBac The piggyBac (PB) transposon system was initially identified as an endogenous TE in the genome of the cabbage looper moth, Trichoplusia ni (Cary et al., 1989). Subsequent work showed that the PB transposon is active in a variety of insect species, including several Drosophila species (Lobo et al., 1999). Ding et al. (2005) were the first to demonstrate the use of PB for germline mutagenesis in the mouse. Although the overall transposition efficiency was modest (1 de novo insertion per gamete), nearly 70% of PB-mediated insertions occurred within genes. Subsequent work by Liang et al. (2009) using PB in mouse ES cells showed a similar, albeit weaker, preference for insertion within transcription units. This disproportionate rate of gene disruption suggests that PB will be useful for mouse mutagenesis. However, the molecular origin of this integration bias has not been studied, which makes it difficult to determine how significantly this integration site bias will impact the number of gene targets that can be mutated using PB mutagenesis. Nevertheless, several features of the PB transposon make it appealing for a variety of applications. Most transposases leave behind a footprint at the site of transposon excision (Table 4.1). Thus, transposon skipping—the ongoing remobilization of transposons within the host cell genome—introduces transposon footprints at sites once occupied by transposons. This is a particular concern when using TEs to perform mutagenesis screens, as many footprints introduce a frameshift mutation when they occur within a coding exon. Such mutations are not tagged by a transposon, and this complicates their identification. The PB transposon is unique among characterized TEs as it does not generate a footprint; instead it cleanly repairs the excision site. Another advantage of the PB transposon is that it does not display a strong local hopping tendency. A recent report by Liang et al. (2009) failed to detect any local hopping preference when PB transposons were mobilized in mouse ES cells, whereas a previous mutagenesis screen had detected a weak preference (9% of insertion events) for integration within a

60

Adam J. Dupuy

100-kilobase (kb) region flanking the original donor site (Wang et al., 2008). Despite the slight differences in these studies, both indicate that PB has the lowest rate of local hopping among the DNA transposons currently in use. 2.2.4. Sleeping Beauty The origin of the Sleeping Beauty (SB) transposon system is unique. One trait common to Tol2, Minos, and PB is that each is an endogenous TE that was isolated from a host organism and subsequently shown to be active in mice. By contrast, SB is the first example of an engineered TE. It was initially created by Ivics et al. (1997), based on a consensus sequence generated from the alignment of inactive Tc1/mariner elements cloned from multiple species of salmonid fish. This initial work also demonstrated that SB is active in cultured mouse and human cells. Additional experiments studied the molecular characteristics of SB in cultured cells. Ivics et al. (1997) had already demonstrated that the target sequence for SB insertion is a TA dinucleotide pair. Due to the limited number of cloned SB insertion events, it was not immediately clear if SB showed any integration site bias aside from the requirement for this dinucleotide. Subsequent studies investigated the integration site preferences of SB in cultured mouse cells, revealing that SB does not have a strong integration site bias (Yant et al., 2005). This is a desirable quality, which makes SB a candidate system for use in performing unbiased insertional mutagenesis screens. In spite of these important advantages, SB has also been shown to have several notable limitations. First, the efficiency of SB-induced transposition is inversely correlated with transposon size, with an upper limit of  6–7 kb (Geurts et al., 2003). This restricted cargo capacity limits the flexibility in designing SB transposons for a variety of applications. Second, several groups examined the donor sites following SB-mediated excision, and found that SB produces a characteristic footprint that leads to the addition of five nucleotides (Izsvak et al., 2004; Luo et al., 1998). As previously discussed, the introduction of a footprint at a donor site has the potential to cause a frameshift mutation in a coding exon at a site once occupied by a transposon. Notwithstanding these technical concerns, several groups began to explore the use of SB for mutagenesis of the mouse genome. Luo et al. (1998) were the first to characterize SB for this purpose, introducing SB transposons into mouse embryonic stem cells. Unfortunately, the results of this experiment showed that although SB is active in mouse embryonic stem cells, it is not efficient enough in this context to be useful. However, three independent groups demonstrated that SB functions at a much higher efficiency in the mouse germline than in mouse ES cells, with efficiencies ranging from 1 to 2 de novo insertions per gamete (Dupuy et al., 2001; Fischer et al., 2001; Horie et al., 2001). An important observation in this

Transposons in Mouse Genetics

61

regard was that SB shows a local hopping tendency, with 50–75% of integration events occurring on the donor chromosome (Carlson et al., 2003; Horie et al., 2003). Overall, these publications suggested that SB could be useful as an insertional mutagen in the mouse. However, its moderate to low efficiency has prevented its wide use in studying mouse genetics. In an effort to improve the SB system, several groups have studied its transposition mechanism in greater detail. Some of these efforts led to the production of modified versions of the transposon IRs and of the SB transposase, and these tools have improved transposition efficiencies in some applications (Baus et al., 2005; Cui et al., 2002; Geurts et al., 2003; Mates et al., 2009). Other experiments have focused on protein interactions between the SB transposase and host cell proteins. For example, using a knockdown approach, Zayed et al. (2003) showed that the efficiency of SB-induced transposition is highly dependent on the DNA-bending protein HMGB1. Another series of experiments showed that efficient SB-induced transposition requires members of the nonhomologous end-joining DNA repair pathway (Izsvak et al., 2004). Although the identification of host cell factors required for efficient SB transposition has not yet had a practical outcome with respect to improving transposition efficiency, it has provided significant insight into how the SB transposase functions.

3. Applications of TEs in Mouse Genetics The application of DNA transposons in mouse genetics is a recent achievement. Nevertheless, they—and in particular PB and SB—are now routinely used for certain applications. Currently, there are five major areas in which transposons are being used: germline mutagenesis, transgenesis, gene therapy, the production of induced pluripotent stem (iPS) cells, and cancer genetics. In general, the technical requirements for each of these applications have determined which transposons are best suited to the task.

3.1. Germline mutagenesis A major goal of the mouse genetics community has been to develop approaches that facilitate phenotype-driven forward genetic screens. It is not surprising that DNA transposons have been investigated for this application since an efficient insertional mutagen, such as a transposon, would facilitate this goal. In fact, germline mutagenesis is the only application in which all four transposons have been evaluated to some extent.

62

Adam J. Dupuy

Generally, the same approach has been used for each transposon system to determine the overall rate of mutagenesis. This approach involves the production of transgenic mice that carry one or more copies of a mutagenic (i.e., a gene trap) transposon. These nonautonomous transposons are then mobilized by breeding these animals to a second mouse strain that expresses the specific transposase, either ubiquitously or specifically in the germline. The resulting double transgenic mice are then bred to wild-type mates to isolate de novo transposition events that occurred in the germline of the double transgenic parent. Mutagenized offspring are then evaluated to estimate the overall mutagenesis frequency and to determine the genomic distribution of transposon insertion events. Fischer et al. (2001) were the first to demonstrate germline transposition in the mouse using SB. Soon thereafter, two additional independent groups also showed the feasibility of germline SB mutagenesis (Dupuy et al., 2001; Horie et al., 2001). Likewise, germline mutagenesis in the mouse has been demonstrated for Tol2, Minos, and PB (de Wit et al., 2010; Ding et al., 2005; Keng et al., 2009a; Wu et al., 2007; Zagoraiou et al., 2001). Despite the increasing number of DNA transposons with demonstrated activity in the mouse germline, none has yet to be proven useful in performing whole-genome forward genetic screens in mice. First, current approaches for transposition by each system give rise to a suboptimal rate of transposition (1–3 de novo insertions per gamete). Thus, the number of mice would be required to obtain a reasonable coverage of the genome in a mutagenesis screen would be prohibitive. This inefficiency is exacerbated by the local hopping tendency exhibited by most of the transposons. With the exception of insertion by PB, which does not display local hopping, insertion by DNA transposons result in 25–85% mobilization into a region flanking the donor site (Table 4.1). Thus, a large number of transposition events fail to contribute to genome-wide mutagenesis. While genome-wide transposon mutagenesis in the mouse germline is not currently feasible, several groups have taken advantage of this large percentage of local hopping to saturate a genomic region. For example, Keng et al. (2005) obtained germline transposon-induced mutations in all genes within a 4-mb region around a SB transposon donor site. A second study combined SB mutagenesis with a balancer chromosome to obtain recessive mutations within a defined region of mouse chromosome 11 (Geurts et al., 2006). In addition to demonstrating the feasibility of regional mutagenesis using SB, this approach also revealed that the transposon donor site was frequently subject to microdeletions, inversions, and translocations. These events were likely caused by the frequent double-stranded DNA breaks introduced by the SB transposase during transposon excision. While these complex transposition-induced chromosomal rearrangements complicate the use of SB for regional mutagenesis, this problem may not apply to all DNA transposons. For instance, a recent report

Transposons in Mouse Genetics

63

revealed that Tol2 mutagenesis in the mouse germline does not cause similar chromosomal rearrangements at the transposon donor site (Keng et al., 2009a). This observation, taken together with the relatively high rate of local hopping by Tol2, suggests that Tol2 is well suited for regional mutagenesis. Transposon mutagenesis of mouse ES cells is another approach that has been utilized to perform forward genetic screens. For example, Wang et al. (2009) performed PB mutagenesis in Blm-deficient ES cells in order to obtain homozygous insertion alleles. One major advantage of transposon mutagenesis of ES cells is that the transposons are delivered via transfection with transposition then occurring from the episomal plasmid DNA into the ES cell genome. Therefore, local hopping is not a concern with this type of transposon delivery. In terms of forward genetic screens, this approach would only be useful for isolating mutations producing a phenotype that can be scored in ES cells, and thus would not be practical for many applications. Although transposon mutagenesis is currently not feasible as a means of performing whole-genome forward genetic screen, this goal could eventually be achieved if transposons with higher efficiencies in the mouse germline were identified. Another alternative is to take advantage of regional mutagenesis with transposons like Tol2. Genome-wide mutagenesis could be achieved by generating mice that harbor transposons at numerous sites throughout the genome instead of at a single transposon donor site.

3.2. Transgenesis The production of transgenic mice was first reported nearly 30 years ago (Gordon et al., 1980). Since that time, transgenesis has become a standard approach that is widely used in virtually all areas of research involving mice. The efficiency of transgenic mouse production has steadily improved, and current approaches can introduce even large transgenes, such as a bacterial artificial chromosome (BAC). Nevertheless, current methods of transgenesis have several notable limitations. First, standard transgenesis relies on random incorporation of recombinant DNA. Thus, transgene insertions can lead to complex chromosomal rearrangements at the site of transgene insertion. Some transgene insertion events produce undesirable phenotypes when the insertion site lies within a critical gene. From a more practical perspective, the unpredictable structure of transgene insertion sites makes it difficult to determine transgene zygosity in later generations. A second limitation is that efficient production of transgenic founder animals can be achieved in only a limited number of mouse strains. Also, many backcross generations may be required to move the transgene onto the desired genetic background. Finally, although the recent development of bacterial recombination systems

64

Adam J. Dupuy

has allowed investigators to easily generate BAC transgenes, BAC transgenesis in mice is still inefficient not widely available. Nearly all of these limitations can likely be addressed by using transposons in generating transgenic mice. The first demonstration of such an approach was published in 2002. In this study, plasmids carrying a SB transposon and an mRNA encoding the transposase enzyme were coinjected into one-cell mouse embryos (Dupuy et al., 2002). This experiment produced a transgenesis rate of 45%—significantly higher than can be achieved by a standard approach. In addition, the resulting transgenic founder animals showed multiple transposase-mediated transgene insertions that were cleanly inserted into TA dinucleotide sites with no indication of chromosomal rearrangements. The PB transposon system has also been used to produce transgenic mice (Shinohara et al., 2007). Shinohara et al. used an intracytoplasmic sperm injection method to deliver a plasmid carrying a transgene embedded in a PB transposon, as well as a PB transposase expression cassette, into one-cell mouse embryos. This method produced a transgenesis frequency as high as 70%, suggesting that PB could be extremely effective in the production of transgenic mice. Finally, recent work has demonstrated the great potential of the Tol2 transposon system in producing transgenic mice (Suster et al., 2009). In this case, the significant cargo capacity of Tol2 transposons was used to introduce a 70-kb BAC into one-cell mouse embryos in a transposase-dependent manner. This approach also led to a surprisingly high rate of transgenesis ( 30%), and facilitated the production of transgenic mice containing a single-copy of a BAC transgene cleanly inserted into the genome by Tol2.

3.3. Gene therapy The use of transposons as a nonviral gene delivery tool has become more common in recent years. In particular, numerous groups have begun to explore the use of transposons in gene therapy applications. Yant et al. (2000) were the first to demonstrate such an application, correcting a mouse model of hemophilia B using the SB system to stably introduce transposons expressing coagulation factor IX to mouse hepatocytes in vivo. Later work used a similar approach to correct a mouse model of tyrosinemia type I, hemophilia A, and mucopolysaccharidosis (Aronovich et al., 2007; Liu et al., 2006; Montini et al., 2002). While both of these studies targeted transposon delivery to hepatocytes, other works have shown that SB transposons can be introduced into type II pneumocytes and endothelial cells in the lung (Belur et al., 2003; Liu et al., 2004). Although SB is the transposon that has been predominantly used in gene therapy applications thus far, it is likely that other transposons, such as PB, will also be effective.

Transposons in Mouse Genetics

65

3.4. Induced pluripotent stem cells Stem cell research has recently become a major area of interest for the biomedical research community. In part, this interest is driven by the potential use of stem cells to treat a panoply of human diseases. However, the use of human embryonic stem cells to achieve this goal has initiated intense debate. In 2006, Takahashi and Yamanaka reported a method for converting primary mouse fibroblasts into pluripotent stem cells Takahashi and Yamanaka, 2006. These cells, referred to as iPS cells, were created by expressing four genes (Oct3/4, Sox2, c-Myc, Klf4) in fibroblasts grown under standard mouse ES cell culture conditions. Later work showed that a similar approach could be used to generate human iPS cells (Yu et al., 2007). This work established the possibility to use somatic cells as the source of stem cells for therapeutic purposes, thus bypassing the ethical concerns raised by the use of stem cells of embryonic origin. Despite the initial enthusiasm for iPS cells, there were some safety concerns about the initial method used to produce them. Takahashi and Yamanaka (2006) had used retroviral vectors to deliver the gene cocktail into fibroblasts to convert them to iPS cells. Thus the iPS cells, and any cells derived from them, would carry proviruses that could potentially continue to express these genes. The major concern was that the combination of insertional mutagenesis induced by the recombinant retroviruses, along with the continued expression of the c-Myc proto-oncogene, would lead to increased cancer risk when iPS cells were introduced into patients. Three independent groups recently reported that efficient iPS cells can be generated using a PB transposon to deliver the reprogramming factors (Kaji et al., 2009; Woltjen et al., 2009; Yusa et al., 2009). This approach has several key advantages over the retroviral delivery system that was used in the initial studies. First, the cargo capacity of the PB system makes it possible to deliver all four reprogramming factors on a single transposon. Second, the PB transposon can be efficiently removed from the iPS cell genome once reprogramming is complete, by simply reexpressing the PB transposase in the iPS cells. Finally, PB does not leave a footprint at the transposon insertion site, and thus the iPS cell genome does not retain any vector sequences. In light of these advantages, use of PB transposons will likely play an important role in the development of iPS cell technology for many applications.

3.5. Cancer genetics Retroviral insertional mutagenesis has been used to study the genetics of hematopoietic and mammary cancers in mice for several decades (Uren et al., 2005). This approach has made use of naturally occurring retroviruses, such as the murine leukemia virus and the mouse mammary tumor virus, to

66

Adam J. Dupuy

induce tumors in mice. However, similar retroviruses are not available to model other forms of cancer commonly seen in the human population. Previous work characterizing the SB transposon system showed that it can function in a variety of tissues (Horie et al., 2001). This suggested that, like retroviruses, transposons can be used to study cancer genetics—but in the absence of many of the limitations imposed by retroviruses. Two publications recently demonstrated the validity of transposonmediated insertional mutagenesis as an approach to study cancer in mice (Collier et al., 2005; Dupuy et al., 2005). Each of these studies used transposons (e.g., T2/Onc) that had been engineered to induce both gain- and loss-of-function mutations when mobilized in mouse somatic cells. Collier et al. (2005) showed that mobilizing these mutagenic transposons could accelerate tumors induced by loss of the p19 tumor suppressor. Dupuy et al. (2005) showed that tumors could be induced in wild-type mice by increasing the number of mutagenic transposons and using a ubiquitously expressed transposase knock-in allele (e.g., RosaSB). In both cases, molecular characterization of transposon-induced mutations identified both known and novel cancer genes involved in sarcoma and lymphoma. While these early studies were an important demonstration of the usefulness of transposon mutagenesis to identify cancer genes, the approaches used had several significant limitations. First, ubiquitous transposition of mutagenic transposons frequently induced embryonic lethality. Furthermore, mice that escaped embryonic lethality invariably went on to develop aggressive lymphomas before other tumor types could develop (Dupuy et al., 2005). Recent work by Collier et al. (2009), however, demonstrated that the embryonic lethality could be eliminated by reducing the number of mutagenic transposons mobilized in the somatic cells of mice. Nevertheless, the inability to control the tumor type produced by transposon mutagenesis prevented the broader use of these mouse strains to model cancer. Fortunately, DNA transposons provide significant flexibility for modifying the transposon structure or providing better control of the transposase expression. A modified version of the original T2/Onc transposon has recently been described (Dupuy et al., 2009). This transposon, called T2/ Onc3, was designed to produce stronger gain-of-function mutations in epithelial cells, and ubiquitous T2/Onc3-induced mutagenesis was subsequently shown to induce a variety of carcinomas in mice. Dupuy et al. (2009) also described a lox-stop-lox approach to produce a Cre-inducible SB transposase allele (RosaSBaseLsL), which provides the ability to restrict transposon mutagenesis to specific sites within the mouse when it is combined with a tissue-specific Cre strain. Two recent publications demonstrated the validity of this concept, using the RosaSBaseLsL allele to generate models of both colorectal cancer and hepatocellular carcinoma (Keng et al., 2009b; Starr et al., 2009). Combining the use of the modified T2/Onc3

Transposons in Mouse Genetics

67

transposon and the Cre-inducible RosaSBaseLsL allele will likely make it possible to model many forms of human cancer in the mouse.

4. The Future of TEs in Mouse Genetics The development of TEs for use in studying mouse genetics is a relatively recent development occurring within the past decade. In addition, several DNA transposon systems—Frog Prince, Hsmar1, Passport, Harbinger3, Mos1—have been shown to function in cell culture, but have not yet been tested in mice (Clark et al., 2009; Emelyanov et al., 2006; Miskey et al., 2003, 2007; Sinzelle et al., 2008; Wu et al., 2006). One or more of these transposons may also be useful in performing mouse mutagenesis experiments. As described here, the application of TEs in some areas of research has already begun to have a significant impact. It is likely that the use of TEs will continue to expand within the mouse genetics community as additional genetic tools and mouse strains become available.

REFERENCES An, W., et al. (2006). Active retrotransposition by a synthetic L1 element in mice. Proc. Natl. Acad. Sci. USA 103, 18662–18667. Aronovich, E. L., et al. (2007). Prolonged expression of a lysosomal enzyme in mouse liver after Sleeping Beauty transposon-mediated gene delivery: Implications for non-viral gene therapy of mucopolysaccharidoses. J. Gene Med. 9, 403–415. Balciunas, D., et al. (2006). Harnessing a high cargo-capacity transposon for genetic applications in vertebrates. PLoS Genet. 2, e169. Baus, J., et al. (2005). Hyperactive transposase mutants of the Sleeping Beauty transposon. Mol. Ther. 12, 1148–1156. Bazopoulou, D., and Tavernarakis, N. (2009). The NemaGENETAG initiative: Large scale transposon insertion gene-tagging in Caenorhabditis elegans. Genetica 137, 39–46. Belur, L. R., et al. (2003). Gene insertion and long-term expression in lung mediated by the Sleeping Beauty transposon system. Mol. Ther. 8, 501–507. Carlson, C. M., et al. (2003). Transposon mutagenesis of the mouse germline. Genetics 165, 243–256. Cary, L. C., et al. (1989). Transposon mutagenesis of baculoviruses: Analysis of Trichoplusia ni transposon IFP2 insertions within the FP-locus of nuclear polyhedrosis viruses. Virology 172, 156–169. Clark, K. J., et al. (2009). Passport, a native Tc1 transposon from flatfish, is functionally active in vertebrate cells. Nucleic Acids Res. 37, 1239–1247. Collier, L. S., et al. (2005). Cancer gene discovery in solid tumours using transposon-based somatic mutagenesis in the mouse. Nature 436, 272–276. Collier, L. S., et al. (2009). Whole-body sleeping beauty mutagenesis can cause penetrant leukemia/lymphoma and rare high-grade glioma without associated embryonic lethality. Cancer Res. 69, 8429–8437. Cooley, L., et al. (1988). Insertional mutagenesis of the Drosophila genome with single P elements. Science 239, 1121–1128.

68

Adam J. Dupuy

Cui, Z., et al. (2002). Structure-function analysis of the inverted terminal repeats of the sleeping beauty transposon. J. Mol. Biol. 318, 1221–1235. de Wit, T., et al. (2010). Tagged mutagenesis by efficient Minos-based germ line transposition. Mol. Cell. Biol. 30, 68–77. Ding, S., et al. (2005). Efficient transposition of the piggyBac (PB) transposon in mammalian cells and mice. Cell 122, 473–483. Dupuy, A. J., et al. (2001). Transposition and gene disruption in the male germline of the mouse. Genesis 30, 82–88. Dupuy, A. J., et al. (2002). Mammalian germ-line transgenesis by transposition. Proc. Natl. Acad. Sci. USA 99, 4495–4499. Dupuy, A. J., et al. (2005). Mammalian mutagenesis using a highly mobile somatic Sleeping Beauty transposon system. Nature 436, 221–226. Dupuy, A. J., et al. (2009). A modified sleeping beauty transposon system that can be used to model a wide variety of human cancers in mice. Cancer Res. 69, 8150–8156. Emelyanov, A., et al. (2006). Trans-kingdom transposition of the maize dissociation element. Genetics 174, 1095–1104. Fischer, S. E., et al. (2001). Regulated transposition of a fish transposon in the mouse germ line. Proc. Natl. Acad. Sci. USA 98, 6759–6764. Flajnik, M. F., and Kasahara, M. (2010). Origin and evolution of the adaptive immune system: Genetic events and selective pressures. Nat. Rev. Genet. 11, 47–59. Franz, G., and Savakis, C. (1991). Minos, a new transposable element from Drosophila hydei, is a member of the Tc1-like family of transposons. Nucleic Acids Res. 19, 6646. Geurts, A. M., et al. (2003). Gene transfer into genomes of human cells by the sleeping beauty transposon system. Mol. Ther. 8, 108–117. Geurts, A. M., et al. (2006). Gene mutations and genomic rearrangements in the mouse as a result of transposon mobilization from chromosomal concatemers. PLoS Genet. 2, e156. Gordon, J. W., et al. (1980). Genetic transformation of mouse embryos by microinjection of purified DNA. Proc. Natl. Acad. Sci. USA 77, 7380–7384. Hamlet, M. R., et al. (2006). Tol2 transposon-mediated transgenesis in Xenopus tropicalis. Genesis 44, 438–445. Horie, K., et al. (2001). Efficient chromosomal transposition of a Tc1/mariner-like transposon Sleeping Beauty in mice. Proc. Natl. Acad. Sci. USA 98, 9191–9196. Horie, K., et al. (2003). Characterization of Sleeping Beauty transposition and its application to genetic screening in mice. Mol. Cell. Biol. 23, 9189–9207. Ivics, Z., et al. (1997). Molecular reconstruction of Sleeping Beauty, a Tc1-like transposon from fish, and its transposition in human cells. Cell 91, 501–510. Izsvak, Z., et al. (2004). Healing the wounds inflicted by sleeping beauty transposition by double-strand break repair in mammalian somatic cells. Mol. Cell 13, 279–290. Kaji, K., et al. (2009). Virus-free induction of pluripotency and subsequent excision of reprogramming factors. Nature 458, 771–775. Kawakami, K., and Noda, T. (2004). Transposition of the Tol2 element, an Ac-like element from the Japanese medaka fish Oryzias latipes, in mouse embryonic stem cells. Genetics 166, 895–899. Kawakami, K., et al. (1998). Excision of the tol2 transposable element of the medaka fish, Oryzias latipes, in zebrafish, Danio rerio. Gene 225, 17–22. Kawakami, K., et al. (2000). Identification of a functional transposase of the Tol2 element, an Ac-like element from the Japanese medaka fish, and its transposition in the zebrafish germ lineage. Proc. Natl. Acad. Sci. USA 97, 11403–11408. Kawakami, K., et al. (2004). A transposon-mediated gene trap approach identifies developmentally regulated genes in zebrafish. Dev. Cell 7, 133–144. Keng, V. W., et al. (2005). Region-specific saturation germline mutagenesis in mice using the Sleeping Beauty transposon system. Nat. Methods 2, 763–769.

Transposons in Mouse Genetics

69

Keng, V. W., et al. (2009a). Efficient transposition of Tol2 in the mouse germline. Genetics 183, 1565–1573. Keng, V. W., et al. (2009b). A conditional transposon-based insertional mutagenesis screen for genes associated with mouse hepatocellular carcinoma. Nat. Biotechnol. 27, 264–274. Kondrychyn, I., et al. (2009). Genome-wide analysis of Tol2 transposon reintegration in zebrafish. BMC Genomics 10, 418. Liang, Q., et al. (2009). Chromosomal mobilization and reintegration of Sleeping Beauty and PiggyBac transposons. Genesis 47, 404–408. Liu, L., et al. (2004). Endothelial targeting of the Sleeping Beauty transposon within lung. Mol. Ther. 10, 97–105. Liu, L., et al. (2006). Sustained FVIII expression and phenotypic correction of hemophilia A in neonatal mice using an endothelial-targeted sleeping beauty transposon. Mol. Ther. 13, 1006–1015. Lobo, N., et al. (1999). Transposition of the piggyBac element in embryos of Drosophila melanogaster, Aedes aegypti and Trichoplusia ni. Mol. Gen. Genet. 261, 803–810. Luo, G., et al. (1998). Chromosomal transposition of a Tc1/mariner-like element in mouse embryonic stem cells. Proc. Natl. Acad. Sci. USA 95, 10769–10773. Mates, L., et al. (2009). Molecular evolution of a novel hyperactive Sleeping Beauty transposase enables robust stable gene transfer in vertebrates. Nat. Genet. 41, 753–761. McClintock, B. (1950). The origin and behavior of mutable loci in maize. Proc. Natl. Acad. Sci. USA 36, 344–355. Miskey, C., et al. (2003). The Frog Prince: A reconstructed transposon from Rana pipiens with high transpositional activity in vertebrate cells. Nucleic Acids Res. 31, 6873–6881. Miskey, C., et al. (2007). The ancient mariner sails again: Transposition of the human Hsmar1 element by a reconstructed transposase and activities of the SETMAR protein on transposon ends. Mol. Cell. Biol. 27, 4589–4600. Montini, E., et al. (2002). In vivo correction of murine tyrosinemia type I by DNA-mediated transposition. Mol. Ther. 6, 759–769. Ostertag, E. M., and Kazazian, H. H., Jr. (2001). Biology of mammalian L1 retrotransposons. Annu. Rev. Genet. 35, 501–538. Ostertag, E. M., et al. (2002). A mouse model of human L1 retrotransposition. Nat. Genet. 32, 655–660. Roberts, A. P., et al. (2008). Revised nomenclature for transposable genetic elements. Plasmid 60, 167–173. Ross-Macdonald, P., et al. (1999). Large-scale analysis of the yeast genome by transposon tagging and gene disruption. Nature 402, 413–418. Sato, Y., et al. (2007). Stable integration and conditional expression of electroporated transgenes in chicken embryos. Dev. Biol. 305, 616–624. Searles, L. L., et al. (1982). Molecular cloning of sequences from a Drosophila RNA polymerase II locus by P element transposon tagging. Cell 31, 585–592. Seberg, O., and Petersen, G. (2009). A unified classification system for eukaryotic transposable elements should reflect their phylogeny. Nat. Rev. Genet. 10, 276. Shinohara, E. T., et al. (2007). Active integration: New strategies for transgenesis. Transgenic Res. 16, 333–339. Siguier, P., et al. (2006). Insertion sequences in prokaryotic genomes. Curr. Opin. Microbiol. 9, 526–531. Sinzelle, L., et al. (2008). Transposition of a reconstructed Harbinger element in human cells and functional homology with two transposon-derived cellular genes. Proc. Natl. Acad. Sci. USA 105, 4715–4720. Starr, T. K., et al. (2009). A transposon-based genetic screen in mice identifies genes altered in colorectal cancer. Science 323, 1747–1750.

70

Adam J. Dupuy

Suster, M. L., et al. (2009). Transposon-mediated BAC transgenesis in zebrafish and mice. BMC Genomics 10, 477. Takahashi, K., and Yamanaka, S. (2006). Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663–676. Urasaki, A., et al. (2008). Transposition of the vertebrate Tol2 transposable element in Drosophila melanogaster. Gene 425, 64–68. Uren, A. G., et al. (2005). Retroviral insertional mutagenesis: Past, present and future. Oncogene 24, 7656–7672. Wang, W., et al. (2008). Chromosomal transposition of PiggyBac in mouse embryonic stem cells. Proc. Natl. Acad. Sci. USA 105, 9290–9295. Wang, W., et al. (2009). A piggyBac transposon-based genome-wide library of insertionally mutated Blm-deficient murine ES cells. Genome Res. 19, 667–673. Woltjen, K., et al. (2009). piggyBac transposition reprograms fibroblasts to induced pluripotent stem cells. Nature 458, 766–770. Wu, S. C., et al. (2006). piggyBac is a flexible and highly active transposon as compared to sleeping beauty, Tol2, and Mos1 in mammalian cells. Proc. Natl. Acad. Sci. USA 103, 15008–15013. Wu, S., et al. (2007). Toward simpler and faster genome-wide mutagenesis in mice. Nat. Genet. 39, 922–930. Yant, S. R., et al. (2000). Somatic integration and long-term transgene expression in normal and haemophilic mice using a DNA transposon system. Nat. Genet. 25, 35–41. Yant, S. R., et al. (2005). High-resolution genome-wide mapping of transposon integration in mammals. Mol. Cell. Biol. 25, 2085–2094. Yu, J., et al. (2007). Induced pluripotent stem cell lines derived from human somatic cells. Science 318, 1917–1920. Yusa, K., et al. (2009). Generation of transgene-free induced pluripotent mouse stem cells by the piggyBac transposon. Nat. Methods 6, 363–369. Zagoraiou, L., et al. (2001). In vivo transposition of Minos, a Drosophila mobile element, in mammalian tissues. Proc. Natl. Acad. Sci. USA 98, 11474–11478. Zayed, H., et al. (2003). The DNA-bending protein HMGB1 is a cellular cofactor of Sleeping Beauty transposition. Nucleic Acids Res. 31, 2313–2322.

C H A P T E R

F I V E

Functional Genomics in the Mouse using the Sleeping Beauty Transposon System Kyoji Horie,* Chikara Kokubu,† and Junji Takeda† Contents 1. Introduction 2. Genome-Wide Germline Mutagenesis with the SB Transgenic Approach 2.1. Distribution of SB transposition sites 2.2. Overview of the mutagenesis scheme 2.3. Vector structures and gene trap scheme 2.4. Consideration for vector construction 2.5. Generation and breeding of transgenic lines 2.6. Identification of transposition sites by linker ligation-mediated PCR 3. Region-Specific Chromosome Engineering with the SB Knock-in Approach 3.1. Exploitation of SB local hopping 3.2. Vector design 3.3. Cell culture and gene targeting 3.4. In vitro mobilization of transposons in ES cells 3.5. Determination of transposon insertion sites 3.6. Chromosomal engineering by site-specific recombination 3.7. Embryo production by tetraploid complementation 4. Concluding Remarks Acknowledgments References

72 74 74 74 76 76 77 78 82 82 82 85 85 86 86 87 87 88 88

Abstract The resurrection of the Sleeping Beauty (SB) transposon from molecularly extinct salmonid transposons at the end of last century opened the door for mouse geneticists to develop various transposon-based genetic tool kits, which * Department of Social and Environmental Medicine, Graduate School of Medicine, Osaka University, Suita, Osaka, Japan Center for Advanced Science and Innovation, Osaka University, Suita, Osaka, Japan

{

Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77005-1

#

2010 Elsevier Inc. All rights reserved.

71

72

Kyoji Horie et al.

had already been proven instrumental in Drosophila and other invertebrate model organisms. Since then, transposon technologies have been successfully applied to many aspects of functional genomics, in combination with various well-established tools of mouse genetics including transgenesis and gene targeting. In the SB system, a substantial fraction of the transposition events occurs on the same chromosome, predominantly within 3–4 megabases, while the remainder occurs between different chromosomes in a genome-wide manner. By taking advantage of the two types of transposition, we have developed applications of the SB system for genome-wide mutagenesis as well as regionspecific functional analysis of the mouse genome.

1. Introduction Transposable elements (TEs) are intrinsic components of most genomes and are capable of jumping to different locations within the genome of a single cell. In mammals, including mice and humans, TEs make up nearly half of the genome (Lander et al., 2001; Waterston et al., 2002) and are believed to have played pivotal roles in shaping genome evolution, although most of the TEs are currently inactive due to accumulation of mutations. Based on the mode of transposition, TEs can be divided into two classes: retrotransposons, which transpose via an RNA intermediate in a ‘‘copyand-paste’’ manner, and DNA transposons, which transpose via a DNA intermediate in a ‘‘cut-and-paste’’ manner with a few exceptions such as the recently characterized atypical copy-and-paste transposons Helitrons and Mavericks (Feschotte and Pritham, 2007). Retrotransposons rapidly increase in copy number following their replication cycle, while the cut-and-paste transposons move through a nonreplicative mechanism and occasionally fail to land in the genome. Nevertheless, many DNA transposons did not become extinct probably due to some as-yet-undefined mechanisms that increase their copy number; for example, transposition during host DNA replication, or the repair of the double-strand break left by excision of the element (Feschotte and Pritham, 2007). The DNA cut-and-paste transposons have been widely used as tools for genetic manipulation in invertebrate model organisms such as plants, Drosophila, and nematodes, but it was not until the molecular reconstruction of the salmonid transposon Sleeping Beauty (SB) in 1997 (Ivics et al., 1997) that the transposon technology was implemented into the geneticists’ toolbox for vertebrate model organisms. The SB transposon was the first transposon shown to have activity in mouse and human cells (Ivics et al., 1997), though the transposition frequency was relatively low (3.5
10 5 excisions/cell per generation) in mouse embryonic stem (ES) cells (Luo et al., 1998; Table 5.1). Following these pioneering experiments, the SB

73

The Sleeping Beauty Transposon System

Table 5.1 Comparison of transgenic versus knock-in approaches for the introduction of a transposon donor site into the mouse chromosome Transgenic

Knock-in

Host DNA delivery method Mode of integration

Fertilized oocytes Embryonic stem cells Pronuclear injection Electroporation Random Targeted by homologous recombination Transposon copy number Multiple Single Total transposition frequency High Low Transposition-associated Possible Unlikelya undesirable rearrangement

a

In the knock-in approach, ES cells that contain, if any, undesirable rearrangements at the transposon donor site could be eliminated by a built-in selection step for correct rejoining. See text for details.

transposon has proven to be active in the mouse male germ line in vivo with an unexpectedly high efficiency (0.2–2 transpositions per gamete), demonstrating its potential for use in mutagenesis applications (Dupuy et al., 2001; Fischer et al., 2001; Horie et al., 2001). To date, other vertebrate transposon systems, such as Tol2, Minos, Frog Prince, and piggyBac, have also been developed as reviewed elsewhere (Ivics et al., 2009). This chapter will focus on the use of the SB transposon system for mouse functional genomics. The SB system consists of two functional components: a transposon DNA element and the transposase enzyme (Ivics et al., 1997). In the mouse, there are two types of approaches to introduce the transposon sequence into the mouse genome, a transgenic approach through pronuclear injection of fertilized oocytes and a knock-in approach through homologous recombination in ES cells (Table 5.1). In the transgenic approach, pronuclear injection usually leads to a tandem integration of multiple copies of the transposon transgene, and the integration site could serve as a donor site from which mutagenic transposons jump to new positions. The relatively high transposition frequency from the transposon transgene concatemer allows efficient germ line mutagenesis (Geurts et al., 2006; Keng et al., 2005) as well as somatic mutagenesis (Collier et al., 2005; Dupuy et al., 2005) in vivo simply by animal breeding. On the other hand, the knock-in approach usually introduces a single-copy transposon as a donor site. Although transposition from a single-copy donor does not occur frequently, it allows more sophisticated engineering of specific chromosomal loci (Kokubu et al., 2009). Among the wide variety of applications, here we discuss two SB protocols based on each of the abovementioned genetic approaches: a genome-wide germline mutagenesis with the SB transgenic approach and region-specific chromosome engineering with the SB knock-in approach.

74

Kyoji Horie et al.

2. Genome-Wide Germline Mutagenesis with the SB Transgenic Approach 2.1. Distribution of SB transposition sites In order to harness the SB transposon as a tool for mutagenesis, it is important to be aware of the distribution pattern of the SB transposition. The SB transposon has preference to transpose locally: 60–80% of the transposition sites are located on the donor-site chromosome, especially 3–4 Mb around the donor site (Fig. 5.1). This local hopping feature is utilized for region-specific mutagenesis as described later in Section 3 (Kokubu et al., 2009). On the other hand, the remaining 20–40% are mapped on different chromosomes (Horie et al., 2003; Keng et al., 2005). This type of transposition appears fairly randomly distributed throughout the genome. Here we describe the application of this genome-wide transposition for functional analysis of the mouse genome.

2.2. Overview of the mutagenesis scheme The mutagenesis scheme is outlined in Fig. 5.2. Two types of transgenic lines are generated, one containing the SB transposon and the other expressing the SB transposase (Fig. 5.2A). They are intercrossed to generate doubly

Distance (bp)

Chromosome

3–4 Mb

Donor site

No. of insertions Local transposition 60–80%

Genome-wide transposition 20–40%

Figure 5.1 Distribution of SB transposition sites. Approximately 60–80% of transposition events are located on the chromosome bearing the donor site, especially within 3–4 Mb from the donor site, and the remaining 20–40% are distributed on other chromosomes.

A

Transposon vector SA IRES lacZ pA

P neo pA

P GFP SD

pBluescript IR/DR

IR/DR

Transposon Transposase vector Transposase

P

pA

B E1

E2

E3 SA IRES lacZ pA

P GFP SD

(A)n

(A)n

Promoter trap

Truncated protein

+

PolyA trap

b -Gal

GFP

Screening for mutant mice

Disruption of gene Expression profiling function with X-gal staining C GFP

TPase

GFP-negative

TPase

Germ line transposition

GFP

“Seed mouse” GFP-negative

Mutant

Wild-type

Wild-type

Transmission of germ line mutations to the next generation GFP-positive (gene trap event) GFP-negative Bright field

Fluorescent field

Figure 5.2 Mutagenesis scheme. (A) Vector structures. SA, splice acceptor; IRES, internal ribosome entry site; SD, splice donor; P, promoter; IR/DR, inverted repeat/direct repeat. (B) Gene trap scheme. E, exon. (C) Breeding scheme for production of mutant mice. Germline transposition is induced in the doubly transgenic ‘‘seed mice.’’ Fertilization of transposon-inserted mutant gametes with wild-type gametes results in the generation of mutant mice. Mice are expected to be GFP-positive when the transposon is inserted into genes according to the polyA-trap scheme shown in (B). TPase, SB transposase.

76

Kyoji Horie et al.

transgenic mice, in which the SB transposon is excised by the SB transposase and reinserted into other locations of the genome. Doubly transgenic mice, also called ‘‘seed mice’’ (Fig. 5.2C), are mated with wild-type mice. Transposition in germ cells of seed mice will generate heterozygous mutant mice in the progeny. We routinely use male seed mice because we can set up more breeding pairs compared with female seed mice. Heterozygous mice are mated with wild-type mice to establish each mutant mouse line. It is advisable to select mice that lost the SB transposase during breeding in order to avoid continuous germline transposition. Homozygous mice are obtained by intercross of heterozygous mice and are subjected to phenotypic analyses.

2.3. Vector structures and gene trap scheme To detect transposon insertion into the gene regions, we routinely incorporate the promoter-trap and polyA-trap units into the transposon vector. The promoter-trap unit consists of a splice acceptor (SA) and a reporter gene to monitor the expression of the mutated gene (Fig. 5.2B). We have successfully used the SA from the intron 2/exon 3 junction of the human BCL2 gene (Ishida and Leder, 1999), encephalomyocarditis virus internal ribosome entry site (IRES), the lacZ reporter gene, and polyadenylation signal (pA) from the rabbit b-globin gene. The polyA-trap unit usually contains a ubiquitously active promoter, a reporter gene, and a splice donor (SD). The reporter gene in the polyA-trap unit is expressed in case the vector is inserted into gene regions and the SD in the polyA unit is spliced to the SA site of the inserted genes (Fig. 5.2B). We have used the cytomegalovirus (CMV) enhancer/chicken b-actin chimeric (CAG) promoter (Niwa et al., 1991), the EGFP gene, and the SD from the exon 8/intron 8 junction of the mouse Hprt gene (Ishida and Leder, 1999). The EGFP reporter allows the identification of mutant mice noninvasively and in a high-throughput manner (Fig. 5.2B and C). The mRNA instability signal derived from the human granulocyte-macrophage colony-stimulating factor transcript is also incorporated downstream of the SD in order to minimize the effect of unspliced read-through transcript (Ishida and Leder, 1999). The promoter of the SB transposase should be strong in germ cell lineage. We have mainly used the CAG promoter in our study (Horie et al., 2001, 2003; Keng et al., 2005; Kitada et al., 2007), though other promoters, such as the protamine promoter, have also been reported (Fischer et al., 2001).

2.4. Consideration for vector construction Transgenic lines are generated by the standard method of pronuclear injection. The plasmid backbone sequence is usually removed prior to pronuclear injection, and SB transposase lines are generated in this scheme.

The Sleeping Beauty Transposon System

77

In contrast, the SB transposon lines are generated without removing the plasmid backbone (pBluescript and neo gene cassette; Fig. 5.2A). In this configuration, we observed that the transposon region is heavily methylated after integration into the mouse genome (Horie et al., 2001). Although the CAG promoter is a ubiquitously active promoter, the effect of methylation is usually, though not always, strong enough to silence the promoter activity and prevent EGFP expression. After transposition, however, the CAG promoter is activated at the reinsertion site because of the absence of the flanking plasmid backbone. Accordingly, the gene insertion events can be visualized as ‘‘green mice’’ (Fig. 5.2C). Our previous studies indicated that the DNA methylation and heterochromatinization of the transposon donor site substantially enhance SB transposition (Ikeda et al., 2007; Yusa et al., 2004). It should be noted that the enhancement of transposition by DNA methylation and heterochromatinization is not necessarily seen in other transposon systems (Wang et al., 2008).

2.5. Generation and breeding of transgenic lines 2.5.1. Screening for transposon lines In the mutagenesis scheme displayed in Fig. 5.2C, the transposon lines are GFP-negative prior to breeding with transposase lines. Although silencing of GFP expression by the plasmid backbone is quite efficient, it may not be perfect in some transposon lines and the GFP signal may be observed prior to transposition. Such transposon lines should be excluded before mating with the transposase lines. The GFP-negative mice should be selected using a fluorescent microscope under the same condition as those applied in screening for GFP-positive heterozygous mice, which is described later in Section 2.5.4. It is well known that the genomic structure could be extensively altered during vector integration in pronuclear injection. Such damage may bring about deleterious effects in homozygosity, which may complicate phenotypic analyses of mutant lines. Therefore, it is important to test whether homozygosity of the donor site exhibits any phenotype without any transposition. 2.5.2. Screening of SB transgenic lines Since the expression level of the SB transposase varies among transgenic lines, it is recommended to screen for appropriate transgenic lines with high transposition activities. Each SB transposase line is mated with a particular SB transposon line, and the frequency of transposon excision can be assayed by PCR in germ cells of doubly transgenic lines. It is also helpful to quantitate the expression levels of the SB transposase in germ cells by

78

Kyoji Horie et al.

standard methods such as immunohistochemistry, Northern blotting, or RT-PCR. Anti-SB transposase antibody is commercially available from R&D systems. 2.5.3. Maintenance of transgenic lines The transposon and transposase lines should be maintained independently without intercrossing. Once the SB transposase is introduced, the transposon concatemer at the donor site may be structurally altered or unpredicted mutations may be introduced by transposition. 2.5.4. Detection of GFP-positive mutant mice using fluorescent microscope The GFP-signals in gene-trapped mutant mice are generally weak. Therefore, screening for GFP-positive mice should be performed within a few days after birth before the appearance of the coat color. Fluorescent stereomicroscope has adequate detection sensitivity. We use the WILD M10 microscope (Leica) equipped with GFP-specific filters. In our previous study, approximately 10% of newborn mice from seed mice were GFP-positive (Horie et al., 2003). This frequency seems to vary among transposon lines. It is worthwhile to test several transposon lines and identify the highly active lines. Recently, new versions of SB transposons and SB transposases have been developed (Mates et al., 2009). Exploitation of these systems is likely to enhance the rate of mutagenesis in mice.

2.6. Identification of transposition sites by linker ligation-mediated PCR 2.6.1. Overview of the procedure To identify the mutated genes, we routinely use linker ligation-mediated PCR (LM-PCR) (Fig. 5.3). This method involves digestion of genomic DNAs with restriction enzymes, ligation of linker DNAs, and nested PCR for amplification of the transposon/genomic DNA junctions using transposon-specific and linker-specific primers. The PCR products are sequenced and mutated genes are identified by database search of the mouse genome. The LM-PCR is simple, straightforward, and widely used in other types of gene trap vectors (Horn et al., 2007). However, it should be kept in mind that the transposon concatemer at the donor site may hinder the amplification of transposition sites. Since many copies of transposon sequences exist at the donor site, they may be amplified much more efficiently than the transposition sites and virtually suppress their amplification. Cosegregation of the donor-site concatemer and transposition sites occurs frequently when the SB transposon is reinserted into the same chromosome, which is 60–80% as shown in Fig. 5.1. In contrast, half of the mice will lose the donor

79

The Sleeping Beauty Transposon System

Donor site Concatemer of transposon vectors X

X

Y X

X

Y X

Plasmid backbone

X

Y X

X

Y X

Transposon Transposition

Transposition site X

X

X

Genomic DNA

Transposon Digestion with restriction enzyme “X” X

X

Linker ligation

Primer

Primer 1st PCR

Primer

Primer 2nd PCR

Sequencing Mapping on the mouse genome

Figure 5.3 Linker ligation-mediated PCR. Following restriction digestion and linker ligation, transposition sites are amplified by nested PCR with transposon-specific and linker-specific primers. Note that the plasmid backbone sequence may also be amplified in case the donor transposon concatemer is not segregated out in the mutant mice. Digestion of linker-ligation product with the enzyme ‘Y’ located right next to the transposon sequence reduces this amplification.

concatemer when reinsertion occurs into other chromosomes. For ‘‘genome-wide’’ mutagenesis, one may prescreen for the absence of the donor concatemer by PCR-amplification of the donor-site-specific sequences (e.g., pBluescript, neo; Fig. 5.2A).

80

Kyoji Horie et al.

We provide below details of the splinkerette-PCR protocol developed by Devon et al. (1995) with some modifications.

2.6.2. Protocol of LM-PCR 1. Digest genomic DNA using an appropriate four-base restriction enzyme (enzyme ‘‘X’’ in Fig. 5.3). The restriction site must be absent between the transposon-specific primer and the terminal end of the SB transposon. In the case of our transposon vector reported previously (Keng et al., 2005; Kokubu et al., 2009), we used AluI, MboI, and HaeIIIl restriction enzymes. 2. Heat-inactivate the restriction enzymes and ligate the splinkerette linker (Table 5.2) to the cleavage ends. Use Spl-blunt for AluI and HaeIII, and Spl-sau for MboI. 3. Purify ligation products by standard methods such as QIAquick PCR purification kit (Qiagen) and resuspend in water. 4. Digest the ligation products with an appropriate restriction enzyme (indicated by ‘‘Y’’ in Fig. 5.3) that cuts the transposon vector backbone sequence just outside of the transposon sequence but is absent between the transposon-specific primer and the terminal end of the transposon. A rare cutter enzyme is preferable in order to avoid digestion at the transposition sites. In our SB transposon vector, KpnI was appropriate for that purpose. Theoretically, this step prevents amplification of the transposon vector donor concatemer site. However, it should be kept in mind that even a trace of the undigested DNA could result in substantial amplification of the donor concatemer and reduce the amplification efficiency of transposition sites. As described in the previous section, prescreening for the absence of the donor site helps avoid this problem. 5. Purify DNA using standard protocols such as QIAqucik PCR purification kit (Qiagen) and resuspend in 50 l of water. 6. Conduct nested PCR using transposon-specific primers and linkerspecific primers (Table 5.2). Since several versions of the SB transposon vectors have been developed, one needs to carefully choose appropriate transposon-specific primers. In Table 5.2, we present two versions of SB transposons pT (Ivics et al., 1997) and pT2 (Cui et al., 2002), and the corresponding primers. 7. Check for PCR products by agarose gel electrophoresis. 8. Purify the PCR products using standard protocols such as QIAqucik PCR purification kit and proceed to sequencing. We routinely use transposon-specific second round PCR primer. In case more than one band is observed due to multiple transpositions from the donor concatemer, purify each band using standard protocol such as QIAqucik gel extraction kit (Qiagen) prior to sequencing.

The Sleeping Beauty Transposon System

81

Table 5.2 Oligonucleotide sequence and reaction condition for linker ligation-mediated PCR

Linker sequence Spl-top SplB-blunta SplB-saub

50 -CGAATCGTAACCGTTCGTACGA GAATTCGTACGAGAATCGCTGTC CTCTCCAACGAGCCAAGG-30 50 -CCTTGGCTCGTTTTTTTTTGCAAAAA-30 50 -GATCCCTTGGCTCGTTTTTTTT TGCAAAAA-30

Nested PCR primersc First round PCR Linker-specific primer Spl-P1: 50 -CGAATCGTAACCGTTCG TACGAGAA-30 Transposon-specific primers pT, T/DR2 50 -CTGGAATTGTGATACAGTGAAT TATAAGTG-30 0 pT2, T/KBA 5 -CTAACTGACCTAAGACAGGGAA TTTTTAC-30 Second round PCR Linker-specific primers Spl-P2 50 -TCGTACGAGAATCGCTGTCCTCTCC-30 Transposon-specific primers pT, T/BAL 50 -CTTGTGTCATGCACAAAGTAGATGTCC-30 pT2, T/KB1 50 -ATTTTTACTAGGATTAAATGT CAGGAATTG-30 PCR conditions (Qiagen HotStarTaq) Template DNA 1 l Buffer (10
) 5 l dNTP (10 mM) 1 l Transposon-specific 1 l primer (10 M) Linker-specific primer 1 l (10 M) HotStarTaq 0.25 l Water 40.75 l Total 50 l 95  C
15 min 1 cycle 94  C
1 min, 30 cycles 55  C
1 min, 72  C
1 min 1 cycle 72  C
7 min 25  C Hold a b c

To generate blunt-end linker (Spl-blunt), anneal Spl-top and SplB-blunt. To generate protruding-end linker compatible with MboI cleavage site (Spl-sau), anneal Spl-top and SplB-sau. pT and pT2 are versions of the SB transposon described in Sections 2 and 3 of the text, respectively.

82

Kyoji Horie et al.

9. Map the transposition sites using mouse genome databases (e.g., UCSC genome browser at http://genome.ucsc.edu/). Beware of possible generation of artificial bands during the LM-PCR procedure such as chimeric DNAs from different chromosomal loci. Accordingly, it is important to verify the junction of the transposon end and the genomic sequence.

3. Region-Specific Chromosome Engineering with the SB Knock-in Approach 3.1. Exploitation of SB local hopping As mentioned in Section 2.1, the SB transposon has a strong tendency of ‘‘local hopping’’ when they move from chromosome to chromosome in the mouse (Fig. 5.1). Since the mouse offers the advantage of use of ES cellbased genome manipulation techniques, the combination of the local hopping transposon with standard knock-in technology could allow functional surveillance of defined genomic regions of interest. The range of local hopping, which is usually from a few kilobases to a few megabases, is nicely compatible with a typical size of chromosomal gene clusters and regulatory regions. This motivated us to exploit the SB transposon for mapping longrange cis-regulatory genomic elements, the effective identification of which has been greatly hampered by the paucity of appropriate genetic tools. The additional use of enhancer detection (so-called enhancer trapping) technology allowed us to develop a transposon-based chromosomal engineering method termed the Local Hopping Enhancer Detector (LHED) system (Kokubu et al., 2009).

3.2. Vector design 3.2.1. Homology arm for targeted recombination For efficient targeted insertion of the SB transposon into the genomic regions of choice, the LHED system adopts insertion-type homologous recombination using a single homology arm. One can choose any genomic region encompassing either a BamHI or a HindIII restriction site for the homology arm, and insert it into the generic targeting vector pLHED (Fig. 5.4), which contains neither BamHI nor HindIII sites within the vector itself. When electroporated into mouse ES cells, the unique BamHI (or HindIII) site on the arm serves as the position of linearization of the entire vector. Moreover, preintroduction of a short sequence gap at the

83

The Sleeping Beauty Transposon System

pLHED targeting vector BamHI

loxP loxP PGK neo

FRT

loxP Hsp-lacZ

Puro

PGK

FRT

Transposon

Insertion-type homologous recombination loxP loxP PGK neo

loxP

Hsp-lacZ

Puro

FRT

Transposon

PGK

FRT

Local hopping

Sleeping beauty transposase

loxP

loxP loxP PGK neo

Hsp-lacZ

Puro PGK

FRT

FRT

Transposon

Cre recombinase loxP

loxP

Cre-induced deletion

Hsp-lacZ

FRT

Figure 5.4 The Local Hopping Enhancer Detector (LHED) system. The pLHED targeting vector is knocked into a genomic locus of interest by insertion-type homologous recombination, and correctly recombined clones are selected using G418 resistance. Subsequent in vitro transfection with SB transposase mobilizes the SB transposon predominantly to closely linked sites, accompanied by reactivation of the puromycinresistance gene for positive selection. When needed, additional in vitro transfection with Cre recombinase can introduce a chromosomal deletion between the fixed and transposed loxP sites. Figure taken from Kokubu et al (2009).

linearization site could improve the recombination efficiency by exploiting the cellular gap repair pathways (Valancius and Smithies, 1991). To provide an actual example, let us assume a particular genomic sequence with about 7-kb length. Excision of a short fragment (a few hundred base pairs) from the middle portion of the 7-kb sequence by, for example, BamHI (50 -G/ GATCC-30 ) and KpnI (50 -GGTAC/C-30 ) double digestion, followed by T4 DNA polymerase treatment (producing blunt ends, 50 -GGATC/ and /C-30 ) and self-ligation, results in the regeneration of a unique BamHI (50 -G/GATCC-30 ) site and the establishment of a sequence gap at the junction.

84

Kyoji Horie et al.

3.2.2. Selection of transposition events The relatively high frequency of SB transposition in transgenic animal’s germ cells is due, at least in some part, to the multiplicity of transposon copy number at the donor sites, which is a common feature of chromosomal integration of pronuclear-injected transgenes (Table 5.1). The LHED system, in contrast, utilizes a chromosomally integrated single-copy transposon as the donor and does not offer a high degree of transposition frequency as the multicopy-donor system. Therefore, similar to the previous study (Luo et al., 1998), the LHED system incorporates a selectable marker pac (puromycin resistance, also called puro) gene separated from the Pgk1 (phosphoglycerate kinase-1, also called PGK) promoter by insertion of a TE, of which excision confers puromycin resistance on the host ES cells by recovering the Pgk1–pac junction. This selection scheme also guarantees that the selected ES cells are free of such undesirable rearrangements around the transposon donor site as has been noted previously in the multicopydonor SB system (Geurts et al., 2006; Table 5.1). 3.2.3. Enhancer detection There has been a good match between enhancer detection and transposon technology since the P element–lacZ fusion gene was used, for the first time, to detect genomic regulatory elements in Drosophila (O’Kane and Gehring, 1987). In the early stages, however, the application of enhancer detection placed more emphasis on the isolation of novel ‘genes’ associated with tissue-specific patterns of the reporter expression (Bellen, 1999). For this purpose, however, promoter- or gene-traps would be more straightforward and popular methods, especially in the mouse. In fact, it was not until the completion of the whole genome sequencing in this century that the enhancer detection (or enhancer trap) method could provide a full scope of its ability to explore the vertebrate genomic ‘enhancers’ themselves with reference to the precise information of the transposon insertion sites. To resuscitate the potential of enhancer trap technology, we decided to equip the local hopping transposon with an enhancer detector cassette, consisting of the lacZ reporter gene fused to the mouse Hspa1b (heat shock protein 68, also known as Hsp68) minimal promoter (Kothary et al., 1989). 3.2.4. Site-specific recombination for chromosomal engineering In addition to enhancer detection, the LHED system provides a powerful tool for region-specific chromosome engineering to introduce nested deletions and inversions that anchor one end at the knock-in site of the pLHED targeting vector. The vector contains three loxP sites, one within the SB transposon and two flanking PGKneo cassette outside the TE. After targeted integration of the entire vector, the SB element containing a loxP site is mobilized predominantly to closely linked sites on the same (cis)

The Sleeping Beauty Transposon System

85

chromosome in a local hopping manner. The genomic DNA segment between the fixed donor loxP site and a transposed loxP site can either be excised or inverted by Cre/loxP-mediated recombination, if loxP sites are in the same or opposite orientations, respectively (Schnutgen et al., 2006). The PGKneo cassette is flanked by loxP sites on both sides so that it can be removed by site-specific recombination with another loxP site transposed to either side. Furthermore, the vector is designed so that, if necessary, the vector-derived flanking sequences can be excised from the knock-in locus by Flp/FRT recombination.

3.3. Cell culture and gene targeting The LHED system involves multiple invasive treatments of ES cells, including electroporation and lipofection, which may affect the quality of cellular pluripotency of ES cells. To overcome this pitfall, we use the hybrid ES cell line V6.5 ((C57BL/6
129S4Sv/Jae) F1), which is considered to be extremely robust, especially for use in the subsequent tetraploid complementation assay (Eggan et al., 2001). The targeted genomic integration of the pLHED vector is carried out in the ES cells by standard electroporation and G418 (150 g/ml) selection procedures. As described above, when electroporated, the pLHED targeting vector is linearized by BamHI (or HindIII) at a prepositioned short sequence gap (a few hundred base pairs) in the middle of the homology. Correctly targeted ES cell clones are screened by PCR using primers that amplify the region between the gapfilling sequence and the vector-specific sequence (Fig. 5.4), and further confirmed by Southern DNA blotting. In our experience, more than 40% targeting efficiency can be expected.

3.4. In vitro mobilization of transposons in ES cells After targeted integration of the pLHED vector into the ES cell genome, the SB transposon can be mobilized by in vitro transfection of the knock-in ES cells with a SB transposase expression vector. For the transfection step, we use TransFastTM Transfection Reagent (Promega, Madison, WI) according to the instructions provided by the manufacturer, with some modifications. Here we show the outline: 1. One day before transfection, suspend the TransFastTM Reagent and store at  20  C. 2. On the day of transfection, thaw and vortex the TransFastTM Reagent suspension, and further heat it to 37  C for 30 min to fully suspend any residual lipid. 3. Trypsinize a subconfluent culture of pLHED knock-in ES cells and suspend the cells to a density of 1
106 cells/ml in ES cell medium.

86

Kyoji Horie et al.

4. Combine 10 l of 1 g/l SB transposase expression (pCMV-SB11) (Geurts et al., 2003) plasmid DNA with 1 ml of ES cell medium and vortex. Then, add 30 l of TransFastTM Reagent, vortex immediately, and incubate for 15 min at room temperature. 5. Combine the whole incubation mixture with 1 ml of the pLHED knock-in ES cell suspension (1
106 cells) and return to the incubator for 1 h. 6. At the end of the incubation, add 8 ml of prewarmed ES cell medium and plate onto 10-cm feeder plate, followed by replacement with fresh medium after 5 h. Two days later, start drug selection with puromycin (1 g/ml) and replace the puromycin-containing medium daily for 5 days. 7. After 5-days selection, pick individual colonies, place them into 96-well feeder plates, and allow growth to confluence in a puromycin-containing medium. When confluent, using multichannel pipettes, transfer one-half of the cells from each well into 96-well storage plates (Thermo Fisher Scientific, Waltham, MA) for freezing, and transfer the remainder into new 96-well feeder plates for further expansion. 8. After reaching confluence, isolate the genomic DNA from each well by using the DNeasy 96 Blood & Tissue kit (Qiagen) using the protocol provided by the manufacturer. In our previous study, 100 puromycin-resistant colonies were isolated from 1
106 of LHED knock-in ES cells transfected with the pCMV-SB11 plasmid, indicating that the frequency of transposon excision was approximately 1
10 4 per transfected cell (Kokubu et al., 2009).

3.5. Determination of transposon insertion sites We usually determine the insertion sites of LHED transposon by linker LM-PCR, essentially according to the protocol described in Section 2.6.2. Since the LHED system relies on a single-copy transposon, the digestion step with the restriction enzyme ‘‘Y’’ can be omitted (Fig. 5.3). It should be noted that the LHED system utilizes SB transposon pT2, which is a different version from the pT described in Section 2. Table 5.2 lists the specific primers for pT2.

3.6. Chromosomal engineering by site-specific recombination As mentioned, the LHED system provides a unique opportunity for chromosome engineering in mouse ES cells, which is mediated by site-specific Cre/loxP- or Flp/FRT-recombination technology. The recombination can be induced in vitro by transient transfection of the Cre or Flp recombinase expression plasmid into LHED-inserted ES cells:

The Sleeping Beauty Transposon System

87

1. The first three steps are the same as steps 1–3 listed in Section 3.4. 2. Combine 2 l of 1 g/l Cre expression (pBS185, Invitrogen) or Flpe expression (pCAGGS-Flpe-IRES-puro) (Schaft et al., 2001) plasmid DNA with 192 l of ES cell medium, and vortex. Then, add 6 l of TransFastTM Reagent, vortex immediately, and incubate for 15 min at room temperature. 3. Combine the incubation mixture with 200 l of the pLHED knock-in ES cell suspension (2
105 cells) and return to the incubator. 4. At the end of the 1-h incubation, add 1.6 ml of prewarmed ES cell medium and transfer into a 24-well feeder plate, followed by replacement with fresh medium after 5 h. 5. Two days later, passage the cells onto a 10-cm feeder plate at a low density (1
103 cells/plate) and allow them to grow to colonies. 6. After 7–8 days of culture, isolate individual colonies and transfer to 96well feeder plates. The site-specific recombination-positive clones are screened by PCR, and then confirmed by Southern blotting. It should be noted that some colonies often contain a mixture of recombinant and nonrecombinant ES cells, which can be ruled out by PCR using the floxed region-specific primers. The efficiency of site-specific recombination tends to decrease with increasing genomic distance. In our experience, when two loxPs were 300–400 kb apart, it was necessary to isolate 200–300 clones to obtain 3–5 independent clones with correct recombination.

3.7. Embryo production by tetraploid complementation The LHED-engineered ES cells can mature into embryonic and postnatal mice, in which one can assess in vivo the phenotypic consequences of chromosome engineering, including reporter gene insertion and nested deletion formation. For cis-regulatory analysis, it suffices to generate heterozygous embryos carrying an LHED insertion and no breeding to homozygosity is required. We therefore inject the ES cells into tetraploid blastocysts to obtain almost entirely ES cell-derived embryos (Nagy et al., 1993). The process of tetraploid complementation was described elsewhere (Kokubu et al., 2009; Okada et al., 2007).

4. Concluding Remarks The SB transposon is one of the most well-characterized transposon tool kits used for genetic engineering of mice and other vertebrates. The two applications described in this chapter, genome-wide germline mutagenesis and region-specific chromosome engineering, exemplify the

88

Kyoji Horie et al.

feasibility and performance of this transposon-based approach for functional genomics. In addition to these germlines and ES cell-based mutagenesis, somatic cell mutagenesis within individual mice is also a recently developed and promising application of transposon, especially for cancer gene discovery as reviewed recently (Largaespada, 2009). Together, the successful implementation of transposon technologies will take mouse genetics to the next level, allowing a more efficient and flexible manipulation of the genome.

ACKNOWLEDGMENTS The authors thank the many past and present members of the Takeda laboratory who have contributed to these methods over the years, and also thank the support of the New Energy and Industrial Technology Development Organization of Japan; RIKEN, the Institute of Physical and Chemical Research; and a grant-in-aid for Science Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

REFERENCES Bellen, H. J. (1999). Ten years of enhancer detection: Lessons from the fly. Plant Cell 11, 2271–2281. Collier, L. S., et al. (2005). Cancer gene discovery in solid tumours using transposon-based somatic mutagenesis in the mouse. Nature 436, 272–276. Cui, Z., et al. (2002). Structure-function analysis of the inverted terminal repeats of the sleeping beauty transposon. J. Mol. Biol. 318, 1221–1235. Devon, R. S., et al. (1995). Splinkerettes-improved vectorettes for greater efficiency in PCR walking. Nucleic Acids Res. 23, 1644–1645. Dupuy, A. J., et al. (2001). Transposition and gene disruption in the male germline of the mouse. Genesis 30, 82–88. Dupuy, A. J., et al. (2005). Mammalian mutagenesis using a highly mobile somatic Sleeping Beauty transposon system. Nature 436, 221–226. Eggan, K., et al. (2001). Hybrid vigor, fetal overgrowth, and viability of mice derived by nuclear cloning and tetraploid embryo complementation. Proc. Natl. Acad. Sci. USA 98, 6209–6214. Feschotte, C., and Pritham, E. J. (2007). DNA transposons and the evolution of eukaryotic genomes. Annu. Rev. Genet. 41, 331–368. Fischer, S. E., et al. (2001). Regulated transposition of a fish transposon in the mouse germ line. Proc. Natl. Acad. Sci. USA 98, 6759–6764. Geurts, A. M., et al. (2003). Gene transfer into genomes of human cells by the sleeping beauty transposon system. Mol. Ther. 8, 108–117. Geurts, A. M., et al. (2006). Gene mutations and genomic rearrangements in the mouse as a result of transposon mobilization from chromosomal concatemers. PLoS Genet. 2, e156. Horie, K., et al. (2001). Efficient chromosomal transposition of a Tc1/mariner-like transposon Sleeping Beauty in mice. Proc. Natl. Acad. Sci. USA 98, 9191–9196. Horie, K., et al. (2003). Characterization of Sleeping Beauty transposition and its application to genetic screening in mice. Mol. Cell. Biol. 23, 9189–9207.

The Sleeping Beauty Transposon System

89

Horn, C., et al. (2007). Splinkerette PCR for more efficient characterization of gene trap events. Nat. Genet. 39, 933–934. Ikeda, R., et al. (2007). Sleeping beauty transposase has an affinity for heterochromatin conformation. Mol. Cell. Biol. 27, 1665–1676. Ishida, Y., and Leder, P. (1999). RET: A poly A-trap retrovirus vector for reversible disruption and expression monitoring of genes in living cells. Nucleic Acids Res. 27, e35. Ivics, Z., et al. (1997). Molecular reconstruction of Sleeping Beauty, a Tc1-like transposon from fish, and its transposition in human cells. Cell 91, 501–510. Ivics, Z., et al. (2009). Transposon-mediated genome manipulation in vertebrates. Nat. Methods 6, 415–422. Keng, V. W., et al. (2005). Region-specific saturation germline mutagenesis in mice using the Sleeping Beauty transposon system. Nat. Methods 2, 763–769. Kitada, K., et al. (2007). Transposon-tagged mutagenesis in the rat. Nat. Methods 4, 131–133. Kokubu, C., et al. (2009). A transposon-based chromosomal engineering method to survey a large cis-regulatory landscape in mice. Nat. Genet. 41, 946–952. Kothary, R., et al. (1989). Inducible expression of an hsp68-lacZ hybrid gene in transgenic mice. Development 105, 707–714. Lander, E. S., et al. (2001). Initial sequencing and analysis of the human genome. Nature 409, 860–921. Largaespada, D. A. (2009). Transposon-mediated mutagenesis of somatic cells in the mouse for cancer gene identification. Methods 49, 282–286. Luo, G., et al. (1998). Chromosomal transposition of a Tc1/mariner-like element in mouse embryonic stem cells. Proc. Natl. Acad. Sci. USA 95, 10769–10773. Mates, L., et al. (2009). Molecular evolution of a novel hyperactive Sleeping Beauty transposase enables robust stable gene transfer in vertebrates. Nat. Genet. 41, 753–761. Nagy, A., et al. (1993). Derivation of completely cell culture-derived mice from earlypassage embryonic stem cells. Proc. Natl. Acad. Sci. USA 90, 8424–8428. Niwa, H., et al. (1991). Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene 108, 193–199. Okada, Y., et al. (2007). Complementation of placental defects and embryonic lethality by trophoblast-specific lentiviral gene transfer. Nat. Biotechnol. 25, 233–237. O’Kane, C. J., and Gehring, W. J. (1987). Detection in situ of genomic regulatory elements in Drosophila. Proc. Natl. Acad. Sci. USA 84, 9123–9127. Schaft, J., et al. (2001). Efficient FLP recombination in mouse ES cells and oocytes. Genesis 31, 6–10. Schnutgen, F., et al. (2006). Engineering embryonic stem cells with recombinase systems. Meth. Enzymol. 420, 100–136. Valancius, V., and Smithies, O. (1991). Double-strand gap repair in a mammalian gene targeting reaction. Mol. Cell. Biol. 11, 4389–4397. Wang, W., et al. (2008). Chromosomal transposition of PiggyBac in mouse embryonic stem cells. Proc. Natl. Acad. Sci. USA 105, 9290–9295. Waterston, R. H., et al. (2002). Initial sequencing and comparative analysis of the mouse genome. Nature 420, 520–562. Yusa, K., et al. (2004). Enhancement of Sleeping Beauty transposition by CpG methylation: Possible role of heterochromatin formation. Mol. Cell. Biol. 24, 4004–4018.

C H A P T E R

S I X

The Use of DNA Transposons for Cancer Gene Discovery in Mice George Vassiliou, Roland Rad, and Allan Bradley Contents 1. Introduction 2. Choice of Transposon 3. Transposon Design for Cancer Screens 3.1. Types of transposons 3.2. Generation of transposon mice 4. Whole-Body (Constitutive) Screens 5. Tissue-Specific Mutagenesis 6. Inducible Transposase Expression 7. Mapping of Transposon Integration Sites 8. Statistical Mining of Recurrent Integration Sites 9. Validation of Putative Cancer Genes 10. Idiosyncrasies of Transposons as Cancer Gene Discovery Tools 11. Concluding Remarks References

92 93 95 95 96 98 98 100 100 101 101 102 103 103

Abstract Insertional mutagenesis in mice is a potent instrument for cancer gene discovery. Until recently, retroviruses were the main experimental tools in this field and application of insertional mutagenesis was limited to tissues for which these agents have tropism, namely hemopoietic cells and mammary epithelium. However, the field has been revolutionized and greatly expanded with the recent reanimation of the transposons, a highly flexible group of insertional mutagens first discovered in maize, which have now been adapted for use in mammalian cells. Transposons do not only extend the application of insertional mutagenesis to any tissue of choice, but also allow a more extensive and unbiased coverage of the genome, can be designed to selectively activate or inactivate genes, and are highly amenable to temporal and spatial control. This chapter gives an overview of the design and application of transposons to cancer gene discovery in mice. Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, United Kingdom Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77006-3

# 2010 by Allan Bradley. Published by Elsevier Inc. All rights reserved.

91

92

George Vassiliou et al.

1. Introduction Insertional mutagenesis is a potent tool for the study of gene function and one of its key applications is cancer gene discovery (Carlson and Largaespada, 2005; Kool and Berns, 2009). Cancer is the result of genetic mutations and recent advances in DNA sequencing technologies (Shendure and Ji, 2008) have greatly enhanced our ability to identify these. However, as is evident from several recent whole cancer genome sequencing projects (Bignell et al., 2010; Pleasance et al., 2010a,b), most mutations do not appear to have a role in the pathogenesis of the disease and the big challenge for cancer researchers is to distinguish the few ‘‘driver’’ mutations from the many ‘‘passenger’’ mutations in the same neoplasm (Stratton et al., 2009). Insertional mutagenesis offers a flexible tool for the unbiased interrogation of the genome for driver mutations in ways that can facilitate their differentiation from passengers and aid the dissection of synergistic genetic cancer pathways (Collier et al., 2005; Dupuy et al., 2005; Uren et al., 2008). The key advantage of insertional compared to other mutagens, such as radiation (Evans and DeMarini, 1999) or ENU ( Justice et al., 1999), is the fact that their genomic insertions can be easily mapped and their target genes thus identified. Retroviruses have served as powerful tools for the identification of cancer genes in mammals (Peters et al., 1983; Uren et al., 2005). The key limitation of retroviruses has been their tropism, which restricts their utility to hemopoietic (Steffen, 1984) and mammary tissues (Peters et al., 1983). Recent advances in the use of a transposons, a different group of insertional mutagens first described in maize (McClintock, 1950), are revolutionizing the field and extending the applications of insertional mutagenesis potentially to all tissues and cell types of interest (Ding et al., 2005; Dupuy et al., 2006; Ivics et al., 2004; Kool and Berns, 2009; Largaespada, 2009). In nature, DNA transposons are bipartite genetic elements that consist of the gene for a transposase enzyme flanked by its cognate recognition sequences ( Jacobson et al., 1986). In a host cell, expression of the transposase protein leads to excision of the element from its resident site in the genome and reinsertion into a new site using a ‘‘cut-and-paste’’ mechanism. This leads to (insertional) mutagenesis at the new site, which can in turn lead to lethality or disease consequent upon the specific effects of the inserted transposon(s) on local genes (Ivics et al., 2009). In experimental biology, the two components of the transposon can be separated. This creates a nonautonomous transposon whose mobilization is controlled by the transposase expressed in trans (Ivics et al., 2009). In principle, any type of genetic cargo can be introduced within the

93

The Use of DNA Transposons for Cancer Gene Discovery in Mice

transposon to meet the experimental purpose. Insertional mutagenesis can then be carried out in transposase/transposon double transgenic animals. When insertional mutagenesis is active in the germline, it leads to heritable mutations and this approach has been used (Dupuy et al., 2001; Fischer et al., 2001; Horie et al., 2001) extensively in genetic studies of lower organisms (Cooley et al., 1988; Youngman et al., 1983) and more recently of mice. Somatic insertional mutagenesis can promote cancer and thus be used for cancer gene/pathway discovery. Indeed, transposons are proving to be both extremely powerful and highly flexible molecular instruments for cancer gene discovery (Collier et al., 2005; Dupuy et al., 2005, 2006; Starr and Largaespada, 2005) with many advantages compared to retroviruses (Table 6.1). This review outlines the practical application of transposon-based insertional mutagenesis for cancer gene discovery in mice, focusing on the two main transposable elements used in this context to-date, namely Sleeping Beauty (SB) (Collier et al., 2005, 2009; Dupuy et al., 2005) and PiggyBac (PB) (unpublished).

2. Choice of Transposon Naturally occurring transposons are classified into two classes depending on their mechanism of transposition. Class I, or retrotransposons, transcribe themselves to RNA which is then reverse transcribed into DNA (often by a reverse transcriptase coded for by the transposon itself) and inserted back into the genome in a ‘‘copy-and-paste’’ mechanism (Cordaux and Batzer, 2009). Class II, or DNA transposons, do not use an Table 6.1 A comparison of insertional mutagens used for cancer gene discovery in mice Retroviruses

SB

Insertion site Integration bias

? TA Upstream of genes None

Cargo size

< 9 kb

Local hopping None tendency Amenability to genetic Low manipulation Target tissues Haematopoietic & mammary

Efficiency drops above 2 kb High

PB

TTAA Transcribed genes > 70 kb Low

High

High

Any

Any

94

George Vassiliou et al.

RNA intermediate but are simply moved from one location in the host genome to another by their cognate transposase enzyme coded for from within the transposon itself, a ‘‘cut-and-paste’’ mechanism (Kool and Berns, 2009). Of these two classes, the DNA transposons are those that have been used successfully for cancer gene discovery in higher species and thus form the focus of this review. The DNA transposon family has many members, including Minos, P-element, Tol2, SB, and PB (Ivics et al., 2009). SB was resurrected from dormant DNA transposons of the Tc1/mariner family resident in the genome of salmonid fish (Ivics et al., 1997) and has been found to be active in cells of a number of vertebrates, including zebrafish, mouse, and human (Ivics et al., 1997). PB is a transposon of the cabbage looper moth, Trichoplusia ni, which is also active in cells of several species, including mouse and human (Ding et al., 2005; Wilson et al., 2007). Additionally, both SB (Collier et al., 2005, 2009; Dupuy et al., 2005, 2009) and PB (unpublished) have been used successfully in insertional mutagenesis experiments for cancer gene discovery in mice. PB and SB have different biological properties and several aspects of mobilization and integration differ between the two systems. PB has a more potent transposase with higher mobilization efficiency (Liang et al., 2009; Wang et al., 2008) and is also able to carry larger DNA sequences, thus allowing greater flexibility with regards to transposon design (Wang et al., 2009). Another important difference is the increased tendency of SB for reintegration in the neighborhood of the donor locus (local hopping) (Liang et al., 2009; Wang et al., 2008). Furthermore, unlike PB, SB mobilization leaves a DNA footprint mutation at the donor locus, usually in the form of a 3-bp microdeletion (Liu et al., 2004; Luo et al., 1998), which can potentially result in undesirable occult mutagenesis. The activity of both SB and PB can be improved by modifying specific amino acid residues identified through random mutagenesis (Baus et al., 2005; Mates et al., 2009) or by species optimization for amino acid codon usage (Cadinanos and Bradley, 2007). It is also well recognized that different transposon systems have distinct integration preferences (Daniels and Deininger, 1985; Geurts et al., 2006b; Liu et al., 2005; Wang et al., 2008), such that saturation mutagenesis can only be achieved if multiple transposon systems are deployed in the same context (Thibault et al., 2004). PB integrates into TTAA (Ding et al., 2005) sites, while SB uses TA dinucleotides with a preference for palindromic TA repeats and for specific DNA confirmations (Geurts et al., 2006b). Beyond the basic sequence requirements for insertion, PB has a twofold higher preference for actively transcribed genes versus SB (Liang et al., 2009; Wang et al., 2008), which arguably makes it the preferred tool for tumor suppressor gene discovery.

95

The Use of DNA Transposons for Cancer Gene Discovery in Mice

3. Transposon Design for Cancer Screens 3.1. Types of transposons Transposons are mobilized through binding of the transposase to its cognate inverted terminal DNA repeats (ITRs). For increased flexibility, transposons can be designed with repeats for both SB and PB, so that they can be mobilized by both transposases. As PB is able to carry larger cargo, repeats for the latter are best placed outside those for SB. 3.1.1. Gene-activating transposons The genetic elements carried by individual transposons can be engineered to augment or attenuate gene function in specific ways (Collier et al., 2005; Dupuy et al., 2005). Regulatory elements such as enhancers and promoters can switch on or markedly increase expression of nearby genes. Such activating transposons are engineered with an enhancer placed directly upstream of a splice donor sequence (Fig. 6.1), which enables splicing into downstream exons when inserted upstream or within a gene. As the activity of most regulatory elements varies depending on the nature of the host cell, this can be exploited to promote cancer in cells/ tissues of interest. For example, the MSCV LTR is a potent activator of gene expression in hemopoietic cells leading to leukemia or lymphoma (Collier et al., 2005, 2009; Dupuy et al., 2005), while the hybrid CMV enhancer/chicken b-actin promoter (CAG) is largely inactive in the hemopoietic compartment but retains good activity in most other tissues Activating transposon RE

SD

Gene-trapping transposon SA

pA

pA

SA

Bifunctional transposon SA ITR

pA

pA

RE

SD

SA ITR

Figure 6.1 Types of transposons used in cancer gene discovery. ITR, Inverted Terminal Repeats; RE, regulator element (usually enhancer þ promoter); SD, splice donor; SA, splice acceptor; pA, polyadenylation signal.

96

George Vassiliou et al.

(unpublished). This makes CAG-carrying transposons excellent tools for solid tumor screens using SB (Dupuy et al., 2009) or PB (unpublished). 3.1.2. Gene-trapping transposons Naturally occurring splice acceptor constructs followed by polyadenylation sites (Fig. 6.1) are a highly effective way to trap the splicing machinery and create a 30 truncated mRNA. This leads to translation of a 30 -truncated protein, which is in most cases inactive and for this reason such constructs appear to be effective for the identification of tumor-suppressor genes (unpublished). It is, however, worth noting that in some cases 30 -truncation may activate oncogenes (Grasser et al., 1991). Splice acceptors that have been used in this context are the mouse Engrailed 2 (En2) splice acceptor, the Carp b-actin splice acceptor (Collier et al., 2005; Dupuy et al., 2005), and the adenovirus splice acceptor (unpublished). To enable transcriptional trapping in both orientations, two splice acceptors can be used facing in opposite orientations with bidirectional polyadenylation signals between them. 3.1.3. Bifunctional activating/inactivating transposons Gene-activating and -inactivating elements can be combined effectively in the same transposon (Fig. 6.1), and in fact only such transposons (e.g., T2Onc) have been used in cancer gene discovery to-date(Collier et al., 2005; Dupuy et al., 2005, 2009). As cancer often involves multiple events that can include simultaneous activation of oncogenes and inactivation of tumor suppressor genes, these transposons are highly effective carcinogens. The specific ways in which transposons can interfere with the function of genes is described in detail in Fig. 6.2. It is noteworthy that the 50 ITR for PB has inherent promoter activity in certain tissues (Cadinanos and Bradley, 2007) while its 30 ITR has trapping activity (Wang et al., 2008). These features need to be taken into account when designing gene-activating transposons.

3.2. Generation of transposon mice Mice carrying arrays of a designed transposon are generated by pronuclear injection of fertilized ova. This process leads to the incorporation of multiple copies of the transposon arranged head-to-tail at one or more genomic loci. The mutagenic efficiency of such arrays increases with copy number up to a point, but depending on the transposase used may become lower for very large arrays. Also the chromosomal location can influence mutagenicity depending on the context, albeit to a lesser extent (unpublished observations). In order to establish a comprehensive transposon system for genome-wide mutagenesis, the generation of several mouse lines carrying arrays of different types of transposon at varied copy numbers and chromosomal loci is required.

97

The Use of DNA Transposons for Cancer Gene Discovery in Mice

Gain of function through promoter insertion and overexpression of full length transcript SA pA pA Ex1

RE SD SA Ex2

Ex3

Ex4

Ex5

A

Ex1 B

Ex1 C

Gain of function through distant enhancer effects or through interference with regulatory elements—overexpression of full length transcript

+

D Ex1 Gain of function through directional intronic insertion and overexpression of 5¢-truncated transcript Ex1 E

Gain of function through gene trap insertion and 3¢-truncation of transcript Ex1 F

Loss of function through disruption of full length transcript Ex1 G

Figure 6.2 Potential effects of bi-functional (activating/inactivating) transposons on target genes. (A–C) Insertions in the sense orientation leading to overexpression of an intact open reading frame of target gene mRNA. Up to 40% of mouse genes have their first initiation codon (ATG) in the second or later exons (estimated using Ensemble’s Biomart on mouse datasets (NCBIM37)). For these genes transposon integration

98

George Vassiliou et al.

4. Whole-Body (Constitutive) Screens Control of transposase expression enables targeting of mutagenesis to the whole organism or to a specific tissue or cell type of interest. In both cases, the level of transposase expression relates to its mutagenic efficiency, and for this reason, knock-in alleles into high-expressing loci are generally preferred (e.g., Rosa26). Several mouse lines are available each carrying different versions of SB or PB transposases, including CAGGS-SB10 (Collier et al., 2005) transgenic, Rosa-SB11 (Dupuy et al., 2005) and Rosa-SB100 (unpublished) knock-in, and Rosa-PB knock-in (unpublished). All of these except Rosa-SB100 have been used successfully for cancer gene discovery, although CAGGS-SB10 only induced tumors in animals with predisposing mutations (Collier et al., 2005). Whole-body mutagenesis can lead to carcinogenesis in many tissues, therefore facilitating cancer gene discovery relevant to many different cancers (Collier et al., 2009; Fig. 6.3). Some control over the types of cancer generated can be exerted through the use of enhancer/promoter elements, which promote cancer in defined tissue types such as hemopoietic (Dupuy et al., 2005) or mesenchymal/epithelial (Dupuy et al., 2009). However, an important constraint in the application of whole-body screens is the incidence of embryonic lethality due to mutagenesis of the germline and early embryo. The incidence of embryonic lethality is influenced by the activity of the transposase, the transposon copy number, the genetic elements carried by the transposon, and the specific chromosomal location of the transposon donor array. Most important among these are an increased transposase activity and transposon copy number (Collier et al., 2009).

5. Tissue-Specific Mutagenesis For the specific study of individual types of cancer, mutagenesis can be restricted to the tissue/cell type of interest through conditional transposase expression. Tissue-specific transposon mutagenesis can be achieved through

upstream of exon 2 results in the expression of a full-length mRNA (A, B). This can also occur with some genes whose ORF starts in exon 1 but also have a (cryptic) splice site upstream of their first ATG (C). (D) Insertions leading to overexpression of a fulllength transcript due to distant enhancer effects. (E) Gain-of-function intronic insertions due to overexpression of a 50 -truncated mRNA. (F–G) Intronic insertions leading to gain (F) or loss (G) of function due to 30 truncation of the mRNA.

The Use of DNA Transposons for Cancer Gene Discovery in Mice

99

Constitutive mutagenesis Transposase mouse line Rosa26

SA

Transposase

Transposon mouse line SA pA pA

RE SD SA

Tissue-specific mutagenesis A Tissue-specific transposase line TS-promoter

Transposase

Transposon mouse line SA pA pA

RE SD SA

Tissue-specific mutagenesis B Conditional transposase mouse line Rosa26

SA LSL

Transposase

Cre mouse line TS-promoter Cre

Transposon mouse line SA pA pA RE SD SA

Figure 6.3 Alleles and crosses for constitutive and tissue-specific mutagenesis using transposons; TS, tissue-specific and LSL, loxP–stop–loxP cassette.

the use of conditional transposase alleles in combination with tissue-specific Cre-recombinase expression or the expression of transposase under the control of tissue-specific promoters (Fig. 6.3). For conditional alleles, two major approaches for controlling expression can be used: (i) the use of a lox–stop–lox cassette upstream of the transposase cDNA and (ii) the use of an inverted transposase cDNA flanked by mutant loxP sites that enables permanent inversion and expression of the sense-orientated sequence. Alternatively, when Cre lines are not available for the tissue/cell of interest, constructs expressing the transposase from a tissue-specific promoter could be used. Tissue-specific mutagenesis has been successfully used in the intestine (Starr et al., 2009), liver (Keng et al., 2009b), and hemopoietic lineage (unpublished) using SB and the intestine and pancreas using PB (unpublished).

100

George Vassiliou et al.

6. Inducible Transposase Expression Activation of conditional transposase alleles leads to permanent expression of the enzyme and therefore continuous mutagenesis in the tissue(s) of interest from early on in development. The use of inducible constructs allows a further level of control of the process. This can be done in two ways: (1) through the use of an inducible Cre-recombinase allele that can be activated at a time of choice using an inducer (e.g., Mx1-Cre induced by pIpC) or (2) through the use of an inducible transposase such as a chimaeric transposase–estrogen receptor that can be ‘‘activated’’ by its nuclear translocation by the action of tamoxifen. Unlike the first example, the use of an inducible transposase allows the removal of the mutagenic stimulus by withdrawing tamoxifen. This has been done for the hemopoietic system (unpublished). A tamoxifen-inducible PB transposase (PBER) has been described (Cadinanos and Bradley, 2007) and mouse lines carrying either a constitutive or a conditional version of PBER have been recently generated (unpublished).

7. Mapping of Transposon Integration Sites Activation of somatic mutagenesis in mice by either SB or PB leads to carcinogenesis with a speed and penetrance that is directly related to the mutagenic drive. Animals are therefore put on ‘‘tumor watch’’ and culled when they become unwell or develop large tumors. Tumors are harvested for pathological analysis and for the extraction of nucleic acids and protein. In order to identify which genes were disrupted by the transposons, their location in the genome needs to be identified. The most widely used method for this process is a ligation-based amplification of transposon– genome junctions known as ‘‘Splinkerette PCR.’’ This involves digestion of tumor genomic DNA with a restriction endonuclease and the ligation of linkers to the ‘‘sticky ends’’ so generated, followed by specific amplification of transposons–genome junctions by PCR (Qureshi et al., 1994). The mixed PCR products can then be cloned and individually sequenced. Alternatively, the protocol can be adapted for the application of highthroughput sequencing technologies, which allow the analysis of large numbers of sequences from multiple tumor DNAs simultaneously through the use of molecularly ‘‘bar-coded’’ primers (Uren et al., 2009).

The Use of DNA Transposons for Cancer Gene Discovery in Mice

101

8. Statistical Mining of Recurrent Integration Sites Depending on the starting transposon copy number and the activity of the transposase, a variable number of genomic integrations will be identified in any transposon-driven tumor. As with naturally occurring cancers, some of these integrations will be ‘‘drivers’’ of the carcinogenic process while others will be ‘‘passengers.’’ Genes that are hit recurrently in independent tumors are likely to represent ‘‘drivers’’ while ‘‘passenger’’ ones will be hit at random and therefore differ between different tumors. Importantly, however, the number of total mutations is much lower than the number in many human cancers (Pleasance et al., 2010a,b) and is decreasing further with the advent of higher efficiency transposases able to promote carcinogenesis with small numbers of transposons. Statistical processes are used to identify genomic loci that are hit more often than by chance and referred to as Common or Recurrent Integration Sites (CISs or RISs). One approach is based on the Monte Carlo simulation, which assumes that any genomic locus is as likely as any other to be hit by a transposon (described in detail in Keng et al., 2009b). More recently, a different approach has been developed known as the Kernel Convolution algorithm. This approach can examine the genome as a continuum rather than in predetermined fragment sizes and attempts to correct for insertion biases (de Ridder et al., 2006).

9. Validation of Putative Cancer Genes The first clue to the possible effects of transposons on genes near their CISs lies in the specific location and orientation of the integrated transposon. For example, if all or nearly all individual integrations from different tumors have the transposon in the sense orientation upstream the gene or in one of its first exons/introns, one can predict that the resulting molecular effect is overexpression of the gene, which thus may be an oncogene. By contrast, if integration is scattered throughout a gene locus and has no consistent orientation, the prediction can be made that this gene is a tumor suppressor gene (Fig. 6.2). High-throughput approaches for validation of the apparent behavior of putative target genes at CISs need to take into account the probable effect of the transposon on them. For example, putative hemopoietic oncogenes can be validated through retrovirus-mediated overexpression in hemopoietic

102

George Vassiliou et al.

cells. The study of tumor suppressor genes is usually more difficult and although shRNA-based approaches can be tried, targeted disruption of the endogenous gene may be the best approach. The latter will be increasingly facilitated by high-throughput mouse gene knock-out programs such as EUCOMM (Friedel et al., 2007) and KOMP (Guan et al., 2010).

10. Idiosyncrasies of Transposons as Cancer Gene Discovery Tools Carefully designed transposons can lead to activation or inactivation of genes in specific ways (Collier et al., 2005; Dupuy et al., 2005). However, certain types of specific human mutations cannot be recapitulated and most important among these are activating point mutations. For example, activating point mutations in the Ras genes are widespread in human cancer (Bos, 1989) and have specific effects on cellular proliferation and transformation (Haigis et al., 2008), and although simple overexpression of Ras can also be oncogenic (Mangues et al., 1992), its molecular effects differ from those of point mutations (Hancock et al., 1988; Malumbres and Barbacid, 2003). Nevertheless, the Ras genes do not evade identification by insertional mutagens in cancer screens (Akagi et al., 2004) and this is also true of other genes usually activated in by point mutation in human cancers, including Flt3 (Uren et al., 2008) and Braf (Collier et al., 2005). It is, however, possible that some genes can only be activated in very specific ways and therefore are not susceptible to identification by insertional mutagens. This may hold true for a small subset of fusion genes where both partners are required for transformation. Another limitation when using transposons is the frequent occurrence of chromosomal deletions at the site of the transposon array, which occur relatively frequently after transposon mobilization and can sometimes be several megabases long (Geurts et al., 2006a). This phenomenon can itself promote carcinogenesis due to loss of proximal tumor suppressor genes in a way, which cannot be tracked as is the case for transposon insertions. This problem can be overcome to a significant degree through the use of more than one mouse line with transposon arrays resident on different chromosomes for the same experiment (unpublished). SB and to a lesser extend PB are prone to local hopping, the phenomenon of transposon reintegration to a nearby genomic location. This leads to a higher rate of mutagenesis in the vicinity of the transposon array than anywhere else in the genome, such that recurrent integrations in such loci are more likely to occur by chance and cannot be confidently identified as candidate cancer genes. Additionally, when SB is used local hopping increases the likelihood of untraceable footprint mutations in local genes.

The Use of DNA Transposons for Cancer Gene Discovery in Mice

103

As with chromosomal deletions, these problems can be circumvented through the use of more than one mouse line with transposon arrays resident at different chromosomal loci. It is also worth noting that, while being a disadvantage for some applications, local hopping can also be exploited for regional saturation mutagenesis of specific genomic regions (Izsvak and Ivics, 2005).

11. Concluding Remarks The emergence of transposons as genetic tools active in mammalian cells is revolutionizing cancer gene discovery. Transposons can serve to both complement large cancer sequencing efforts currently in progress (Stratton et al., 2009) and independently validate cancer ‘‘driver’’ mutations in an unbiased and relatively ‘‘passenger-free’’ environment. The controlled design and application of transposon-based mutagenesis can also serve to answer specific questions in cancer biology such the deciphering of collaborating cancer networks in defined mouse backgrounds (Collier et al., 2005; Keng et al., 2009b). Additionally, tumors derived from insertional mutagenesis experiments can be propagated in culture or in vivo and used in the study of the molecular effects of specific mutations. Although transposons have an apparently unbiased access to the whole of the mouse genome (barring local hopping), it is clear that SB an PB have subtle integration preferences which manifest in their ability to identify a different, albeit overlapping, set of putative cancer genes in an otherwise identical context (unpublished). Therefore, use of both PB and SB as well as possibly other systems being introduced at present (de Wit et al., 2010; Keng et al., 2009a), may be necessary for genome-wide saturation mutagenesis. Large insertional mutagenesis projects are currently underway by us and others to systematically interrogate the genome for cancer causing mutations in different organs/tissues. The integration of these findings with those from human cancer sequencing projects promises to transform the landscape of our understanding of cancer and our ability to develop therapeutic interventions.

REFERENCES Akagi, K., et al. (2004). RTCGD: Retroviral tagged cancer gene database. Nucleic Acids Res. 32, D523–D527. Baus, J., et al. (2005). Hyperactive transposase mutants of the Sleeping Beauty transposon. Mol. Ther. 12, 1148–1156. Bignell, G. R., et al. (2010). Signatures of mutation and selection in the cancer genome. Nature 463, 893–898.

104

George Vassiliou et al.

Bos, J. L. (1989). Ras oncogenes in human cancer: A review. Cancer Res. 49, 4682–4689. Cadinanos, J., and Bradley, A. (2007). Generation of an inducible and optimized piggyBac transposon system. Nucleic Acids Res. 35, e87. Carlson, C. M., and Largaespada, D. A. (2005). Insertional mutagenesis in mice: New perspectives and tools. Nat. Rev. Genet. 6, 568–580. Collier, L. S., et al. (2005). Cancer gene discovery in solid tumours using transposon-based somatic mutagenesis in the mouse. Nature 436, 272–276. Collier, L. S., et al. (2009). Whole-body sleeping beauty mutagenesis can cause penetrant leukemia/lymphoma and rare high-grade glioma without associated embryonic lethality. Cancer Res. 69, 8429–8437. Cooley, L., et al. (1988). Insertional mutagenesis of the Drosophila genome with single P elements. Science 239, 1121–1128. Cordaux, R., and Batzer, M. A. (2009). The impact of retrotransposons on human genome evolution. Nat. Rev. Genet. 10, 691–703. Daniels, G. R., and Deininger, P. L. (1985). Integration site preferences of the Alu family and similar repetitive DNA sequences. Nucleic Acids Res. 13, 8939–8954. de Ridder, J., et al. (2006). Detecting statistically significant common insertion sites in retroviral insertional mutagenesis screens. PLoS Comput. Biol. 2, e166. de Wit, T., et al. (2010). Tagged mutagenesis by efficient Minos-based germ line transposition. Mol. Cell Biol. 30, 68–77. Ding, S., et al. (2005). Efficient transposition of the piggyBac (PB) transposon in mammalian cells and mice. Cell 122, 473–483. Dupuy, A. J., et al. (2001). Transposition and gene disruption in the male germline of the mouse. Genesis 30, 82–88. Dupuy, A. J., et al. (2005). Mammalian mutagenesis using a highly mobile somatic Sleeping Beauty transposon system. Nature 436, 221–226. Dupuy, A. J., et al. (2006). Sleeping beauty: A novel cancer gene discovery tool. Hum. Mol. Genet. 15(Spec No 1), R75–R79. Dupuy, A. J., et al. (2009). A modified sleeping beauty transposon system that can be used to model a wide variety of human cancers in mice. Cancer Res. 69, 8150–8156. Evans, H. H., and DeMarini, D. M. (1999). Ionizing radiation-induced mutagenesis: Radiation studies in Neurospora predictive for results in mammalian cells. Mutat. Res. 437, 135–150. Fischer, S. E., et al. (2001). Regulated transposition of a fish transposon in the mouse germ line. Proc. Natl. Acad. Sci. USA 98, 6759–6764. Friedel, R. H., et al. (2007). EUCOMM–the European conditional mouse mutagenesis program. Brief Funct. Genomic. Proteomic. 6, 180–185. Geurts, A. M., et al. (2006a). Gene mutations and genomic rearrangements in the mouse as a result of transposon mobilization from chromosomal concatemers. PLoS Genet. 2, e156. Geurts, A. M., et al. (2006b). Structure-based prediction of insertion-site preferences of transposons into chromosomes. Nucleic Acids Res. 34, 2803–2811. Grasser, F. A., et al. (1991). Protein truncation is required for the activation of the c-myb proto-oncogene. Mol. Cell Biol. 11, 3987–3996. Guan, C., et al. (2010). A review of current large-scale mouse knockout efforts. Genesis 48, 73–85. Haigis, K. M., et al. (2008). Differential effects of oncogenic K-Ras and N-Ras on proliferation, differentiation and tumor progression in the colon. Nat. Genet. 40, 600–608. Hancock, J. F., et al. (1988). Mutant but not normal p21 ras elevates inositol phospholipid breakdown in two different cell systems. Oncogene 3, 187–193. Horie, K., et al. (2001). Efficient chromosomal transposition of a Tc1/mariner-like transposon Sleeping Beauty in mice. Proc. Natl. Acad. Sci. USA 98, 9191–9196.

The Use of DNA Transposons for Cancer Gene Discovery in Mice

105

Ivics, Z., et al. (1997). Molecular reconstruction of Sleeping Beauty, a Tc1-like transposon from fish, and its transposition in human cells. Cell 91, 501–510. Ivics, Z., et al. (2004). The Sleeping Beauty transposable element: Evolution, regulation and genetic applications. Curr. Issues Mol. Biol. 6, 43–55. Ivics, Z., et al. (2009). Transposon-mediated genome manipulation in vertebrates. Nat. Methods 6, 415–422. Izsvak, Z., and Ivics, Z. (2005). Sleeping Beauty hits them all: Transposon-mediated saturation mutagenesis in the mouse germline. Nat. Methods 2, 735–736. Jacobson, J. W., et al. (1986). Molecular structure of a somatically unstable transposable element in Drosophila. Proc. Natl. Acad. Sci. USA 83, 8684–8688. Justice, M. J., et al. (1999). Mouse ENU mutagenesis. Hum. Mol. Genet. 8, 1955–1963. Keng, V. W., et al. (2009a). Efficient transposition of Tol2 in the mouse germline. Genetics 183, 1565–1573. Keng, V. W., et al. (2009b). A conditional transposon-based insertional mutagenesis screen for genes associated with mouse hepatocellular carcinoma. Nat. Biotechnol. 27, 264–274. Kool, J., and Berns, A. (2009). High-throughput insertional mutagenesis screens in mice to identify oncogenic networks. Nat. Rev. Cancer 9, 389–399. Largaespada, D. A. (2009). Transposon mutagenesis in mice. Methods Mol. Biol. 530, 379–390. Liang, Q., et al. (2009). Chromosomal mobilization and reintegration of Sleeping Beauty and PiggyBac transposons. Genesis 47, 404–408. Liu, G., et al. (2004). Excision of Sleeping Beauty transposons: Parameters and applications to gene therapy. J. Gene. Med. 6, 574–583. Liu, G., et al. (2005). Target-site preferences of Sleeping Beauty transposons. J. Mol. Biol. 346, 161–173. Luo, G., et al. (1998). Chromosomal transposition of a Tc1/mariner-like element in mouse embryonic stem cells. Proc. Natl. Acad. Sci. USA 95, 10769–10773. Malumbres, M., and Barbacid, M. (2003). RAS oncogenes: The first 30 years. Nat. Rev. Cancer 3, 459–465. Mangues, R., et al. (1992). Overexpression of the N-ras proto-oncogene, not somatic mutational activation, associated with malignant tumors in transgenic mice. Oncogene 7, 2073–2076. Mates, L., et al. (2009). Molecular evolution of a novel hyperactive Sleeping Beauty transposase enables robust stable gene transfer in vertebrates. Nat. Genet 41, 753–761. McClintock, B. (1950). The origin and behavior of mutable loci in maize. Proc. Natl. Acad. Sci. USA 36, 344–355. Peters, G., et al. (1983). Tumorigenesis by mouse mammary tumor virus: Evidence for a common region for provirus integration in mammary tumors. Cell 33, 369–377. Pleasance, E. D., et al. (2010a). A comprehensive catalogue of somatic mutations from a human cancer genome. Nature 463, 191–196. Pleasance, E. D., et al. (2010b). A small-cell lung cancer genome with complex signatures of tobacco exposure. Nature 463, 184–190. Qureshi, S. J., et al. (1994). Alu-based vectorettes and splinkerettes. More efficient and comprehensive polymerase chain reaction amplification of human DNA from complex sources. Genet. Anal. Tech. Appl. 11, 95–101. Shendure, J., and Ji, H. (2008). Next-generation DNA sequencing. Nat. Biotechnol. 26, 1135–1145. Starr, T. K., and Largaespada, D. A. (2005). Cancer gene discovery using the Sleeping Beauty transposon. Cell Cycle 4, 1744–1748. Starr, T. K., et al. (2009). A transposon-based genetic screen in mice identifies genes altered in colorectal cancer. Science 323, 1747–1750.

106

George Vassiliou et al.

Steffen, D. (1984). Proviruses are adjacent to c-myc in some murine leukemia virus-induced lymphomas. Proc. Natl. Acad. Sci. USA 81, 2097–2101. Stratton, M. R., et al. (2009). The cancer genome. Nature 458, 719–724. Thibault, S. T., et al. (2004). A complementary transposon tool kit for Drosophila melanogaster using P and piggyBac. Nat. Genet. 36, 283–287. Uren, A. G., et al. (2005). Retroviral insertional mutagenesis: Past, present and future. Oncogene 24, 7656–7672. Uren, A. G., et al. (2008). Large-scale mutagenesis in p19(ARF)- and p53-deficient mice identifies cancer genes and their collaborative networks. Cell 133, 727–741. Uren, A. G., et al. (2009). A high-throughput splinkerette-PCR method for the isolation and sequencing of retroviral insertion sites. Nat. Protoc. 4, 789–798. Wang, W., et al. (2008). Chromosomal transposition of PiggyBac in mouse embryonic stem cells. Proc. Natl. Acad. Sci. USA 105, 9290–9295. Wang, W., et al. (2009). A piggyBac transposon-based genome-wide library of insertionally mutated Blm-deficient murine ES cells. Genome Res. 19, 667–673. Wilson, M. H., et al. (2007). PiggyBac transposon-mediated gene transfer in human cells. Mol. Ther. 15, 139–145. Youngman, P. J., et al. (1983). Genetic transposition and insertional mutagenesis in Bacillus subtilis with Streptococcus faecalis transposon Tn917. Proc. Natl. Acad. Sci. USA 80, 2305–2309.

C H A P T E R

S E V E N

A Practical Summary of Site-Specific Recombination, Conditional Mutagenesis, and Tamoxifen Induction of CreERT2 Konstantinos Anastassiadis,* Stefan Glaser,†,1 Andrea Kranz,† Kaj Bernhardt,† and A. Francis Stewart† Contents 110 110 112 112 114 114 114 117 117 118 119 119 119

1. 2. 3. 4. 5.

Introduction Recombinase Target Sites Applications Allele Design Recombinase Properties 5.1. Flp recombinase 5.2. Cre zoo and conditional mutagenesis 5.3. Dre recombinase 6. Tamoxifen Administration in Mice and Cultured Cells 7. Problems 8. Concluding Remarks Acknowledgments References

Abstract The site-specific recombinases, Cre and Flp, are essential tools for altering the mouse genome. Since the pioneering work with these enzymes, much progress has been made regarding their strengths and weaknesses, as well as how they should be applied. Cre recombinase is vital for conditional mutagenesis, including for temporal mutagenesis via tamoxifen induction of recombinase—estrogen receptor fusion proteins. Recently a Cre-like recombinase, Dre has been

* Center for Regenerative Therapy, Technische Universitaet Dresden, Dresden, Germany Genomics, BioInnovationsZentrum, Technische Universitaet Dresden, Dresden, Germany Current address: The Walter and Eliza Hall Institute, Melbourne, Victoria, Australia

{ 1

Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77007-5

#

2010 Elsevier Inc. All rights reserved.

109

110

Konstantinos Anastassiadis et al.

added to the toolbox. Here, we summarize current knowledge about applications of Cre, Flp, and Dre in the mouse and present a protocol for tamoxifen induction of conditional mutagenesis.

1. Introduction Site-specific recombinases fall into two distinct classes termed tyrosine and serine according to the catalytic amino acid involved in their enzymatic mechanisms (Grindley et al., 2006). The tyrosine recombinases, including Cre and Flp, have proven to be the most useful for DNA and genomic engineering (Branda and Dymecki, 2004; Glaser et al., 2005). The serine recombinases were long held to be unsuitable for eukaryotic applications until the discovery of the subclass termed the large serine recombinases, which include phiC31 and pBT1 (Smith and Thorpe, 2002). In this chapter, we will only deal with the tyrosine recombinases.

2. Recombinase Target Sites Most tyrosine recombinases mediate recombination through palindromic recombinase target sites (RTs) based on
13 bp inverted repeats that flank a nonpalindromic central spacer of 6–8 bp (Table 7.1). Each half of the inverted repeat binds a monomer and recombination occurs when two dimer bound RTs contact each other after random collision (Ringrose et al., 1998, 1999) to establish a tetrameric, square planar synapse (Chen and Rice, 2003; Ennifar et al., 2003; Ghosh et al., 2007; van Duyne, 2001). Because the spacer is not palindromic, RTs are represented as arrowheads to indicate the directionality of the spacer. Recombination occurs through the spacer, so that the product of recombination contains a hybrid of one spacer strand from each substrate. Consequently, recombination can only occur between two RTs that have the same spacer sequence. This fact has been exploited to establish homotypic and heterotypic RTs having the same or different spacer sequences, respectively. Thereby, two or more different recombination events using a single recombinase can be programmed into the same cell (or test tube) simply by employing different pairs of homotypic RTs. For Cre recombinase, two heterotypic lox sites have been established termed lox5171 and lox2272 (Lee and Saito, 1998; Table 7.1). Usually, a pair of lox5171 or 2272 sites has been employed together with a pair of loxP sites. Notably, another heterotypic site, lox511, has been reported by several authors to retain significant ability to recombine with loxP sites (e.g., Kolb, 2001) and is no longer used as a heterotypic lox site.

Table 7.1 The RT of Cre, Flp, and Dre recombinases are shown, including the most commonly used variations RT

Inverted repeat

Spacer

Inverted repeat

Reference

loxP lox5171 lox2272 lox71 lox66 loxJTZ17 rox FRT F3 F5

ATAACTTCGTATAG ATAACTTCGTATAG ATAACTTCGTATAG TACCGTTCGTATAG ATAACTTCGTATAG ATAACTTCGTATAG TAACTTTAAATAA GAAGTTCCTATTC GAAGTTCCTATTC GAAGTTCCTATTC

CATACA TAGACA GATACT CATACA CATACA CATACA TTGGCA TCTAGAAA TTTAAATA TTCAAAAG

TTATACGAAGTTAT TTATACGAAGTTAT TTATACGAAGTTAT TTATACGAAGTTAT TTATACGAACCGTA TTATAGCAATTTAT TTATTTAAAGTTA GTATAGGAACTTC GTATAGGAACTTC GTATAGGAACTTC

Hoess et al. (1986) Lee and Saito (1998) Lee and Saito (1998) Albert et al. (1995) Albert et al. (1995) Thomson et al. (2003) Sauer and McDermott (2004) Senecoff et al. (1988) Schlake and Bode (1994) Schlake and Bode (1994)

Each RT comprises an inverted repeat flanking a spacer sequence. One recombinase monomer binds each inverted repeat half site and covalently attaches to its 30 end, freeing the 50 end of the spacer. Consequently, the spacer for loxP and presumably also rox is here defined as 6 bp, whereas the spacer for FRT is 8 bp. Variations from the wild type sequences are shown in bold.

112

Konstantinos Anastassiadis et al.

3. Applications Most tyrosine recombinases, including Cre and Flp, do not require cofactors and bind their RTs as homodimers. With the exception of l-Int, which requires cofactors and is the basis of the in vitro DNA engineering method called Gateway (Invitrogen), the useful tyrosine recombinases are therefore inherently reversible. This fact limits the way in which they can be applied (Glaser et al., 2005; Logie and Stewart, 1995; Schnu¨tgen et al., 2003). The most common application employs the RTs as direct repeats (i.e., with the spacers arranged in the same orientation). Recombination then excises the DNA between the repeats. When employed as indirect repeats, recombination inverts the DNA between the RTs. When employed on different DNA molecules, for example different chromosomes, recombination directs a translocation according to the orientation of the spacers. When employed as combinations of RTs (either heterotypic or for different SSRs), a variety of applications can be achieved. Because Cre and Flp recombinations are inherently reversible, two strategies have been developed to impose directionality. The first strategy, which was developed by Ow and colleagues (Albert et al., 1995), utilizes two mutant lox sites termed lox66 and lox71 (Table 7.1). Because Cre binding to loxP sites is cooperative (Ringrose et al., 1998), binding of one Cre monomer to a wildtype (wt) half site in an inverted repeat will promote Cre occupancy of the other half site even when it is mutated. Hence, binding of a Cre dimer to a wt-mutant lox site (lox66) or a mutant-wt lox site (lox71) permits recombination between these lox sites. However, after recombination, one product site will contain both mutant half sites and so will not serve to mediate the reverse recombination reaction. A second strategy, termed FlEx for Flip-Excision, which employs interdigitated pairs of heterotypic RTs or wt RTs from two different SSRs (usually loxP/Cre and FRT/Flp), also delivers directional recombination (Schnu¨tgen et al., 2003, 2005). The interested reader can find a more detailed presentation of these possibilities in earlier reviews (Glaser et al., 2005; Schnu¨tgen et al., 2006).

4. Allele Design The inclusion of RTs in an allele adds the further dimension of sitespecific recombination to allele design. There are at least four main applications. The accompanying chapter by Fu et al. (Chapter 8) in this volume also addresses issues of allele design for conditional mutagenesis.

Site-Specific Recombination, Conditional Mutagenesis, and Tamoxifen Induction

113

First, the simplest application involves using site-specific recombination to remove the selectable gene from targeted alleles. The continued presence of foreign DNA encoding a selectable gene, particularly sequences that encode cis elements like promoters, enhancers, or splice junctions, can seriously interfere with normal expression of the targeted gene (e.g., Kaul et al., 2000). Removal of these operational sequences is essential to eliminate potential artifacts in functional analysis. This task is most often achieved by flanking the selectable gene with FRTs (Flp RTs) and effecting Flp recombination in ES cells or mice (Kranz et al., 2010; Rodriguez et al., 2000; Schaft et al., 2001; Wu et al., 2009). The second main application is conditional mutagenesis. This task is usually achieved by putting loxP sites (RTs for Cre) into introns flanking a chosen exon(s) of the target gene. The exon(s) is usually chosen to promote a frameshift in the mRNA after removal. Alternatively, the exon(s) is chosen to remove an essential protein function, an alternative splicing event or a stop codon. Hence, the target gene is poised for mutagenesis by Cre recombinase and the mode of Cre recombinase delivery determines the place and time of the mutagenic event. Multipurpose alleles present a third application of RTs, which follows from the above two cases. An allele with an embedded pair of RTs has two states, namely before and after recombination. If the allele has RTs for both Flp and Cre, then it has four states, namely before recombination, after Flp recombination only, after Cre recombination only, and after both. This multipurpose strategy can be used to create an allelic series (Nagy et al., 1998) or as a control to evaluate whether the allele is a knock-out (i.e., a complete loss of function), hypomorph, or dominant negative (Testa et al., 2003, 2004). Recently, we established that a new Cre-like recombinase, Dre, together with its RTs, rox, can be used efficiently in mice (Anastassiadis et al., 2009). Among other applications, Dre now permits even more elaborate designs for multipurpose alleles. Further applications include RMCE (recombinase-mediated cassette exchange), which is a method for insertion of transgenes into genomic loci utilizing heterotypic RTs (Baer and Bode, 2001; Wallace et al., 2007), FlEx, which is a method for directional inversion of a DNA segment (Schnu¨tgen et al., 2003, 2005), and chromosomal engineering (van der Weyden et al., 2009; Yu and Bradley, 2001). Combinations of these applications are also feasible. Hence SSR technology permits a variety of useful and sophisticated strategies. The key to many of these applications is the ability to put the various RTs exactly where intended in the targeting construct or transgene. Thereby designs can be achieved according to concept rather than being limited by the technical limitations of DNA engineering. Herein lies one of the advantages of recombineering over conventional DNA engineering methods (see Chapter 8 for further information).

114

Konstantinos Anastassiadis et al.

5. Recombinase Properties 5.1. Flp recombinase Because the original Flp has an unsuitable enzymatic temperature optimum for application in mice (Buchholz et al., 1996), an improved mutagenic version termed Flpe was developed using molecular evolution (Buchholz et al., 1998). Flpe has been recently further improved by codon optimization that is termed ‘‘Flpo’’ (Kranz et al., 2010; Raymond and Soriano, 2007). In mammalian cells, Flpo appears to be about five times more active than Flpe, which in turn was about five times more active than wt Flp. Flp/FRTs have been primarily used for the easiest applications of site-specific recombination, such as the removal of the selectable gene after gene targeting. This can be achieved by transfection of an expression construct into ES cells or crossing to a Flp deleter mouse line. Currently, the best expression vectors are PGK-Flpo (Raymond and Soriano, 2007) or pCAGGS-Flpo (Kranz et al., 2010), and good Flpo deleter mice are available on the C57BL6N background (Kranz et al., 2010; Wu et al., 2009), so are suitable for use with the IKMC (International Knock-out Mouse Consortium) resources (www. ikmc.org). In mice, Flpe has also been successfully applied for lineage tracing studies, notably, in combination with Cre to identify ‘‘intersectional’’ cells characterized by the expression of two chosen genes (Awatramani et al., 2003; Kim and Dymecki, 2009).

5.2. Cre zoo and conditional mutagenesis Many remarkable genome feats have been accomplished with Cre recombinase. It is the sharpest tool in the box and works efficiently in every organism tested so far, including prokaryotes, plants, flies, fish, and mammals. Consequently, in mice, Cre is used for the most demanding applications such as conditional mutagenesis or genome engineering (Branda and Dymecki, 2004; Glaser et al., 2005; van der Weyden et al., 2009). As opposed to Flpe/Flpo, codon optimization of the original, prokaryotic, Cre coding region to mammalian codon usage delivered only a very small benefit. Consequently, the codon altered version, termed ‘‘iCre’’ (Shimshek et al., 2002) is only in sporadic usage. Conditional mutagenesis in the animal is usually achieved by crossing a mouse line carrying a ‘‘floxed’’ gene (i.e., a gene with loxP sites flanking an exon or exons) with a mouse line carrying a Cre transgene. Many Cre expressing mouse lines have been established thereby creating the ‘‘Cre zoo’’ (Rajewsky et al., 1996) and efforts are underway to systematically organize this important resource for easier access by researchers (http://dev. creline.org/home). There are four different types of mice in the Cre zoo.

Site-Specific Recombination, Conditional Mutagenesis, and Tamoxifen Induction

115

(i) Deleters. Cre deleters express Cre so that a cross with a floxed mouse results in complete deletion in the germline. Several Cre deleter lines are in use; however, a Cre deleter made on the standardized IKMC C57BL6N background (Pettitt et al., 2009) is urgently needed. As opposed to Flpe deleters, all published Cre deleters provoke complete germline recombination of small loxP-flanked cassettes. Indeed excess Cre activity may be the source of problems such as unwanted rearrangements (Schmidt et al., 2000). Certain deleter lines can be used to provoke germline rearrangements to delete, invert, or translocate sizable regions up to a megabase scale (He´rault et al., 1998; Wu et al., 2007; Yu and Bradley, 2001). (ii) Spatial regulation. Most of the existing Cre zoo is based on mouse lines that express Cre in a limited number of cell types. Cell-type-restricted Cre expression is often referred to as ‘‘spatial’’ regulation. This straightforward objective has proven difficult to achieve in many cases because conventional transgenes (i.e., small ones based on plasmid constructs) often fail to recapitulate the intended gene expression pattern and/or ectopically express Cre early in development. If transient Cre expression occurs during early development, recombination can be widespread even though the Cre transgene shows the intended restricted expression later in development or in the adult. To avoid some of these small transgene problems, most new Cre zoo mouse lines are based on BAC transgenes, which are now easy to generate due to the development of recombineering (Copeland et al., 2001; Muyrers et al, 1999; Zhang et al., 1998). BAC transgenes usually recapitulate the expression pattern of the chosen gene precisely with copy number dependence. Some Cre zoo lines have been generated by knock-in targeting of Cre into an endogenous locus. This approach entails more work and is more likely to cause unwanted phenotypes than random insertional transgenesis. Hence BAC transgenes have emerged as the preferred option (Vintersten et al., 2008). (iii) Temporal regulation. Ubiquitous temporal regulation of Cre recombination in mice has been (almost) achieved by expression of a Cremutant estrogen receptor fusion protein, CreERT2, from the ROSA26 locus. Regulation of a site-specific recombinase by fusion to the ligand binding domain of a steroid receptor (Kellendonk et al., 1996; Logie and Stewart, 1995) delivers robust ligand inducible recombination in mice (Feil et al., 1996; Schwenk et al., 1998). The first version used in mice, CreERT, employed a point mutated estrogen binding domain (G400V) that is insensitive to endogenous estrogens but still activatable by a synthetic antiestrogen, 4-OH tamoxifen. Subsequently, Metzger and Chambon developed CreERT2, which has a higher affinity for 4-OH tamoxifen (Feil et al., 1997). CreERT2 is now the most popular and reliable way to temporally regulate recombination by ligand administration.

116

Konstantinos Anastassiadis et al.

Temporal regulation to achieve ubiquitous recombination after ligand administration was first attempted using an inducible promoter (Ku¨hn et al., 1995). After a great deal of work, one mouse line with reasonable properties, Mx-Cre, was established. However, this mouse delivers useful inducible recombination only in a few tissues/cell types. For reasons that are unclear, attempts to establish a tetracycline inducible Cre have been largely unrewarding (Scho¨nig et al., 2002). By targeting CreERT2 to the ROSA26 locus, the ROSA26 CreERT2 mouse has been generated more than once (Hameyer et al., 2007; Seibler et al., 2003; Ventura et al., 2007). Whether these lines are identical remains to be established. We have worked with the version made at Artemis GmbH (Seibler et al., 2003) and found that it delivers near-ubiquitous expression and tamoxifen-induced recombination in adults (Glaser et al., 2009). However, recombination can only be poorly induced in the brain because the ROSA26 locus expresses only weakly in postmitotic neurons (Hameyer et al., 2007). Tamoxifen can readily induce CreERT2 in the brain (Badea et al., 2009; Casanova et al., 2002), showing that the lack of tamoxifen induction of recombination in ROSA26 CreERT2 brains is not due to the blood-brain barrier. Tamoxifen can be used to induce recombination in utero by administration to pregnant mothers (Danielian et al., 1998; Glaser et al., 2009; Hameyer et al., 2007). However, the dose must be limited otherwise rapid abortion will occur. This problem is severe before E9.5 but less so later in pregnancy. Even if rapid abortion is avoided, the embryos will die about 4 days after tamoxifen administration. During this time, the embryos continue to develop so certain experiments can be accomplished, particularly if heterozygous littermates serve as controls. In one case using ROSA26 CreERT2, we found that 1 mg of tamoxifen given to the mother by gavage twice, 24 h apart, induced efficient recombination in utero without provoking rapid abortion (Glaser et al., 2009). However, this floxed allele recombines readily. Possibly, other floxed alleles will require stronger doses such that complete recombination cannot be attained. (iv) Combined spatial–temporal regulation. Many combinations of spatial and temporal regulations have been used. Expressing CreERT2 from a cell type specific transgene has proven to be a reliable way to achieve very precise control over recombination (Branda and Dymecki, 2004). Notably, this approach circumvents the difficulties encountered when a Cre transgene is transiently expressed early in development. Recently a tetracycline inducible CreERT2 transgene has been targeted to ROSA26 (Badea et al., 2009). Upon introduction of a tetracycline-regulated transactivator, this line permits highly selective dual ligand (tetracycline, tamoxifen) regulation of loxP recombination, which could be particularly useful for studies in the brain.

Site-Specific Recombination, Conditional Mutagenesis, and Tamoxifen Induction

117

5.3. Dre recombinase Dre recombinase was identified by an explicit search for another genome engineering tool as good as Cre (Sauer and McDermott, 2004). Recently we showed that Dre works efficiently in mice and have established deleter and ROSA26 rox (the Dre RT; Table 7.1) reporter mouse lines (Anastassiadis et al., 2009). Furthermore, Dre and Cre do not cross recombine under any circumstance tested, despite similarities in their RTs. The acquisition of a second tool as good as Cre (Kranz et al., 2010) permits further complexities and specializations. For example, Cre/loxP can now be dedicated to conditional mutagenesis and Dre dedicated to genome engineering. For another example, double conditional strategies to dissect alternatively spliced functions can be developed.

6. Tamoxifen Administration in Mice and Cultured Cells Please note that the cognate ligand is 4-OH tamoxifen. Tamoxifen administered to mice is metabolized in the liver to 4-OH tamoxifen. For experiments with cultured cells, 4-OH tamoxifen must be used. Tamoxifen can be administered to mice with food by purchasing custom-made chow (containing 0.4–1 g/kg tamoxifen, Harlan) or water intake (0.5–1 mg/ml tamoxifen) from water bottles. We prefer direct administration by oral gavage or intraperitoneal injection because the dose is more precise. Tamoxifen (Sigma, T5648) does not dissolve well in water but is soluble in peanut oil (Sigma, P2144) containing 10% ethanol at a concentration of 60 mg/ml. This solution can be obtained by weighing 90 mg tamoxifen into an Eppendorf tube, adding 150 l 100% EtOH first and topping up to 1.5 ml with oil. It is recommended to use the volume labeling on the Eppendorf tube because pipetting of oil is inaccurate. Dissolve the tamoxifen at 50  C (we use an Eppendorf heat block) with occasional vortexing. It should remain in solution at room temperature. We recommend four or five daily doses by oral gavage of 200 mg/kg or 5 mg for a 25 g mouse. Cut a P200 pipette tip and divide the 60 mg/ml Tamoxifen stock into aliquots of 4, 5, or 6 mg (67, 83, or 100 l, respectively) depending on the average body weight of the mouse cohort (20, 25, or 30 g, respectively). Keep the aliquots at 20  C and thaw at 37  C prior to use. For gavage, we prefer a reusable, bulb-tipped feeding needle (Agnthos, catalogue number 7920 or 7902; 20–22 gauge  3.7 cm long, 2.25 mm Ball diameter). The conscious mouse is manually restrained by grasping the skin on the neck with the left hand if right-handed. It is essential to pinch a large fold of skin to immobilize the head and forelimbs, but without choking the

118

Konstantinos Anastassiadis et al.

mouse. Immobilize the lower body of the mouse in the same hand by pinning the tail between little finger and palm. Carefully introduce the needle (with your right hand if right-handed) into the mouth. Slide the tip past the back of the tongue. Gently push the head and nose back using the needle as a lever. The needle should easily slide down the upper esophagus. If any resistance is met, remove the needle and reinsert. Once the needle is properly placed, administer the solution. In contrast to intragastric administration, which requires a lot of care and practice to insert the needle into the stomach, this procedure does not require insertion into the esophagus and therefore much easier. To gain experience, mark the needle and aim to insert approximately 1 cm. Following this procedure, common complications associated with gastric intubation such as damage to the esophagus and administration of substance into the trachea are very unlikely. An alternative to oral gavage is intraperitoneal injection of the tamoxifen solution. Ensure that the needle has passed through the skin into the peritoneal cavity without entering an organ or blood vessel. We prefer gavage because the oil accumulates in the peritoneum after intraperitoneal injection. For tamoxifen induction in cultured cells, use 4-OH tamoxifen (Sigma, H7904). Make the stock at 1 mM in absolute ethanol (5 mg in 12.9 ml EtOH) and store aliquots at 20  C. The concentration on cells is 10 7 M. To avoid tamoxifen precipitation, it is a good idea to pipette from the stock into a volume of cell culture media with swirling, and then add this to the cell culture plate. Do not exceed 10 6 M on the cultured cells. We have evidence that there is a low affinity receptor for 4-OH tamoxifen, which is not the endogenous estrogen receptor and can cause cell death at 10 6 M, depending on the cell type. Replace the culture media with fresh tamoxifen after two population doublings if continuous activation of CreERT2 is required.

7. Problems Toxicity of Cre recombinase has been reported by several labs (Loonstra et al., 2001; Naiche and Papaioannou, 2007; Schmidt et al., 2000; Schmidt-Supprian and Rajewsky, 2007). On the other hand, the large number of highly successful applications, together with the robust nature of most transgenic Cre mice in the Cre zoo, indicates that the problems are not common and can be managed. Because Cre can be toxic, it is essential to plan experiments with exacting controls. We recommend the use of heterozygous conditionals as controls (Glaser et al., 2009). That is, a littermate in which the same Cre transgene is present and conditional recombination has occurred but only on one allele with the other allele wild type. In cases when this is not possible, for example

Site-Specific Recombination, Conditional Mutagenesis, and Tamoxifen Induction

119

conditional mutagenesis on the X chromosome in males or heterozygous phenotypes, the control animals must include the same Cre transgene. Although largely anecdotal, all evidence of Cre toxicity relates to Cre expression level. Overexpression of Cre appears to promote unwanted genome rearrangement due, at least in part, to recombination between cyptic lox-like sites (Higashi et al., 2009). Whether toxicity is also a problem provoked by Flp or Dre recombinases is not known. Tamoxifen also causes several problems, largely due to the fact that it blocks the action of endogenous estrogen receptors. Hence tamoxifen administration to mice causes transient male and female infertilities (lasting about one month after a standard course of CreERT2 tamoxifen induction, i.e., 4 or 5 daily doses of 4–5 mg). With the ROSA26 CreERT2 mice, an intestinal problem has also been noted. After tamoxifen administration to homozygous ROSA26 CreERT2 animals, intestinal peristalsis stops, the gall bladder swells with apparent blockage of the bile duct, and wasting with diarrhea in most animals. This problem occurs only rarely in heterozygous animals and does not depend upon introduced loxP sites in the genome (Glaser et al., 2009). Clearly, a ligand more neutral than tamoxifen would be an improvement.

8. Concluding Remarks The tyrosine site-specific recombinases have proven to be remarkable tools for many aspects of mouse experimentation, in particular mouse models of human diseases and conditional mutagenesis. However, these tools have several limitations. Knowledge of their strengths and weaknesses is vital for optimal application of these powerful instruments.

ACKNOWLEDGMENTS The work described here was funded by the Deutches Forschung Gemeinschaft grant to the Center for Regenerative Therapies, Dresden and the EU 6th Framework Program, EUCOMM (European Conditional Mouse Mutagenesis Consortium www.eucomm.org; now part of IKMC).

REFERENCES Albert, H., Dale, E. C., Lee, E., and Ow, D. W. (1995). Site-specific integration of DNA into wild-type and mutant lox sites placed in the plant genome. Plant J. 7, 649–659. Anastassiadis, K., Fu, J., Patsch, C., Hu, S., Weidlich, S., Duerschke, K., Buchholz, F., Edenhofer, F., and Stewart, A. F. (2009). Dre recombinase, like Cre, is a highly efficient

120

Konstantinos Anastassiadis et al.

site specific recombinase in E. coli, mammalian cells and mice. Dis. Model Mech. 2, 508–515. Awatramani, R., Soriano, P., Rodriguez, C., Mai, J. J., and Dymecki, S. M. (2003). Cryptic boundaries in roof plate and choroid plexus identified by intersectional gene activation. Nat. Genet. 35, 70–75. Badea, T. C., Hua, Z. L., Smallwood, P. M., Williams, J., Rotolo, T., Ye, X., and Nathans, J. (2009). New mouse lines for the analysis of neuronal morphology using CreER(T)/loxP-directed sparse labeling. PLoS ONE 16, e7859. Baer, A., and Bode, J. (2001). Coping with kinetic and thermodynamic barriers: RMCE, an efficient strategy for the targeted integration of transgenes. Curr. Opin. Biotechnol. 12, 473–480. Branda, C. S., and Dymecki, S. M. (2004). Talking about a revolution: The impact of site-specific recombinases on genetic analyses in mice. Dev. Cell 6, 7–28. Buchholz, F., Ringrose, L., Angrand, P.-O., Rossi, F., and Stewart, A. F. (1996). Different thermostabilities of FLP and Cre recombinases: Implications for applied site specific recombination. Nucleic Acids Res. 24, 4256–4262. Buchholz, F., Angrand, P.-O., and Stewart, A. F. (1998). Improved properties of FLP recombinase evolved by cycling mutagenesis. Nat. Biotechnol. 16, 657–662. Casanova, E., Fehsenfeld, S., Lemberger, T., Shimshek, D. R., Sprengel, R., and Mantamadiotis, T. (2002). ER-based double iCre fusion protein allows partial recombination in forebrain. Genesis 34, 208–214. Chen, Y., and Rice, P. A. (2003). New insight into site-specific recombination from FLP recombinase-DNA structures. Annu. Rev. Biophys. Biomol. Struct. 32, 135–159. Copeland, N. G., Jenkins, N. A., and Court, D. L. (2001). Recombineering: A powerful new tool for mouse functional genomics. Nat. Rev. Genet. 2, 769–779. Danielian, P. S., Muccino, D., Rowitch, D. H., Michael, S. K., and McMahon, A. P. (1998). Modification of gene activity in mouse embryos in utero by a tamoxifeninducible form of Cre recombinase. Curr. Biol. 8, 1323–1326. Ennifar, E., Meyer, J. E., Buchholz, F., Stewart, A. F., and Suck, D. (2003). Crystal structure of a wild-type Cre recombinase-loxP synapse reveals a novel spacer conformation suggesting an alternative mechanism for DNA cleavage activation. Nucleic Acids Res. 31, 5449–5460. Feil, R., Brocard, J., Mascrez, B., LeMeur, M., Metzger, D., and Chambon, P. (1996). Ligand-activated site-specific recombination in mice. Proc. Natl. Acad. Sci. USA 93, 10887–10890. Feil, R., Wagner, J., Metzger, D., and Chambon, P. (1997). Regulation of Cre recombinase activity by mutated estrogen receptor ligand-binding domains. Biochem. Biophys. Res. Commun. 237, 752–757. Ghosh, K., Guo, F., and Van Duyne, G. D. (2007). Synapsis of loxP sites by Cre recombinase. J. Biol. Chem. 282, 24004–24016. Glaser, S., Anastassiadis, K., and Stewart, A. F. (2005). Current issues in mouse genome engineering. Nat. Genet. 37, 1187–1193. Glaser, S., Lubitz, S., Loveland, K. L., Ohbo, K., Robb, L., Schwenk, F., Seibler, J., Ro¨llig, D., Kranz, A., Anastassiadis, K., and Stewart, A. F. (2009). The histone 3 lysine 4 methyltransferase, Mll2, is only required briefly during development and spermatogenesis. Epigenetics Chromatin 2, 5. Grindley, N. D., Whiteson, K. L., and Rice, P. A. (2006). Mechanisms of site-specific recombination. Annu. Rev. Biochem. 75, 567–605. Hameyer, D., Loonstra, A., Eshkind, L., Schmitt, S., Antunes, C., Groen, A., Bindels, E., Jonkers, J., Krimpenfort, P., Meuwissen, R., Rijswijk, L., Bex, A., Berns, A., and Bockamp, E. (2007). Toxicity of ligand-dependent Cre recombinases and generation

Site-Specific Recombination, Conditional Mutagenesis, and Tamoxifen Induction

121

of a conditional Cre deleter mouse allowing mosaic recombination in peripheral tissues. Physiol. Genomics 31, 32–41. He´rault, Y., Rassoulzadegan, M., Cuzin, F., and Duboule, D. (1998). Engineering chromosomes in mice through targeted meiotic recombination (TAMERE). Nat. Genet. 20, 381–384. Higashi, A. Y., Ikawa, T., Muramatsu, M., Economides, A. N., Niwa, A., Okuda, T., Murphy, A. J., Rojas, J., Heike, T., Nakahata, T., Kawamoto, H., Kita, T., et al. (2009). Direct hematological toxicity and illegitimate chromosomal recombination caused by the systemic activation of CreERT2. J. Immunol. 182, 5633–5640. Hoess, R. H., Wierzbicki, A., and Abremski, K. (1986). The role of the loxP spacer region in P1 site-specific recombination. Nucleic Acids Res. 14, 2287–2300. Kaul, A., Ko¨ster, M., Neuhaus, H., and Braun, T. (2000). Myf-5 revisited: loss of early myotome formation does not lead to a rib phenotype in homozygous Myf-5 mutant mice. Cell 102, 17–19. Kellendonk, C., Tronche, F., Monaghan, P.-A., Angrand, P.-O., Stewart, A. F., and Schu¨tz, G. (1996). Regulation of Cre recombinase activity by the synthetic steroid RU 486. Nucleic Acids Res. 24, 1404–1411. Kim, J. C., and Dymecki, S. M. (2009). Genetic fate-mapping approaches: new means to explore the embryonic origins of the cochlear nucleus. Methods Mol. Biol. 493, 65–85. Kolb, A. F. (2001). Selection-marker-free modification of the murine beta-casein gene using a lox2272 [correction of lox2722] site. Anal. Biochem. 290, 260–271. Kranz, A., Fu, J., Duerschke, K., Weidlich, S., Naumann, R., Anastassiadis, K., and Stewart, A. F. (2010). An improved Flp deleter mouse in C57Bl/6 based on Flpo recombinase. Genesis [Epub ahead of print]. Ku¨hn, R., Schwenk, F., Aguet, M., and Rajewsky, K. (1995). Inducible gene targeting in mice. Science 269, 1427–1429. Lee, G., and Saito, I. (1998). Role of nucleotide sequences of loxP spacer region in Cre-mediated recombination. Gene 216, 55–65. Logie, C., and Stewart, A. F. (1995). Ligand-regulated site-specific recombination. Proc. Natl. Acad. Sci. USA 92, 5940–5944. Loonstra, A., Vooijs, M., Beverloo, H. B., Allak, B. A., van Drunen, E., Kanaar, R., Berns, A., and Jonkers, J. (2001). Growth inhibition and DNA damage induced by Cre recombinase in mammalian cells. Proc. Natl. Acad. Sci. USA 98, 9209–9214. Muyrers, J. P., Zhang, Y., Testa, G., and Stewart, A. F. (1999). Rapid modification of bacterial artificial chromosomes by ET recombination. Nucleic Acids Res. 27, 1555–1557. Nagy, A., Moens, C., Ivanyi, E., Pawling, J., Gertsenstein, M., Hadjantonakis, A. K., Pirity, M., and Rossant, J. (1998). Dissecting the role of N-myc in development using a single targeting vector to generate a series of alleles. Curr. Biol. 8, 661–664. Naiche, L. A., and Papaioannou, V. E. (2007). Cre activity causes widespread apoptosis and lethal anemia during embryonic development. Genesis 45, 768–775. Pettitt, S. J., Liang, Q., Rairdan, X. Y., Moran, J. L., Prosser, H. M., Beier, D. R., Lloyd, K. C., Bradley, A., and Skarnes, W. C. (2009). Agouti C57BL/6N embryonic stem cells for mouse genetic resources. Nat. Methods 6, 493–495. Rajewsky, K., Gu, H., Ku¨hn, R., Betz, U. A., Mu¨ller, W., Roes, J., and Schwenk, F. (1996). Conditional gene targeting. J. Clin. Invest. 98, 600–603. Raymond, C. S., and Soriano, P. (2007). High-efficiency FLP and PhiC31 site-specific recombination in mammalian cells. PLoS ONE 2, e162. Ringrose, L. H., Lounas, V., Ehrlich, L., Buchholz, F., Wade, R., and Stewart, A. F. (1998). Comparative kinetic analysis of FLP and Cre recombinases: Mathematical models for DNA binding and recombination. J. Mol. Biol. 284, 363–384.

122

Konstantinos Anastassiadis et al.

Ringrose, L., Chabanis, S., Angrand, P.-O., Woodroofe, C., and Stewart, A. F. (1999). Quantitative comparison of DNA looping in vitro and in vivo: Chromatin increases effective DNA flexibility at short distances. EMBO J. 23, 6630–6641. Rodriguez, C. I., Buchholz, F., Galloway, J., Sequerra, R., Kasper, J., Ayala, R., Stewart, A. F., and Dymecki, S. M. (2000). High efficiency FLPe deleter mice provide a complement to Cre-loxP for in vivo genetic engineering. Nat. Genet. 25, 139–140. Sauer, B., and McDermott, J. (2004). DNA recombination with a heterospecific Cre homolog identified from comparison of the pac-c1 regions of P1-related phages. Nucleic Acids Res. 32, 6086–6095. Schaft, J., Ashery-Padan, R., van der Hoeven, F., Gruss, P., and Stewart, A. F. (2001). Efficient FLP recombination in mouse ES cells and oocytes. Genesis 31, 6–10. Schlake, T., and Bode, J. (1994). Use of mutated FLP recognition target (FRT) sites for the exchange of expression cassettes at defined chromosomal loci. Biochemistry 33, 12746–12751. Schmidt, E. E., Taylor, D. S., Prigge, J. R., Barnett, S., and Capecchi, M. R. (2000). Illegitimate Cre-dependent chromosome rearrangements in transgenic mouse spermatids. Proc. Natl. Acad. Sci. USA 97, 13702–13707. Schmidt-Supprian, M., and Rajewsky, K. (2007). Vagaries of conditional gene targeting. Nat. Immunol. 8, 665–668. Schnu¨tgen, F., Doerflinger, N., Calle´ja, C., Wendling, O., Chambon, P., and Ghyselinck, N. B. (2003). A directional strategy for monitoring Cre-mediated recombination at the cellular level in the mouse. Nat. Biotechnol. 21, 562–565. Schnu¨tgen, F., De-Zolt, S., Van Sloun, P., Hollatz, M., Floss, T., Hansen, J., Altschmied, J., Seisenberger, C., Ghyselinck, N. B., Ruiz, P., Chambon, P., Wurst, W., and von Melchner, H. (2005). Genomewide production of multipurpose alleles for the functional analysis of the mouse genome. Proc. Natl. Acad. Sci. USA 102, 7221–7226. Schnu¨tgen, F., Stewart, A. F., von Melchner, H., and Anastassiadis, K. (2006). Engineering embryonic stem cells with recombinase systems. Methods Enzymol. 420, 100–136. Scho¨nig, K., Schwenk, F., Rajewsky, K., and Bujard, H. (2002). Stringent doxycycline dependent control of CRE recombinase in vivo. Nucleic Acids Res. 30, e134. Schwenk, F., Ku¨hn, R., Angrand, P.-O., Rajewsky, K., and Stewart, A. F. (1998). Temporally and spatially regulated somatic mutagenesis in mice. Nucleic Acids Res. 26, 1427–1432. Seibler, J., Zevnik, B., Ku¨ter-Luks, B., Andreas, S., Kern, H., Hennek, T., Rode, A., Heimann, C., Faust, N., Kauselmann, G., Schoor, M., Jaenisch, R., et al. (2003). Rapid generation of inducible mouse mutants. Nucleic Acids Res. 31, e12. Senecoff, J. F., Rossmeissl, P. J., and Cox, M. M. (1988). DNA recognition by the FLP recombinase of the yeast 2 mu plasmid. A mutational analysis of the FLP binding site. J. Mol. Biol. 201, 405–421. Shimshek, D. R., Kim, J., Hu¨bner, M. R., Spergel, D. J., Buchholz, F., Casanova, E., Stewart, A. F., Seeburg, P. H., and Sprengel, R. (2002). Codon-improved Cre recombinase (iCre) expression in the mouse. Genesis 32, 19–26. Smith, M. C., and Thorpe, H. M. (2002). Diversity in the serine recombinases. Mol. Microbiol. 44, 299–307. Testa, G., Zhang, Y., Vintersten, K., van der Hoeven, F., Benes, V., Chambers, I., Smith, A. J. H., Smith, A. A., and Stewart, A. F. (2003). Engineering the mouse genome with bacterial artificial chromosomes to create multi-purpose alleles. Nat. Biotechnol. 21, 443–447. Testa, G., Schaft, J., van der Hoeven, F., Glaser, S., Anastassiadis, K., Zhang, Y., Hermann, T., Stremmel, W., and Stewart, A. F. (2004). A reliable expression reporter cassette for multipurpose, knock-out/conditional mouse alleles. Genesis 38, 151–158.

Site-Specific Recombination, Conditional Mutagenesis, and Tamoxifen Induction

123

Thomson, J. G., Rucker, E. B., III, and Piedrahida, A. (2003). Mutational analysis of loxP sites for efficient Cre-mediated insertion into genomic DNA. Genesis 36, 162–167. van der Weyden, L., Shaw-Smith, C., and Bradley, A. (2009). Chromosome engineering in ES cells. Methods Mol. Biol. 530, 49–77. van Duyne, G. D. (2001). A structural view of Cre-loxP site-specific recombination. Annu. Rev. Biophys. Biomol. Struct. 30, 87–104. Vintersten, K., Testa, G., Naumann, R., Anastassiadis, K., and Stewart, A. F. (2008). Bacterial artificial chromosome transgenesis through pronuclear injection of fertilized mouse oocytes. Methods. Mol. Biol. 415, 83–100. Wallace, H. A., Marques-Kranc, F., Richardson, M., Luna-Crespo, F., Sharpe, J. A., Hughes, J., Wood, W. G., Higgs, D. R., and Smith, A. J. (2007). Manipulating the mouse genome to engineer precise functional syntenic replacements with human sequence. Cell 128, 197–209. Wu, S., Ying, G., Wu, Q., and Capecchi, M. R. (2007). Toward simpler and faster genomewide mutagenesis in mice. Nat. Genet. 39, 922–930. Wu, Y., Wang, C., Sun, H., LeRoith, D., and Yakar, S. (2009). High-efficient FLPo deleter mice in C57BL/6J background. PLoS ONE 4, e8054. Yu, Y., and Bradley, A. (2001). Engineering chromosomal rearrangements in mice. Nat. Rev. Genet. 2, 780–790. Zhang, Y., Buchholz, F., Muyrers, J. P., and Stewart, A. F. (1998). A new logic for DNA engineering using recombination in Escherichia coli. Nat. Genet. 20, 123–128.

C H A P T E R

E I G H T

A Recombineering Pipeline to Make Conditional Targeting Constructs Jun Fu,* Madeleine Teucher,* Konstantinos Anastassiadis,† William Skarnes,‡ and A. Francis Stewart* Contents 1. Introduction 2. Recombineering 3. Methods 3.1. Allele design and in silico work on the targeting construct 3.2. Generate the targeting construct in silico 3.3. Ordering mouse BACs from the genome resources 3.4. Generating the PCR products for recombineering 3.5. PCR purification and yield 3.6. Preparation of the lacZ-neo cassette 3.7. Building the targeting construct by recombineering 3.8. Plating instead of liquid processing 4. Standard Recombineering Electroporation Protocol 5. Concluding Remarks Acknowledgments References

126 126 127 127 131 132 132 135 136 136 139 140 142 142 142

Abstract Here we describe an optimized recombineering protocol for the creation of conditional alleles. Recombineering is a DNA engineering methodology based on homologous recombination in Escherichia coli. It presents a number of advantages over conventional engineering methods, including the ability to place target sites for site-specific recombinases at exactly chosen positions. Consequently, complicated designs, such as conditional alleles, can be rapidly achieved. We have applied these advantages to establish a pipeline method for the fluent generation of conditional alleles which employs Cre, Flp, and Dre recombinase target sites. * Genomics, BioInnovationsZentrum, Technische Universitaet Dresden, Dresden, Germany Center for Regenerative Therapy, Technische Universitaet Dresden, Dresden, Germany Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, England

{ {

Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77008-7

#

2010 Elsevier Inc. All rights reserved.

125

126

Jun Fu et al.

1. Introduction The power and utility of the mouse as a model system has been greatly enhanced by progress in the technology of DNA recombination. Starting from the first transgenic approaches, which used plasmid DNA constructs and illegitimate recombination for genomic integration, the manipulation of the mouse genome has become more sophisticated at each of the three stages of the process: (i) the design and construction of recombinant DNA molecules has become significantly more sophisticated, particularly due to the development of recombineering, which is the use of phage recombinase-mediated homologous recombination in Escherichia coli (Copeland et al., 2001; Muyrers et al., 1999; Zhang et al., 1998); (ii) the integration of recombinant DNA into living genomes no longer solely relies upon illegitimate recombination but can be mediated by homologous (gene targeting), site-specific (RMCE; recombinase-mediated cassette exchange) or transpositional (piggyBac, Sleeping Beauty) recombination (Glaser et al., 2005; Ivics et al., 2009); (iii) after the recombinant DNA has been integrated into the genome, its structure can be altered by site-specific recombination (SSR) for conditional mutagenesis or to establish new alleles (Branda and Dymecki, 2004; Glaser et al., 2005). With these options, it is now possible to design complex alleles that incorporate various components and aspects. Recombineering facilitates complex allele design because the steps involved in building the recombinant DNA can be readily achieved. Consequently, allele design is now conceptual—virtually any design is achievable. This is particularly evident for SSR applications. These remarkable enzymes, including Cre, Flp, and more recently Dre recombinase (Anastassiadis et al., 2009; Sauer and McDermott, 2004), mediate recombination through specific
32 bp recognition targets (RTs). The deployment of the RTs dictates the mode of SSR. Recombineering permits the exact positioning of RTs in the recombinant DNA construct according to design (Zhang et al., 1998). Consequently, there is a synergy between recombineering and SSR applications.

2. Recombineering The original methods for the production of recombinant DNA, complimented by oligonucleotide synthesis and PCR, utilize restriction enzymes and DNA ligase to ‘‘cut-and-paste’’ DNA. These methods all suffer from a size limitation. It becomes increasingly difficult to achieve a precise product as the DNA engineering tasks becomes larger. To address this problem, we initially tried a popular method for homologous

A Recombineering Pipeline to Make Conditional Targeting Constructs

127

recombination in E. coli (Buchholz et al., 1996; Hamilton et al., 1989; Yang et al., 1997). Although this method has been developed for DNA engineering with some success, we found it to be laborious and inefficient. Subsequently, we found that phage recombinase systems from the rac prophage and the closely related l-phage Red operon mediate extremely efficient homologous recombination that requires only short homology arms (Muyrers et al., 1999; Zhang et al., 1998). These findings initiated the DNA engineering methodology termed ‘‘recombineering’’ (Copeland et al., 2001; Court et al., 2002; Datsenko and Wanner, 2000; Yu et al., 2000), which now includes a broad range of applications, including subcloning by gap repair (Zhang et al., 2000), oligonucleotide-directed mutagenesis (Ellis et al., 2001), BAC engineering for gene targeting (Testa et al., 2003; Valenzuela et al., 2003; Wu et al., 2008; Yang and Seed, 2003), highthroughput DNA engineering (Poser et al., 2008; Sarov et al., 2006; Skarnes et al., 2010), and a variety of other, often complex, applications. Here we do not aim to present a broad primer into recombineering methodologies, which is beyond the scope of this book (the interested reader is referred to Sawitzke et al., 2007). Rather, we focus upon a method for fluent generation of targeting constructs for conditional alleles. Nevertheless, this protocol is composed of steps that serve as good examples for several recombineering applications, such as BAC modification, subcloning from a BAC, and the use of counterselection. Consequently, these methods and associated reagents have utility for other recombineering exercises.

3. Methods 3.1. Allele design and in silico work on the targeting construct The allele design described here is based on the ‘‘knock-out first’’ strategy (Testa et al., 2004), which is also the basic design being used for the International Knock-out Mouse Consortium (IKMC; www.ikmc.org). In brief, after selection of a frameshifting exon in the target gene, a genetrap-style stop cassette flanked by FRT sites and a loxP site is placed in the intron on the 50 side and a second loxP site is placed in the intron on the 30 side. The general scheme is presented in Fig. 8.1. The work described here includes a variety of improvements over Testa et al. (2004). 1. Exon selection. Ideally, an exon whose removal from the mRNA provokes a frameshift is selected. These exons start and finish in different reading frames, which are annotated as either frames 0, 1, or 2. Hence, a frameshifting (or ‘‘critical’’) exon will be ‘‘0,1’’ or ‘‘2,0,’’ etc., and not ‘‘0,0,’’ ‘‘1,1,’’ or ‘‘2,2’’. The 50 most frameshifting exon is selected unless the flanking introns are too small. The flanking introns need to be at least

128

A

Jun Fu et al.

5'HA

3'HA

Critical exon

BAC

rpsL Genta

1.1 kb PCR

B BAC

rpsL Genta

p15A pA DTA pTK amp

3.3 kb PCR R2

R1

C p15A

rpsL Genta

6.1 kb fragment released

sA lacZ neo Stop from R6Ky ori plasmid T2A FRT T2A FRT loxP

D sA

lacZ

p15A

neo Stop pA BSD PGK 1.4 kb PCR rox rox loxP

R1

E

R2 p15A pA DTA pTK Amp

1st allele targeting construct sA

lacZ

FRT T2A

neo Stop

T2A

FRTloxP

pA BSD PGK rox loxP

rox

Figure 8.1 Diagram of the construction scheme for the conditional targeting construct. In (A), the rpsL-gentamycin selectable/counterselectable cassette is inserted in the intron 50 to the chosen frameshifting (‘‘critical’’) exon. In (B), a region of about 10 kb around the critical exon is subcloned into the p15A-pTK-DTA-amp vector by gap repair. In (C), the rpsL-gentamycin cassette is replaced by the lacZ-neo stop cassette. In (D), the PGK-BSD cassette is inserted into the intron on the 30 side of the critical exon, to complete the targeting construct, as illustrated in (E).

150 bp long so that the cassette insertion site is at least 50 bp from the 50 splice site (i.e., the 50 end of the intron) and 100 bp from the 30 splice site. Also, 50 bp upstream and downstream of the cassette insertion site must be sequence that is unique within the parental BAC (i.e., within about 150 kb of the chosen exon). If the flanking introns are too small to permit these choices, then either look for another frameshifting exon or

A Recombineering Pipeline to Make Conditional Targeting Constructs

129

use the neighboring intron (i.e., jump the difficult intron and use the next intron). In this case, (i) make sure that the removal of the two exons together will still provoke a frameshift and (ii) do not include more than about 2 kb between the two cassette insertion sites because the efficiency of Cre-recombinase-mediated conditional mutagenesis diminishes with increasing distance between the loxP sites (Ringrose et al., 1999). Optimally, we aim to place the two loxP sites less than 1 kb apart. The attentive reader will realize that this design is not applicable to single exon genes. For single intron genes, we place the 30 loxP site into the 30 noncoding region or downstream of the polyadenylation signal if possible. For genes that do not present a useable frameshifting exon, selecting an exon that encodes an essential protein function or the last exon can be reasonable compromises, albeit with the risk of generating hypomorphic or dominant negative mutations rather than a knock-out. 2. The lacZ-neo stop cassette (Fig. 8.1) has been altered in several ways compared to our previous version (Testa et al., 2004). The original stop cassette included the Engrailed-2 splice acceptor to capture the transcript of the target gene, followed by an internal ribosome entry site (IRES) promoting the expression of a b-galactosidase–neomycin resistance fusion protein and then a 30 noncoding region ending in a polyadenylation signal (Chen and Soriano, 2003). This cassette was flanked by FRTs with a loxP site on the 30 end. In the new version presented here, the splice acceptor has been shortened and the IRES has been replaced by a 2A ribosomal-error peptide (Szymczak et al., 2004). These two changes make the cassette shorter and remove the IRES, which was occasionally unpredictable. The 2A peptide provokes the ribosome into skipping a peptide bond, thereby producing two polypeptides from one open reading frame. However, in the absence of an IRES, the cassette must be in frame with the coding region of the targeted gene (i.e., in frame with the exon upstream of the chosen exon). Consequently, the stop cassette has been made in three versions differing according to the three different reading frames of the 2A–lacZ–neo coding region. During design, the correct version needs to be selected according to the 0, 1, or 2 reading frame of the upstream exon. The new stop cassette also differs at the lacZ–neo junction. Previously, the two protein reading frames were fused, connected by a region that encoded the gb3 E. coli promoter for selection of kanamycin resistance during recombineering. Now a 2A peptide is included so that b-galactosidase and neomycin resistances are expressed as separate proteins (because the performance of both proteins is enhanced when they are separate rather than fused). 3. The 30 loxP site. The 30 loxP site is introduced into the 30 intron alongside a cassette that includes the PGK promoter driving the expression of the blasticidin resistance gene (BSD, Fig. 8.1). Between the PGK promoter and BSD gene is an E. coli promoter for blasticidin selection during

130

Jun Fu et al.

recombineering. This cassette is flanked by rox sites (the recombination target sites for Dre recombinase; Sauer and McDermott, 2004). There are two reasons for this design. A. For genes that are expressed in ES cells and can be targeted by selection of neomycin/G418 resistance driven from the endogenous target gene promoter (termed ‘‘promoterless selection’’), the rox cassette should be removed in E. coli by expressing Dre recombinase from pSC101-BAD-Dre (Anastassiadis et al., 2009) to leave the 30 loxP site in the targeting construct. Hence, the rox cassette serves only as a convenient way to insert the 30 loxP site. However, in a few cases after targeting in ES cells, we have found that all targeted clones omit the 30 loxP site. That is, the stop cassette has been targeted to the correct site; however, the 30 loxP site has been excluded. In these cases, the targeting can be repeated using the construct, including the rox cassette, and selecting for both neomycin and blasticidin resistances to force the inclusion of the 30 loxP site. B. For genes that are not expressed in ES cells, the targeting construct must include a promoter to drive expression of the selectable gene (termed ‘‘promoter-driven selection’’). Here the rox cassette provides this function. In cases where it is not clear whether the endogenous gene is sufficiently expressed to drive selection, experiments selecting for G418, blasticidin or both can be tried in parallel. In the cases when the rox cassette remains after targeting, it must be removed either by expressing Dre in ES cells or by crossing to a Dre deleter (Anastassiadis et al., 2009). 4. The subcloning vector, p15A-pTK-DTA-ampR (Fig. 8.1), is based on the p15A plasmid origin. Consequently, it does not give the very high copy yields delivered by the mutated colE1 origins in pUC-type plasmids. We have found that these unnaturally very high copy plasmids can provoke recombineering problems so prefer p15A or the unmutated colE1 origin from pBR322 for recombineering applications, particularly subcloning by gap repair. In p15A-pTK-DTA-ampR, the diptheria toxin A chain (DTA) coding region under the HSVtk promoter lies between the p15A origin and the ampicillin resistance gene. For subcloning, the vector is amplified by PCR to attach the homology arms. The presence of a DTA gene permits the use of positive/negative selection (Mansour et al., 1988; Yagi et al., 1990). We have found that negative selection can benefit promoter-driven selection, but has little effect on promoterless selection. Therefore, we recommend that the vector be entirely cut off for promoterless selection or linearized for promoter-driven selection. For promoterless targeting, the final targeting construct should be cut with SgrA1 to release the plasmid backbone entirely (unless the targeting construct unluckily includes an SgrA1 site, in which case you should select a new unique cutter and include these

A Recombineering Pipeline to Make Conditional Targeting Constructs

131

sites in the oligonucleotides used for PCR amplification). For promoterdriven targeting, restriction sites have to be included enabling the vector to be linearized on one side of the DTA gene or the other.

3.2. Generate the targeting construct in silico The first step towards building the targeting construct involves designing the entire recombineering procedure in silico using a DNA cloning program (e.g., Gene Construction Kit, Vector NTI). The points that need to be considered are: (i) Centered on the chosen exon, enter about 30 kb of sequence into your DNA management program. (ii) To insert the rpsL-gen cassette (Fig. 8.1) into the intron on the 50 side of the chosen exon, select a 100nt long sequence upstream of the 50 end of the exon so that the 30 end of this sequence is at least 50 bp from the 50 end of the exon. Check that the selected 100nt sequence does not include repetitive sequences by inspection and also BLASTing against the DNA sequence of the BAC containing your target gene. Divide the selected 100nt sequence in half. The 50 50nt of this sequence will be included in an oligonucleotide that will become, via PCR of the pR6K-rpsL-gen counterselection template, the 50 homology arm for the recombineering step. The inverse compliment to the 30 50nts will be included in an oligonucleotide that will become, via the same PCR reaction, the 30 homology arm. (iii) The selection of the site and homology arms to insert the PGK-BSD cassette carrying a loxP site (Fig. 8.1) on the 30 side of the chosen exon follows essentially the same procedure as detailed in (ii) except that the 100nt can start at the 30 end of the chosen exon. Again select 100nt without repetitive sequence. The 50 50nt and the inverse compliment of the 30 50nt will be included in the oligonucleotides that serve as PCR primers on the pR6K-PGK-BSD template. (iv) The backbone of the targeting construct is obtained by subcloning an approximately 10 kb genomic region around the chosen exon. Inspect the genomic regions between 4 and 6 kb upstream and downstream of the chosen exon for repeated sequences. The subcloned region should end with unique sequence so avoid the inclusion of repeats at or close to the ends. The outermost 50nt at each end of the subcloned region defines the homology arms for subcloning by recombineering (gap repair). These regions are included in oligonucleotides used to PCR the subcloning vector, p15A-pTK-DTA-ampR. As illustrated in Fig. 8.1, please note that for subcloning by gap repair, the sequences included in the oligonucleotides should be the inverse compliment to the orientations described in (ii) and (iii), which are for insertions.

132

Jun Fu et al.

(v) Inspect the restriction map for about 15 kb either side of the chosen exon to design a Southern strategy for use after targeting in ES cells that will discriminate between homologous and random integration. Avoid restriction sites with CpG dinucleotides because they may be partially or fully methylated, which can lead to incomplete restriction digestion. The Southern probes should be made from unique regions of at least 200 bp, ideally about 400 bp, lying between the chosen external genomic restriction sites and the ends of the targeting construct. These regions can be easily obtained using the parent BAC as the template for PCR reactions. Further details for designing targeting constructs can be found at www. biotec.tu-dresden.de/research/stewart/group-page/genetargeting.

3.3. Ordering mouse BACs from the genome resources The availability of whole genomes as indexed BACs is a very convenient starting point for the construction of a targeting construct. The arrays of available BACs can be viewed on a genome browser like Ensembl (http:// www.ensembl.org/index.html). For the current Ensembl release, after finding the gene of interest on the ‘‘Location-based display—Region in detail’’ page, go to ‘‘Configure this page.’’ Then on the main panel page, select ‘‘External Data.’’ Click on ‘‘BAC map’’ and close the window. After the page renews, the region in detail will display the two C57BL/6J BAC libraries available from CHORI. Either library RP23 or RP24 is suitable. Select one or two BACs and follow the instructions for ordering from CHORI (http://bacpac.chori.org/order.php). For BACs made from 129 DNA, look under ‘‘other DNA alignments’’ and switch on ‘‘M37-129AB22.’’ This reveals the bMQ(129) library, which can be ordered from http://www.geneservice.co.uk/home/. Sometimes it is a good idea to select two BACs because about 5% of BACs that you receive will be incorrect (due to contamination, clerical, or recombinational errors). To retrieve the sequence of a chosen BAC from ENSEMBL, right click on the BAC to obtain its end points, and ‘‘export data’’ into your DNA management program.

3.4. Generating the PCR products for recombineering Synthetic oligonucleotides carrying the 50 bp recombineering homology arms at their 50 ends are attached to the standard cassettes by PCR. The PCR primer sequences for the standard cassettes are presented in Table 8.1. Sometimes it will be useful to include an extra sequence between the homology arm and the PCR primer in the oligonucleotide. For example,

A Recombineering Pipeline to Make Conditional Targeting Constructs

133

Table 8.1 Sequences of PCR primers for the cassettes shown in Fig. 8.1 rpsL-gen 50

rpsL-gen 30 p15A-tkDTAampR 50 p15A-tkDTAampR 30 PGK-BSD 50 PGK-BSD 30

GCGTGTTTCGAGCATGTTTCTGCGTAGTGTC AGCTCATCC

GTGTCCATCATCCTGTAGGTGTAGAC GACGACGAACAGAG TTAATAAGATGATCTTCTTGAGATCG TTACCAATGCTTAATCAGTGAGG TGACTACACCAAGGTGGAAGACAGAGAAAT TCGTCACAATAATCTTCCACCGATCGCTTC

All sequences are 50 –30 and should be preceded on their 50 ends by the specific homology arms.

this is a very convenient way to include another restriction enzyme site if needed. Care should be taken to ensure that the correct orientation and strand is included in the oligonucleotide design before ordering. Therefore in silico creation of the final construct vector map is recommended. Three PCR products are required to generate the first allele targeting construct (Fig. 8.1), hence three pairs of oligonucleotides need to be synthesized. The templates for the PCR reactions are the plasmids shown in Figure 8.2, the oligonucleotide primers are presented in Table 8.1 and the reaction conditions in Tables 8.2 and 8.3. 1. The first PCR product is amplified from the rpsL-gentamycin selectable/counterselectable cassette (Fig. 8.2). The required oligonucleotides contain the 50nt homology arms from the 50 intron selected in (ii) above and the PCR primers for the rpsL-gen cassette (Table 8.1). The PCR amplified rpsL-gen cassette already contains the homology arms for the stop cassette (Fig. 8.1, step 3). Consequently, incorporation of the rpsL-gen cassette into the BAC by selection for gentamycin resistance incorporates the homology arms for the stop cassette. We include this step because it avoids PCR amplification of the 6.1 kb stop cassette. Instead, a restriction fragment from a pR6K-amp-lacZ-neo plasmid preparation is used, thereby reducing unwanted, PCR-based, mutagenesis. 2. The second PCR product is the amplification of the pre-restricted subcloning vector, p15A-pTK-DTA-ampR. The oligonucleotides used for PCR amplification include the 50nt homology arms selected in (iv) above and the PCR primers for each end of p15A-pTK-DTAampR (Table 8.1). Note, as stated above and illustrated in Fig. 8.1, for subcloning the homology arms must be inversely orientated.

134

Jun Fu et al.

A

B

araC BAD-promotor redg

tetR

R6K ori redb

pSC101-BAD-gbaA-tet 7.4 kb

pR6K-rpsL-gen 1.8 kb

Tn10promotor

repA

rpsL

reda 150 50

recA

Genta

pSC101-ori Swal

C

D

Swal

R6K ori

CoTC synt pA SV40-pA

ampR

FRT loxP Swal SA

Swal ampR HSVtk promoter

FRT

p15A-ori

p15A-pTK-DTA-ampR 4.4 kb

pR6K-amp-lacZneo 7.9 kb

neo T2A gb3promotor

GT0/1/2

T2A lacZ

DTA 100 150

pA

E

PGK promotor

R6K-ori

loxP ROX ROX

pR6K-PGK-BSD 2.1 kb pA

50

Em7-promotor

bsdR

Figure 8.2 Diagrams of the plasmids used to build the targeting construct. A. The pSC101-BAD-gbaA-tet plasmid that is used to express the l Red gba proteins after L-arabinose induction. B. The pR6K-rpsL-gen plasmid that is used as a PCR template for the first step in Fig. 8.1. C. The p15A-pTK-DTA-ampR plasmid that is used after SwaI digestion as a PCR template for the second step in Fig. 8.1. D. The pR6K-amplacZ-neo plasmid that is digested with Swa1 to provide the restriction fragment for the third step in Fig. 8.1. E. The pR6K-PGK-BSD plasmid that is used as a PCR template for the fourth step in Fig. 8.1.

135

A Recombineering Pipeline to Make Conditional Targeting Constructs

Table 8.2 PCR reactions

Plasmid template

2 Phusion Flash (FINNZYME, includes buffer, dNTPs, polymerase) Oligo 50 (50 pmol/l) Oligo 30 (50 pmol/l) Template DMSO H2O to Total

pR6K-rpsL-gen or pR6K-PGK-BSD

p15A-pTKDTA-ampR

25 l

50 l

1 l 1 l 1 l of 50 ng/l 2.5 l 50 l

2 l 2 l 2 l of 3 ng/l 5 l 100 l

Conditions for rpsL-gen PCR and PGK-BSD PCR are the same. For rpsL-gen PCR and PGK-BSD PCR, simply use 50 ng from standard plasmid preparations as PCR template. For the p15A-pTK-DTAampR PCR, digest about 10 g plasmid preparation with SwaI, precipitate and dissolve in H2O to the final concentration of 3 ng/l as PCR template.

Table 8.3 Cycle settings for PCR reactions rpsL-gen or PGK-BSD 

98 C 98  C 54  C 72  C 72  C 10  C

1 min ) 1s 5s 2 min 1 min end

p15A-pTK-DTA-ampR

30

98  C 98  C 54  C 68  C 68  C 10  C

2 min ) 20 s 20 s 4 min 5 min end

30

3. The third PCR product is amplified from the pR6K-PGK-BSD template, which will insert a loxP site on the 30 side of the chosen exon.

3.5. PCR purification and yield Check 3 l of the PCR reaction on a 0.7% agarose gel. Purify the rest using Invitek MSB Spin PCRapace Kit for column purification and elute with 20 l H2O. Do not elute with buffer—use only H2O. The purification is necessary to get rid of unincorporated oligonucleotides which compete with recombination, and salt, which inhibits transformation. The yields should be more than 2 g at 100–300 ng/l. For rpsL-gen and PGK-BSD, about 300 ng is needed for one electroporation. For p15A-pTK-DTA-ampR, about 600 ng is needed for one electroporation.

136

Jun Fu et al.

3.6. Preparation of the lacZ-neo cassette Because the homology arms for integrating the lacZ-neo cassette are included in the rpsL-gen cassette, the lacZ-neo cassette does not need to be PCR amplified but should be prepared by SwaI restriction digestion of 20 g or more of the pR6K-amp-lacZ-neo plasmid. After checking for complete digestion, ethanol precipitate using one-tenth volume of 3 M NaAcetate and three volumes of ethanol. Rinse the pellet with ethanol, air dry, and resuspend the DNA in H2O at about 0.3 g/l. Two microliters will be used for electroporation.

3.7. Building the targeting construct by recombineering To capacitate the E. coli host for recombineering, the l-phage recombination proteins Red abg need to be expressed. However, it is advantageous to limit their expression to only the time interval required because this also limits the chances of unwanted recombination. Consequently, we employ the excellent properties of the arabinose inducible promoter, BAD, to tightly regulate Red expression (Zhang et al., 1998). Here the BAD abg operon is either present on the low copy plasmid, pSC101-BAD-abgA (Wang et al., 2006) or has been stably integrated into the genome (E. coli GB08-red, unpublished). The pSC101 plasmid is low copy (
5 per cell) and encodes a temperature sensitive replication protein so that it replicates at 30  C and not at 37  C (Hashimoto-Gotoh and Sekiguchi, 1977). Consequently, it can be easily eliminated from the host by temperature shift. Transforming pSC101-BAD-abgA into the host containing the BAC is a convenient way to start recombineering the BAC. However, the BAC is no longer needed once the plasmid subclone has been made. Also, the subclone needs to be retrieved by miniprep purification and transformation into a new host. Hence, we generated the E. coli strain GB08-red by integrating the BAD-abgA operon into the genome, so that the transformed subclone could be conveniently further recombineered. The procedure detailed below includes the time frame of the procedure as a further guide for the work involved. This is particularly important for the liquid selection steps because host cells carrying the intended (homologous) recombination product grow up more quickly than cells carrying the rare unintended, random recombination events. Hence, if cultures are harvested as early as possible they will be virtually always correct whereas later cultures will include rubbish. A second important point to appreciate is that the most common source of background resistance (i.e., cells expressing resistance to the antibiotic used for selection of the intended recombination event) is due to carry over of the parent plasmid in which the antibiotic resistance gene is cloned. In the early versions of recombineering methods, parental plasmids were eliminated by gel purification or restriction enzyme

A Recombineering Pipeline to Make Conditional Targeting Constructs

137

cleavage steps prior to electroporation of the selectable linear DNA fragment. Often, however, carryover background still occurred. We now routinely solve this problem by cloning all cassettes into R6K plasmids. R6K plasmids require a specific Rep1 protein called Pi to replicate. Hence R6K plasmids can be grown in pir þ E. coli hosts but will not replicate at all in normal E. coli cloning strains (Filutowicz et al., 1986). We use an unpublished pir þ E. coli strain, GB05-pir. Prior to beginning the protocol below, plate the E. coli host containing the BAC obtained from the BAC provider onto chloramphenicol (10 g/ ml) LB plates. Day 1 Inoculate three colonies of the BAC into a 1.5 ml reaction tube containing 1 ml LB with chloramphenicol (10 g/ml). Incubate for 20 h at 30  C with 950 rpm shaking on an Eppendorf Thermomixer. Set up duplicate tubes for each BAC. Day 2—Transformation of the recombineering plasmid Inoculate 30–40 l1 of the ON-culture into a new tube containing 1.4 ml LB with chloramphenicol (10 g/ml). Culture and make cells electrocompetent according to the standard protocol (see below). Electroporate 200 ng of pSC101-BAD-gbaA-tet plasmid DNA and allow the cells to recover at 30  C in 1 ml SOC for 2 h. Electroporate water as a negative control. After recovery, inoculate 100 l of the culture into a new tube containing 900 l LB with a final concentration of chloramphenicol (10 g/ml) and tetracycline (4 g/ml). Incubate at 30  C for 20 h. Day 3—Integration of the counterselection cassette Inoculate 30–40 l1 of the ON-culture into a new tube containing 1.4 ml LB with chloramphenicol (10 g/ml) and tetracycline (4 g/ml). Culture grow and make cells electrocompetent according to the standard protocol. Electroporate 200 ng of the rpsL-gen PCR product. Recover at 30  C in 1 ml SOC for 2 h. Omit the addition of L-arabinose to generate a negative control. After recovery, inoculate 100 l of the culture into a new tube containing 900 l LB with a final concentration of chloramphenicol (10 g/ml), tetracycline (4 g/ml), and gentamycin (1 g/ml). Incubate at 30  C for 20 h. Day 4—Subcloning by gap repair Inoculate 30–40 l1 of the ON-culture into a new tube containing 1.4 ml LB with chloramphenicol (10 g/ml), tetracycline (4 g/ml), and 1

30–40 l: Measure the OD600 (A) of the ON-culture. It should be between 3 and 4. Transfer X l of the ON-culture into the 1.4 ml LB for the final cell density of OD600 ¼ 0.085 according to this formula: X ¼ 1400 l  0.085/A.

138

Jun Fu et al.

gentamycin (1 g/ml). Culture and make cells electrocompetent according to the standard protocol. Electroporate 600 ng of the p15ApHSVtk-DTA-ampR PCR product. Recover at 37  C in 1 ml SOC for 1 h. After recovery inoculate 100 l of the culture into a new tube containing 900 l LB with a final concentration of ampicillin (50 g/ml) and gentamycin (1 g/ml). Incubate at 37  C for 20 h. In parallel, inoculate 3 colonies of E. coli GB08-red into 1 ml LB and incubate at 37  C for 16 h. Day 5—Separation of subclone products Isolate the plasmid products of the subcloning reaction from the ON-culture using a plasmid minipreparation kit with size exclusion columns (e.g., Invitek Invisorb Spin Plasmid Mini Two kit). Elute the plasmid DNA with 20 l water. Inoculate 30–40 l1 of the GB08-red ON-culture into a new tube containing 1.4 ml LB. Culture and make cells electrocompetent according to the standard protocol but omitting the L-arabinose induction. Electroporate 2 l of the plasmid DNA from the miniprep and recover at 37  C in 1 ml SOC for 1 h. After recovery, inoculate 100 l of the culture into a new tube containing 900 l LB plus ampicillin (50 g/ml) and gentamycin (1 g/ml). Incubate at 37  C for 20 h. Day 6—Replacement of the rpsL-gen cassette by the lacZ-neo cassette Inoculate 30–40 l1 of the ON-culture into a new tube containing 1.4 ml LB with ampicillin (50 g/ml) and gentamycin (1 g/ml). Culture and make cells electrocompetent according to the standard protocol. Electroporate 600 ng of the linearized lacZ-neo cassette. Recover at 37  C in 1 ml SOC for 1 h. After recovery, inoculate 100 l of the culture into a new tube containing 900 l LB with a final concentration of ampicillin (50 g/ml) and kanamycin (5 g/ml). Incubate at 37  C for 4 h then add streptomycin to a final concentration of 200 g/ml and incubate at 37  C for 16 h. Day 7—Insertion of the PGK-BSD cassette Inoculate 30–40 l1 of the ON-culture into a new tube containing 1.4 ml LB with ampicillin (50 g/ml) and kanamycin (5 g/ml). Culture and make cells electrocompetent according to the standard protocol. Electroporate 400 ng of the PGK-BSD PCR product. Recover at 37  C in 1 ml SOC for 1 h. After recovery, inoculate 100 l of the culture into a new tube containing 1800 l low-salt LB (pH 8.0) with a final concentration of ampicillin (50 g/ml) and blasticidin (40 g/ml). Incubate at 37  C for 16 h. In parallel, inoculate three colonies of E. coli GB05 into 1 ml LB and incubate at 37  C for 16 h. Day 8—Separation of final construct from the intermediates Prepare the plasmid DNA from the 1.9 ml overnight culture according to standard protocols and dissolve in 20 l water.

A Recombineering Pipeline to Make Conditional Targeting Constructs

139

Inoculate 30–40 l1 of the GB05 ON-culture into a new tube containing 1.4 ml LB. Culture and make cells electrocompetent according to the standard protocol. Electroporate 2 l of the plasmid DNA and recover at 37  C in 1 ml SOC for 1 h. After recovery, inoculate 100 l of the culture into a new tube containing 1800 l low-salt LB (pH 8.0) with a final concentration of ampicillin (50 g/ml), kanamycin (5 g/ ml), and blasticidin (30 g/ml). Incubate at 37  C for 16 h. Day 9—Analysis of the final construct and preparation of stocks Take 500 l of overnight culture plus 500 l 50% sterilized glycerol to make a bacterial stock for 80  C storage. Prepare the plasmid DNA from the remaining 1.4 ml ON-culture according to standard protocols. Dissolve the DNA in 16 l TE buffer. Use 4 l DNA for a restriction analysis. Keep the remaining 12 l as a mixed DNA stock. To evaluate the targeting construct, streak 10 l of the glycerol stock onto a low-salt LB plate with ampicillin (100 g/ml) and blasticidin (40 g/ml). Pick single colonies for restriction analysis and sequence verification of all oligonucleotide recombineering junctions using the primers listed in Table 8.4. Prepare a plasmid maxiprep for ES cell electroporation.

3.8. Plating instead of liquid processing The above procedure can be adjusted from continuous processing by liquid culture into stages by plating after any of the recovery steps. Plating will slow the procedure but permits selection of single colonies for subsequent culture and checking by restriction analysis to evaluate each step for the correct recombinants before the next step. However, we find that successful overnight growth in liquid is usually sufficient to indicate that the intended recombination step has occurred. Although the liquid culture may include some unwanted products, the next recombineering step usually serves to Table 8.4 Sequencing oligos Recombineering junction to be sequenced

Sequencing primer (50 ! 30 )

50 junction of subclone backbone 30 junction of subclone backbone 50 junction of lacZ-neo cassette Reading frame check (frame 0/1/2) 30 junction of lacZ-neo cassette 50 junction of PGK-BSD cassette 30 junction of PGK-BSD cassette

TCAAAGAGTTGGTAGCTCAGAG GATAAATCTGGAGCCGGTGAG TTGAAGGACTCCAATAGG TTGCACCACAGATGAAACGCC TAAGTGATGATAGAAGGG ATAGTGAAGGACAGTGATGGA GTCTTCAGAGATGGGGATGC

140

Jun Fu et al.

select for the correct product and eliminate the unwanted. If you want to plate out, use the instructions for low-salt LB and temperature incubation exactly as indicated above for the equivalent liquid step. Also, please note that the antibiotic concentrations in plates and liquid culture differ. For plates we recommend tetracycline (5 g/ml), gentamycin (2 g/ml), ampicillin (50 g/ml), kanamycin (15 g/ml), and blasticidin (40 g/ml) except when kanamycin and blasticidin are used together, then kanamycin (6 g/ml) and blasticidin (35 g/ml). When using the plating method, do not limit your procedure to one verified clone for the next step, but pool several verified clones. Thereby unrecognized point mutations in a single clone can be circumvented.

4. Standard Recombineering Electroporation Protocol 1. Start an overnight culture by picking one or more colonies from a plate and inoculate an 1.5-ml reaction tube containing 1.0 ml LB medium plus appropriate antibiotics. For BACs, we prefer to pick three or four separate colonies into one tube because the recombineering protocol selects for the correct product, so the chances to avoid occasional problems are improved. Puncture a hole in the lid of the tubes for air and incubate the cultures while shaking at 30  C overnight. We prefer to use benchtop Eppendorf Thermoshakers (950 rpm) because they are convenient. 2. Next day, put ddH2O (or 10% glycerol) and electroporation cuvettes on ice. Inoculate two 1.5 ml reaction tubes containing 1.4 ml LB plus appropriate antibiotics with 30–40 ml1 of the overnight culture and incubate with shaking for 2 h at 30  C, then add 20 ml 10% L-arabinose to induce the expression from the BAD promoter. (As a control, it is good to have another two tubes in parallel to which you do not add L-arabinose but otherwise treat the same.) Change the temperature to 37  C and incubate with shaking for 40 min. Note: Even though the pSC101 plasmid will be lost upon prolonged incubation at 37  C, the cells must be incubated at 37  C during the L-arabinose induction for strong expression and recombination. There are about five copies of pSC101 plasmid per cell and during 1 h there is approximately one doubling step. Therefore daughter cells will still have on average 2–3 copies left even though the plasmid has not replicated. If you want to retain the plasmid, return the culture to 30  C after the recovery period. Alternatively, to eliminate the plasmid from the host, continue incubating at 37  C.

A Recombineering Pipeline to Make Conditional Targeting Constructs

141

3. Spin down the cells for 30 s at 10,000 rpm in a cooled microfuge benchtop centrifuge (at 2  C). Discard the supernatant by quickly tipping it out twice, and place the tube on ice. Resuspend the pellet with 1 ml chilled ddH2O (or 10% glycerol), pipetting up and down three times to mix the suspension using blue tips. Repeat the centrifugation, tip off and resuspension. Centrifuge and tip off again so that about 20–30 ml will be left in the tube with the pellet. Keep the tubes on ice. 4. Add 100–600 ng (in 1–2 ml) of the cassette DNA/PCR product to the tube, resuspend cells and pipette the entire mixture into a chilled electroporation cuvette. Electroporate at 1350 V, 10 F, 600 O. This setting applies to an Eppendorf Electroporator 2510 using an electroporation cuvette with a slit of 1 mm. Other devices can be used, but high voltage and a 5 ms pulse are recommended. 5. Add 1 ml LB or SOC medium without antibiotics to the cuvette. Mix the cells carefully by pipetting up and down and pipette back into the microfuge tube. Incubate the cultures at 37  C with shaking for 1 h. Recombination occurs at 37  C during this recovery time. Either follow the liquid protocol detailed above or streak the cultures with a loop onto LB agar plates containing the appropriate antibiotics. For simple transformations plating of 50 l is sufficient, for recombineering all bacteria cells should be plated for single colonies (after centrifugation to reduce to volume). Media LB medium (10 g/l tryptone, 5 g/l yeast extract, 10 g/l NaCl, pH 7.0). Low-salt LB (10 g/l tryptone, 5 g/l yeast extract, 5 g/l NaCl, pH 8.0). SOC medium (20 mM glucose, 20 g/l tryptone, 5 g/l yeast extract, 0.5 g/l NaCl, 2.5 mM KCl, 5 mM MgCl2, and 5 mM MgSO4, pH 7.0). E. coli strains GB05 (HS996, DrecET D ybcC). The endogenous recET locus and the DLP12 prophage ybcC, which encodes a putative exonuclease similar to the Red a, were deleted. GB08-red (GB05, DlacZ pBAD abgA). After deletion of lacZ, BAD-abgA was inserted at ybcC locus. GB05-pir (GB05. pir). pir gene, which encodes Pi protein, was inserted at ybcC locus. HS996 (DH10B. fhuAIS2; phage T1-resistant). DH10B (F-mcrA D(mmr-hsdRMS-mcrBC) F80dlacZDM15 DlacX74 endA1 recA1 deoR D(ara, leu)7697 araD139 galU galK nupG rpsL l-).

142

Jun Fu et al.

5. Concluding Remarks The method to build a complex targeting construct for conditional mutagenesis presented here is another example of a recombineering pipeline based on liquid processing at most steps, rather than the conventional approach to plate out and pick colonies at each step. The pipeline is specialized to generate targeting constructs designed to be widely useful. In addition, the principles and reagents presented can be adapted to further variations to achieve alternative products. Hopefully, this exercise will illustrate how straightforward yet powerful recombineering can be.

ACKNOWLEDGMENTS The work described here was funded by the EU 6th Framework Programs, EUCOMM (European Conditional Mouse Mutagenesis Consortium; now part of IKMC), and HEROIC (LSHG-CT-2005-018883).

REFERENCES Anastassiadis, K., Fu, J., Patsch, C., Hu, S., Weidlich, S., Duerschke, K., Buchholz, F., Edenhofer, F., and Stewart, A. F. (2009). Dre recombinase, like Cre, is a highly efficient site specific recombinase in E. coli, mammalian cells and mice. Dis. Model Mech. 2, 508–515. Branda, C. S., and Dymecki, S. M. (2004). Talking about a revolution: The impact of sitespecific recombinases on genetic analyses in mice. Dev. Cell 6, 7–28. Buchholz, F., Angrand, P.-O., and Stewart, A. F. (1996). A simple assay to verify the functionality of Cre and FLP recombination targets in genomic manipulation constructs. Nucleic Acids Res. 24, 3118–3119. Chen, W. V., and Soriano, P. (2003). Gene trap mutagenesis in embryonic stem cells. Meth. Enzymol. 365, 367–386. Copeland, N. G., Jenkins, N. A., and Court, D. L. (2001). Recombineering: A powerful new tool for mouse functional genomics. Nat. Rev. Genet. 2, 769–779. Court, D. L., Sawitzke, J. A., and Thomason, L. C. (2002). Genetic engineering using homologous recombination. Annu. Rev. Genet. 36, 361–388. Datsenko, K. A., and Wanner, B. L. (2000). One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97, 6640–6645. Ellis, H. M., Yu, D., Di Tizio, T., and Court, D. L. (2001). High efficiency mutagenesis, repair and engineering of chromosomal DNA using single stranded oligonucleotides. Proc. Natl. Acad. Sci. USA 98, 6742–6746. Filutowicz, M., McEachern, M. J., and Helinski, D. R. (1986). Positive and negative roles of an initiator protein at an origin of replication. Proc. Natl. Acad. Sci. USA 83, 9645–9649. Glaser, S., Anastassiadis, K., and Stewart, A. F. (2005). Current issues in mouse genome engineering. Nat. Genet. 37, 1187–1193. Hamilton, C. M., Aldea, M., Washburn, B. K., Babitzke, P., and Kushner, S. R. (1989). New method for generating deletions and gene replacements in Escherichia coli. J. Bacteriol. 171, 4617–4622.

A Recombineering Pipeline to Make Conditional Targeting Constructs

143

Hashimoto-Gotoh, T., and Sekiguchi, M. (1977). Mutations of temperature sensitivity in R plasmid pSC101. J. Bacteriol. 131, 405–412. Ivics, Z., Li, M. A., Ma´te´s, L., Boeke, J. D., Nagy, A., Bradley, A., and Izsva´k, Z. (2009). Transposon-mediated genome manipulation in vertebrates. Nat. Methods. 6, 415–422. Mansour, S. L., Thomas, K. R., and Capecchi, M. R. (1988). Disruption of the protooncogene int-2 in mouse embryo-derived stem cells: A general strategy for targeting mutations to non-selectable genes. Nature 336, 348–352. Muyrers, J. P., Zhang, Y., Testa, G., and Stewart, A. F. (1999). Rapid modification of bacterial artificial chromosomes by ET recombination. Nucleic Acids Res. 27, 1555–1557. Poser, I., Sarov, M., Hutchins, J. R., He´riche´, J. K., Toyoda, Y., Pozniakovsky, A., Weigl, D., Nitzsche, A., Hegemann, B., Bird, A. W., Pelletier, L., Kittler, R., et al. (2008). BAC TransgeneOmics: A high-throughput method for exploration of protein function in mammals. Nat. Methods 5, 409–415. Ringrose, L., Chabanis, S., Angrand, P.-O., Woodroofe, C., and Stewart, A. F. (1999). Quantitative comparison of DNA looping in vitro and in vivo: Chromatin increases effective DNA flexibility at short distances. EMBO J. 23, 6630–6641. Sarov, M., Pozniakovski, A., Schneider, S., Roguev, A., Ernst, S., Zhang, Y., Hyman, A. A., and Stewart, A. F. (2006). A recombineering pipeline for protein tagging and functional analysis, as applied to Caenorhabtidis elegans. Nat. Methods 3, 839–844. Sauer, B., and McDermott, J. (2004). DNA recombination with a heterospecific Cre homolog identified from comparison of the pac-c1 regions of P1-related phages. Nucleic Acids Res. 32, 6086–6095. Sawitzke, J. A., Thomason, L. C., Costantino, N., Bubunenko, M., Datta, S., and Court, D. L. (2007). Recombineering: In vivo genetic engineering in E. coli, S. enterica, and beyond. Meth. Enzymol. 421, 171–199. Skarnes, W. C., Rosen, B., Koutsourakis, M., West, A. P., Bushell, W., Iyer, V., Cox, T., Jackson, D., Severin, J., Biggs, P., Thomas, M., Mujica, A., et al. (2010). A conditional knockout resource for genome-wide analysis of mouse gene function. Nature (in press). Szymczak, A. L., Workman, C. J., Wang, Y., Vignali, K. M., Dilioglou, S., Vanin, E. F., and Vignali, D. A. (2004). Correction of multi-gene deficiency in vivo using a single ’selfcleaving’ 2A peptide-based retroviral vector. Nat. Biotechnol. 22, 589–594. Testa, G., Zhang, Y., Vintersten, K., van der Hoeven, F., Benes, V., Chambers, I., Smith, A. J. H., Smith, A. A., and Stewart, A. F. (2003). Engineering the mouse genome with bacterial artificial chromosomes to create multi-purpose alleles. Nat. Biotechnol. 21, 443–447. Testa, G., Schaft, J., van der Hoeven, F., Glaser, S., Anastassiadis, K., Zhang, Y., Hermann, T., Stremmel, W., and Stewart, A. F. (2004). A reliable expression reporter cassette for multipurpose, knock-out/conditional mouse alleles. Genesis 38, 151–158. Valenzuela, D. M., Murphy, A. J., Frendewey, D., Gale, N. W., Economides, A. N., Auerbach, W., Poueymirou, W. T., Adams, N. C., Rojas, J., Yasenchak, J., Chernomorski, R., Boucher, M., et al. (2003). High-throughput engineering of the mouse genome coupled with high-resolution expression analysis. Nat. Biotechnol. 21, 652–659. Wang, J., Sarov, M., Rientjes, J. J., Stewart, A. F., and Zhang, Y. (2006). Improved recombineering by addition of the non-recombinogenic properties of RecA. Mol. Biotechnol. 32, 43–54. Wu, S., Ying, G., Wu, Q., and Capecchi, M. R. (2008). A protocol for constructing gene targeting vectors: Generating knockout mice for the cadherin family and beyond. Nat. Protoc. 3, 1056–1076. Yagi, T., Ikawa, Y., Yoshida, K., Shigetani, Y., Takeda, N., Mabuchi, I., Yamamoto, T., and Aizawa, S. (1990). Homologous recombination at c-fyn locus of mouse embryonic

144

Jun Fu et al.

stem cells with use of diphtheria toxin A-fragment gene in negative selection. Proc. Natl. Acad. Sci. USA 87, 9918–9922. Yang, Y., and Seed, B. (2003). Site-specific gene targeting in mouse embryonic stem cells with intact bacterial artificial chromosomes. Nat. Biotechnol. 21(4), 447–451. Yang, X. W., Model, P., and Heintz, N. (1997). Homologous recombination based modification in Escherichia coli and germline transmission in transgenic mice of a bacterial artificial chromosome. Nat. Biotechnol. 15, 859–865. Yu, D., Ellis, H. M., Lee, E. C., Jenkins, N. A., Copeland, N. G., and Court, D. L. (2000). An efficient recombination system for chromosome engineering in Escherichia coli. Proc. Natl. Acad. Sci. USA 97, 5978–5983. Zhang, Y., Buchholz, F., Muyrers, J. P., and Stewart, A. F. (1998). A new logic for DNA engineering using recombination in Escherichia coli. Nat. Genet. 20, 123–128. Zhang, Y., Muyrers, J. P., Testa, G., and Stewart, A. F. (2000). DNA cloning by homologous recombination in Escherichia coli. Nat. Biotechnol. 18, 1314–1317.

C H A P T E R

N I N E

Confirmation of Recombination Site Functionality in Gene Targeting Vectors using RecombinaseExpressing Bacteria M. David Stewart* and Richard R. Behringer† Contents 146 147 147 148 150 150 150

1. Introduction 2. Materials 3. Method 4. Example of Results 5. Summary Acknowledgments References

Abstract Recognition sequences for the site-specific DNA recombinases Cre and FLP are commonly incorporated into gene targeting vectors for the purposes of removing selection markers or generating conditional alleles. Gene targeting vectors typically contain a positive selection marker, such as the neomycin resistance gene, flanked by loxP sites. Thus, the selection marker can be removed by breeding to a mouse strain which expresses Cre recombinase in its germ line. Conditional knockout vectors typically have one or more exons flanked by loxP sites and the positive selection marker flanked by FRT sites. Thus, the selection marker is removed with FLP recombinase and the knockout allele is generated in tissues expressing Cre recombinase. Because the generation of mice by gene targeting in embryonic stem (ES) cells is an expensive and time-consuming process, it is important to confirm that the recombination sites in your targeting vector are functional prior to electroporation of ES cells. This chapter describes a simple method for testing the functionality of loxP and FRT sites in vivo using Cre- or FLP-expressing bacteria. * Department of Biology and Biochemistry, University of Houston, Houston, Texas, USA Department of Genetics, University of Texas M.D. Anderson Cancer Center, Houston, Texas, USA

{

Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77009-9

#

2010 Elsevier Inc. All rights reserved.

145

146

M. David Stewart and Richard R. Behringer

1. Introduction Technology allowing mutagenesis of the mouse genome via homologous recombination in embryonic stem (ES) cells has yielded thousands of mouse mutants. Phenotypic analysis of these mutant strains has provided important information regarding the physiological functions of individual genes. Gene targeting in ES cells is accomplished by first generating a targeting vector consisting of a positive selection marker, usually the neomycin resistance gene (neo) whose expression is driven by a regulatory sequence such as that of the phosphoglycerate kinase (PGK) gene that is active in mouse ES cells. The PGK–neo cassette is flanked by 50 and 30 arms of homology to the gene of interest. Typically for knockout experiments the arms of homology exclude one or more exons important for gene function. Thus, following homologous recombination, the mutant allele lacks these exons and hopefully produces a nonfunctional protein product or no protein product at all. The targeting vector may also contain a negative selection marker outside the homologous arms to select against random integration. Because the inclusion of PGK–neo in the targeted allele may interfere with gene function, it is common to design the targeting vector such that the PGK–neo cassette is flanked by recognition sequences for either Cre or FLP recombinase (Fig. 9.1A). In this manner the PGK–neo cassette may be removed after identification of correctly targeted ES cell clones. Creation of conditional knockout alleles is also a very common gene targeting strategy. These alleles are designed such that critical exons are flanked by recognition sites for Cre recombinase. Breeding to an appropriate tissue-specific Cre-expressing mouse strain will yield tissue-specific excision of the critical exons and thus a deletion of gene function in that particular tissue. Targeting vectors created for this purpose usually contain A 5⬘ arm of homology

PGK-neo loxP

3⬘ arm of homology loxP

B 5⬘ arm of homology loxP

Exon 2 loxP FRT

PGK-neo

3⬘ arm of homology

FRT

Figure 9.1 Gene targeting strategies using loxP and FRT sites. (A) Traditional targeting vector with PGK–neo flanked by loxP sites. (B) Conditional knockout vector with an exon flanked by loxP sites and PGK–neo flanked by FRT sites.

Confirmation of Recombination Site Functionality

147

both loxP and FRT sites. The loxP sites flank the critical exons and the FRT sites flank the positive selection marker (PGK–neo; Fig. 9.1B). These mice can be bred to germ line FLP-expressing mice to permanently remove the PGK–neo cassette and a tissue-specific Cre-expressing line to generate conditional knockout mice for phenotypic analysis. Generating mutant mice through gene targeting in ES cells is a costly and time-consuming process. Thus, it is advised to take all precautions necessary to ensure that the targeting vector is correctly designed. The method described in this chapter was developed to ensure the recombination sites of your targeting vector are functional in vivo. Nonfunctional recombination sites may occur due to molecular cloning errors while making the targeting vector or uncontrollable errors during homologous recombination. While we absolutely advise sequencing the targeting vector to confirm everything is correct, including the recombination sites, we also recommend using the following simple method to functionally test the recombination sites in vivo prior to electroporation into ES cells.

2. Materials Cre-expressing Escherichia coli strains: BNN132 (ATCC, cat # 47059) (Elledge et al., 1991), EKA133 (ATCC, cat # 47041) (Ayres et al., 1993), BS1365 (Sauer and Henderson, 1988), or 294-CRE (Buchholz et al., 1996). FLP-expressing E. coli strains: PS393 (Babineau et al., 1985) and 294-FLP (Buchholz et al., 1996). Your gene targeting vector.

3. Method Figure 9.2 shows a flow diagram of the basic method described below. Dilute the targeting vector to a concentration of 3 ng/ml. Thaw a 100 ml aliquot of recombinase-expressing competent bacteria on ice (5 min). Add 1 ml of targeting vector (3 ng) and mix by gently flicking the tube with your finger. Incubate on ice for 30 min. Heat shock at 42  C for 45 s. Place immediately back on ice. Keep on ice for 2 min. Add 500 ml of LB or SOC and shake for 1 h at 37  C.

148

M. David Stewart and Richard R. Behringer

Transform Cre or FLP recombinaseexpressing competent cells with your targeting vector

Spread on plates of appropriate antibiotic

Innoculate mini-cultures

Purify plasimd DNA

Analyze for “pop-out” by either restriction enzyme digestion or PCR

Figure 9.2 Outline of experimental procedure.

Spread 20 and 50 ml of cell suspension on LB agar plates containing the appropriate antibiotic for your targeting vector. Incubate overnight at 37  C. Pick 20 colonies and inoculate 2–3 ml LB cultures with appropriate antibiotic. Incubate overnight in a 37  C shaking incubator. Prepare plasmid DNA from the minicultures using standard procedures. Analyze the DNA for recombination by either restriction enzyme digestion or PCR (see example below). Expect mosaicism, that is, there will be a mixture of intact and recombined plasmid DNA, because the Cre and FLP recombinases expressed in the bacteria exhibit a varying degree of activity. In normal practice, they will not recombine and ‘‘pop-out’’ the intervening sequence 100% of the time. The presence of a smaller band of the correct size in a clone indicates that the recombination sites are functional in vivo and it is safe to proceed with ES cell electroporation.

4. Example of Results Figure 9.3A and B illustrates the results of restriction enzyme analysis of a gene targeting vector introduced into Cre-expressing bacteria. In this example, the targeting vector yields fragments of 8.3 and 5.2 kb when

149

Confirmation of Recombination Site Functionality

A

8.3 kb eo K-n PG

C

F1

BamHI

F2 R1 PGK-neo 2.1 kb 550 bp

BamHI

Cre

5.2 kb Cre 6.5 kb

360 bp BamHI

D F1/R1

1 BamHI

Clones 2 3

4 360 bp

Clones 1

2

3

4

Targeting vector

B

F2/R1

5.2 kb 550 bp

8.3 kb 6.5 kb 5.2 kb

Figure 9.3 Analysis of recombination in bacteria. (A) For this targeting vector BamHI digestion yields fragments of 8.3 and 5.2 kb prior to recombination and 6.5 and 5.2 kb following removal of the PGK–neo cassette by Cre recombinase. (B) Example of a restriction enzyme digestion. Clone 1 ¼ complete excision; Clones 2–4 ¼ varying degrees of excision. (C) Testing the functionality of recombinase recognition sequences by PCR. In this example two PCR reactions were performed for each clone. Primer set F1/R1 yields no band if PGK–neo is present (2.1 kb is too long for the extension time) and a 360 bp band if PGK–neo is removed by Cre recombinase. Primer set F2/R1 yields a 550 bp band if PGK–neo is present and no band if it is removed. (D) Clone 1 ¼ complete excision; Clones 2–4 ¼ partial excision.

digested with BamHI. Following removal of PGK–neo by Cre recombinase, the 8.3 kb fragment is reduced to 6.5 kb. This difference can be easily distinguished on an agrose gel. As explained above, there is mosaicism; the results indicate both the presence and absence of Cre-mediated recombination. In this example, Clone 1 shows complete recombination; whereas Clones 2–4 exhibit varying degrees of recombination. Figure 9.3C and D illustrates the results of analysis by PCR. In this experiment, two reactions were performed for each clone. Primer set

150

M. David Stewart and Richard R. Behringer

F1/R1 yields no band in the presence of PGK–neo because the extension time is too short to produce 2.1 kb products. However, in the absence of PGK–neo this primer set yields a product of 360 bp. Primer set F2/R1 yields a 550 bp product in the presence of PGK–neo and no product in the absence. In this experiment, Clone 1 shows complete recombination (presence of 360 bp product, but no 550 bp product); whereas Clones 2–4 exhibit varying degrees of recombination.

5. Summary Gene targeting in ES cells and the subsequent generation of correctly targeted strains of mice is a time-consuming and expensive process. For this reason, it is advisable to take all precautions necessary such that the mice generated will have the precise mutation that you designed. In this chapter we have described a simple method to test the functionality of loxP and FRT recombination sites in vivo. A positive result from this simple procedure will give an investigator confidence that the gene targeting plasmid that is introduced into ES cells has functional recombination sites. Likewise, a negative result saves an enormous amount of time, energy, and money that would have been expended in the screening of ES cells and generation of mice.

ACKNOWLEDGMENTS R. R. B. was supported by National Institutes of Health (NIH) grant HD30284. M. D. S was supported by the NIH/National Cancer Institute Training Program in Molecular Genetics of Cancer CA009299.

REFERENCES Ayres, E. K., Thomson, V. J., Merino, G., Balderes, D., and Figurski, D. H. (1993). Precise deletions in large bacterial genomes by vector-mediated excision (VEX). The trfA gene of promiscuous plasmid RK2 is essential for replication in several Gram-negative hosts. J. Mol. Biol. 230, 174–185. Babineau, D., Vetter, D., Andrews, B. J., Gronostajski, R. M., Proteau, G. A., Beatty, L. G., and Sadowski, P. D. (1985). The FLP protein of the 2-micron plasmid of yeast. Purification of the protein from Escherichia coli cells expressing the cloned FLP gene. J. Biol. Chem. 260, 12313–12319. Buchholz, F., Angrand, P. O., and Stewart, A. F. (1996). A simple assay to determine the functionality of Cre or FLP recombination targets in genomic manipulation constructs. Nucleic Acids Res. 24, 3118–3119.

Confirmation of Recombination Site Functionality

151

Elledge, S. J., Mulligan, J. T., Ramer, S. W., Spottswood, M., and Davis, R. W. (1991). Lambda YES: A multifunctional cDNA expression vector for the isolation of genes by complementation of yeast and Escherichia coli mutations. Proc. Natl. Acad. Sci. U. S. A. 88, 1731–1735. Sauer, B., and Henderson, N. (1988). The cyclization of linear DNA in Escherichia coli by site-specific recombination. Gene 70, 331–341.

C H A P T E R

T E N

Genetic Fate Mapping Using Site-Specific Recombinases Emilie Legue´ and Alexandra L. Joyner Contents 1. Principles Behind Genetic Fate Mapping 2. Genetic Fate Mapping Technique 2.1. Tools for genetic fate mapping 2.2. Genetic fate mapping methods 2.3. Important considerations in designing all types of genetic fate mapping studies 2.4. Special consideration when designing GIFM experiments 2.5. Special consideration when designing genetic inducible clonal analysis experiments 3. Future Applications: Combining Genetic Fate Mapping with Mutant Analysis 3.1. Intrachromosomal recombination: Using existing conditional loxP containing alleles 3.2. Interchromosomal recombination: Mosaic analysis with double makers References

154 156 156 163 164 170 175 177 177 177 178

Abstract Understanding how cells are assembled in three dimensions to generate an organ, or a whole organism, is a pivotal question in developmental biology. Similarly, it is critical to understand how adult stem cells integrate into an existing organ during regeneration or in response to injury. Key to discovering the answers to these questions is being able to study the various behaviors of distinct cell types during development or regeneration. Fate mapping techniques are fundamental to studying cell behaviors such as proliferation, movement, and lineage segregation, as the techniques allow precursor cells to be marked and their descendants followed and characterized over time. The generation of transgenic mice, combined with the use of site-specific

Memorial Sloan-Kettering Cancer Center, New York, USA Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77010-5

#

2010 Elsevier Inc. All rights reserved.

153

154

Emilie Legue´ and Alexandra L. Joyner

recombinases (SSR) in the mouse genome, has provided a means to develop powerful genetic fate mapping approaches. A key advantage of genetic fate mapping is that it allows cells to be genetically marked, and therefore the mark is transmitted to all the descendants of the initially marked cells. By making modifications to the SSRs that render their enzymatic activity inducible, and the development of an assortment of reporter alleles for marking cells, increasingly sophisticated genetic fate mapping studies can be performed. In this chapter, we review the four main genetic fate mapping methods that utilize intrachromosomal recombination to mark cells (cumulative, inducible, clonal, and intersectional) and one interchromosomal method, the tools required to carry out each approach, and the practical considerations that have to be taken into account before embarking on each type of genetic fate mapping study.

1. Principles Behind Genetic Fate Mapping A critical question in developmental biology is how a cell acquires its ultimate differentiated state (or fate), which includes how it becomes assembled in three dimensions along with other cells to generate a functional organ. Equally important to understand is the potential of adult stem cells to replenish tissues during homeostasis and to repair injured tissues. Fate mapping is an essential technique for answering these questions. Fate mapping consists of marking a group of cells, or a single cell, in the embryo and then determining what the cells and all their descendents become after development is complete, or in the adult after a regenerative process has taken place. The final position at which the marked cells and/or their progeny settle is also revealed by fate mapping. By following the fate of the marked cells at different stages of development, the lineage of a cell, that is its genealogical series of ancestors (Fig. 10.1), can be uncovered to provide information on when and how different cell types become segregated from a common ancestor. In the mouse, fate mapping studies using traditional invasive methods, such as viral infection or dye marking, although they have provided invaluable knowledge (Fields-Berry et al., 1992; Golden et al., 1995; Lawson, 1999; Sanes et al., 1986), have been hampered by the inaccessibility of the embryo in the uterus and the limited time during which embryos can be cultured. The development of transgenic mice and site-specific recombinases (SSRs) that are effective in mice provided a new approach for genetically marking cells in the intact embryo or postnatal mouse, without a need for manipulation of the embryo or an organ itself. The application of these new fate mapping approaches is termed genetic fate mapping, since the marking of the cells involves a permanent change in their DNA that allows the cell to be identified from its unmarked neighbors. Furthermore, the genetic nature of the labeling ensures the transmission of the marker to all the progeny of the initially labeled

155

Genetic Fate Mapping

A Cumulative genetic fate mapping

B

Genetic inducible fate-mapping

Time

Time

Induction

M

N

C

N

M

Genetic inducible clonal analysis Time

Induction

N

M

Expression domain of the SSR driver

Unmarked cell

Marked cell

Figure 10.1 Three types of genetic fate mapping lead to marking of different cell populations. A lineage tree is shown of the cells arising from a single ancestor that gives rise to two structures (ovals marked as M and N) in the mouse. The domain of expression of the promoter that drives the SSR is represented by light blue shading. It encompasses cells from several generations (aligned from top to bottom) and cells that are spatially distinct (aligned from left to right). The marked cells are represented with a bold outline and expression of a reporter protein by dark blue after recombination of the reporter allele. (A) In cumulative fate mapping, all the cells that express the SSR at some time point undergo recombination of the reporter allele, and if transcription of the reporter allele is driven by a ubiquitous promoter, all the cells that have ever expressed the SSR and their descendants are marked. (B) In genetic inducible fate mapping, only the cells that express the SSR at the time of induction (here represented over one cell generation by an orange arrow) can undergo recombination of the reporter allele and express the reporter allele, and transmit the recombined reporter allele to all their descendants. (C) In clonal analysis, the dose of ligand administered to the mouse is titrated down such that only one precursor undergoes recombination (is marked), and transmits the recombined allele to its descendants.

cells, circumventing the problem of dilution of a dye marker during cell division. By choosing the appropriate promoter to drive expression of the SSR, a population of cells of interest can be precisely marked genetically at any time point during development or in the adult, and its progeny then followed over time. Thus, genetic fate mapping in mouse is a powerful

156

Emilie Legue´ and Alexandra L. Joyner

method for studying the behavior of cell lineages in the intact mouse at any stage of development, or in the adult during homeostasis and organ regeneration.

2. Genetic Fate Mapping Technique Genetic fate mapping requires the generation of two transgenic mouse lines, and the two transgenes are then combined in one mouse by breeding. One transgene has a promoter that is expressed in the cell type of interest (i.e., the population of cells to be fate mapped) and drives expression of an SSR (SSR line). The second transgene typically contains a ubiquitous promoter, so that all the progeny of the population of interest can express the marker gene and be visualized, upstream of two copies of the DNA sequences recognized by the SSR flanking a cassette that inhibits expression of a downstream marker or reporter gene (reporter line). In a cell that expresses the SSR, the two SSR sites in the reporter allele will undergo recombination and the recombined DNA configuration will then be transmitted to all of its descendants, thus all cells expressing the SSR and all their descendents will be genetically marked (Fig. 10.1). If the promoter in the reporter allele is expressed ubiquitously, all the genetically marked cells will express the reporter protein, and thus can be visualized. In genetic fating mapping, the marking of cells is therefore heritable and permanent due to its genetic nature.

2.1. Tools for genetic fate mapping 2.1.1. The site-specific recombinases Two SSRs have primarily been used for genetic fate mapping in mice; the Cre recombinase from the bacteriophage P1 (Austin et al., 1981) and the Flp recombinase from Saccharomyces cerevisiae (Farley et al., 2000). The Cre recombinase specifically acts on a 34-bp DNA sequence called a loxP site (Hoess et al., 1982), whereas the Flp recombinase acts on a 47-bp FRT site (Andrews et al., 1985). If a DNA sequence is flanked by two loxP or FRT sites in the same orientation, then the sequence will be efficiently excised by Cre or Flp recombination, respectively, leaving one loxP or FRT site behind (Sauer, 1993; Schaft et al., 2001). See Chapter 7 for more details. 2.1.2. Modifications to the sites-specific recombinases A number of critical modifications have been made to the SSRs to maximize the activity of the recombinases (see Chapter 7 for details), as well as to gain temporal control over their activity. An optimized form of Flp that is more stable than the original yeast protein at mouse body temperature is

Genetic Fate Mapping

157

referred to as Flpe (e ¼ enhanced; Buchholz et al., 1998). More recently, new versions of Flpe have been generated in which codon usage was optimized for translation in mouse, and they are referred to as Flpo (o ¼ optimized codon usage; Raymond and Soriano, 2007; Wu et al., 2009), these versions of Flp confer better efficiency to the SSR. Conversely, a thermolabile form of Flp exists that has a low recombinase efficiency (FlpL; Buchholz et al., 1996), and might be useful for clonal analysis (see below) (Dymecki and Kim, 2007). The most versatile modification of the SSRs in terms of genetic fate mapping studies has been the fusion of the SSRs with a protein domain that allows the activity of the SSRs to be temporally controlled. The hormonebinding domains from homodimeric nuclear receptors (nuR) have been used to control Cre and Flp activity. In the absence of the appropriate ligand, the hormone-binding domains ensure that the SSR-fusion protein is sequestered in the cytoplasm (Fig. 10.2A). Upon binding of the hormone to the domain, the cytoplasmic sequestration is relieved and the SSR-fusion protein is translocated into the nucleus (Fig. 10.2B and C). Thus, the fusion of an SSR to an nuR hormone-binding domain allows the subcellular localization of the SSR to be controlled, and therefore its accessibility to DNA. Site-specific recombination is therefore dependent on ligand administration. The hormone-binding domains from the human and mouse estrogen receptor (ER) or the human progesterone receptor (PR) have been used to generate CreER and FlpeER or CrePR fusions (Hunter et al., 2005; Metzger and Chambon, 2001; Tsujita et al., 1999). In order to avoid interaction of the hormone-binding domains with endogenous hormones, the ER or the PR binding domains have been modified such that their affinity for estrogen or progesterone is reduced, and their affinity for another ligand increased (Feil et al., 1996, 1997; Vegeto et al., 1992). A modified ER estrogen-binding domain that preferentially binds an antagonist of estrogen (4-hydroxy-tamoxifen) (Feil et al., 1996, 1997; Indra et al., 1999) or a truncated PR progesterone-binding domain that binds an antagonist of progesterone (RU-486) have been developed (Vegeto et al., 1992). The resulting modified ER hormone-binding domains are named ERT (T for tamoxifen, where a number following T indicates subsequent modifications of the domain). In the absence of ligand, the SSR-nuR hormonebinding domain fusion protein is sequestrated in the cytoplasm in a large protein aggregate that includes heat shock protein chaperones (Fig. 10.2A). Binding of the ligand to the modified nuR hormone-binding domain of the SSR-fusion protein (Fig. 10.2B) induces a change in conformation of the hormone-binding domain and its subsequent release from the chaperone. The SSR-fusion protein then translocates into the nucleus where the SSR can mediate site-specific recombination (Fig. 10.2C). For genetic fate mapping studies, a stop sequence flanked by two loxP or FRT sites is excised from a reporter allele leading to expression of a protein that marks

Ligand

A

B

Prom 1 SSR

SSR

nuR

Prom 1

nuR Hsp90

SSR Prom 2

Stop

SSR

nuR

nuR Hsp90

Reporter

Nucleus

Prom 2

Stop

Reporter

Nucleus

Cytoplasm

Cytoplasm

C

D

Prom 1

SSR

nuR

Prom 2

SSR

Stop

SSR Hsp90

nuR

Prom 1

Reporter

Prom 2

nuR

SSR

nuR

Reporter nuR

Hsp90

Nucleus Cytoplasm

SSR

Nucleus Cytoplasm

Figure 10.2 Genetic inducible fate mapping. (A) In the absence of ligand, the fusion protein between an SSR and nuclear receptor (nuR) is bound to Hsp90 and retained in the cytoplasm. (B) Binding of the ligand induces a conformational change, and Hsp90 is released. (C) The SSR-fusion protein then translocates into the nucleus where it induces recombination between the DNA target sites (c) flanking a stop sequence in a reporter allele that inhibits transcription of the reporter gene. (D) Once the stop sequence is excised, the reporter protein is produced and the cell and its descendents are marked/labeled (e.g., cytoplasm expresses lacZ, indicated by blue).

Genetic Fate Mapping

159

the cell (Fig. 10.2D). Fate mapping approaches that utilize a temporally controlled SSR-fusion protein are referred to as genetic inducible fate mapping (GIFM; Joyner and Zervas, 2006), since the activity of the SSRfusion protein can be induced at any time point (temporal control). GIFM is therefore a powerful approach since the marking of cells is controlled by both the precision of the promoter-driving expression of the SSR-fusion protein and a small time window when the SSR is active (see below). 2.1.3. The reporter alleles 2.1.3.1. Intrachromosomal recombination In a reporter allele that undergoes intrachromosomal recombination, the recombination sites are placed in cis, and in the same orientation, flanking a stop sequence. The stop sequence typically includes a polyadenylation sequence that signals termination of transcription, and thus prevents transcription of a downstream reporter gene. Expression of the reporter protein is therefore dependent on the recombination and excision of the loxP/FRT flanked stop sequence by the SSR. The stop sequence often also contains a protein coding sequence such as neo. If the promoter driving the reporter gene is expressed in all cells of the mouse (ubiquitous promoter), then all cells that have undergone SSR-mediated recombination, as well as their descendants, will be permanently marked by expression of the reporter protein. In the simplest reporter alleles, there is only one reporter gene, for instance, LacZ or GFP (Fig. 10.3A). Some reporter alleles contain an additional reporter gene as part of the stop sequence. For example, a LacZ gene has been incorporated into the stop sequence and an alkaline phosphatase (Z/AP reporter) or an eGFP (Z/EG reporter, Fig. 10.3B) gene placed downstream of the stop sequence (Lobe et al., 1999; Novak et al., 2000). The activity of the SSR thus switches expression (or marking) of cells from one reporter to the other—the reporter gene present in the stop sequence is expressed before site-specific recombination and the reporter gene downstream the stop sequence is expressed after recombination (Fig. 10.3B and B0 ). In an additional approach called intersectional (or combinatorial) genetic fate mapping, the reporter allele has both a pair of loxP and FRT sites and two reporters—the first reporter is preceded by a loxP (or FRT) flanked stop sequence and is also surrounded by FRT (or loxP) sites (Awatramani et al., 2003; Dymecki and Kim, 2007) (Fig. 10.3D) (see below and Chapter 11 for details). The power of intersectional fate mapping is that for the second reporter to be expressed, a cell must have expressed both Cre and Flp (be at the intersection of the two SSR expression domains) (Fig. 10.3D00 ), thus refining the domain or cell population that can be studied (marked). More recently the production of spectral variants of GFP (Chudakov et al., 2005; Pakhomov and Martynov, 2008) and modified recombination sites for Cre (Kolb, 2001) have allowed for the generation of new reporter

A Prom

Stop

A⬘

LacZ

B Prom

Stop

LacZ

Prom

LacZ

Prom

GFP

C⬘1

Prom

M-RFP

C⬘2

Prom

C⬘3

Prom

D⬘

loxP FRT Prom LacZ

B⬘

GFP

C Prom

Stop

OFP

M-RFP

Stop

M-YFP

Stop

M-CFP

Stop Stop

M-YFP

M-YFP

Stop

M-CFP

M-CFP

M-CFP (Adapted from Livet et al., 2007)

D

loxP Prom

loxP FRT Stop

LacZ

FRT Stop

GFP

Cre recombination

FRT Stop GFP

Flp recombination

D⬙

Prom

GFP (Adapted from Dymecki and Kim, 2007)

Figure 10.3 Examples of reporter alleles used for intrachromosomal recombination genetic fate mapping. Four major types of genetic fate mapping reporter alleles for marking cells using intrachromosomal recombination are shown, both before (left panel) and after recombination (right panel). The color in the circles to the right of each DNA configuration represents the protein that labels the cell. (A) The simplest reporter allele contains a stop sequence flanked by recombination sites upstream a reporter gene. (B) An example of a reporter gene (Z/EG; Novak et al., 2000) in which there is a switch from expression of one marker protein (color) to another. (C) The Brainbow reporter allele (Livet et al., 2007) contains three pairs of different loxP sites. It generates three products of recombination (C0 1, C0 2, and C0 3) depending on which pair of loxP sites undergoes Cre-mediated recombination. (D) A reporter allele for intersectional genetic fate mapping is shown (Awatramani et al., 2003; Dymecki and Kim, 2007). Both loxP and FRT site-specific recombination sites are present in the reporter allele. Expression of both Cre and Flp is required for expression of the reporter protein GFP.

Genetic Fate Mapping

161

alleles that lead to random marking of cells with several different reporters within the same animal (Livet et al., 2007). Such reporters can be very useful for clonal analyses (see below). The reporter allele is engineered to comprise several reporter genes encoding fluorescent proteins with different spectral properties in tandem. Three different loxP sites are inserted upstream of the reporters genes, and in addition, each reporter gene is followed by a stop sequence and one of the modified loxP sites (Fig. 10.3C). SSR-mediated recombination leads to random deletion of the reporters situated between a pair of modified loxP sites, leading to expression of the remaining most 50 reporter (Fig. 10.3C0 1, C0 2, and C0 3). If multiple reporter alleles are present in a cell, then multiple fluorescent proteins will be randomly coexpressed (one from each copy of the reporter allele). The various combinations of coexpressed fluorescent proteins result in an array of possible colors a cell and its descendants can be marked with (up to 10 to date; Livet et al., 2007). This additional complexity of marking thus provides a means for many distinct recombination events to be distinguished and studied within the same animal. To date, such reporter lines have been generated with a transgene that contains a Thy1 promoter fragment that is expressed in a variety of neurons, and have been named Brainbow (Livet et al., 2007). Finally, in the future it is likely that reporter alleles will be optimized for live imaging of marked cells by using proteins with enhanced fluorescence, minimal cell toxicity, and robust levels of protein expression (Luche et al., 2007 and see Chapter 20 in volume 476). 2.1.3.2. Interchromosomal recombination—Mosaic analysis with double markers The Luo lab has generated a reporter system that is dependent on Cre activity and two reporter alleles, each carrying a complementary half of a reporter transgene inserted by gene targeting into the same locus on homologous chromosomes (Zong et al., 2005). One transgene contains the coding sequence for the N-terminal part of GFP and the C-terminal part of RFP separated by a loxP site that disrupts translation, while the other allele contains the N-terminal part of RFP and the C-terminal part of GFP separated by a loxP site (Fig. 10.4). When Cre is expressed in a cell, sitespecific recombination between the two loxP sites on homologous chromosomes occurs leading to exchange of the C-terminal and N-terminal portions of the reporters, and production of GFP from one chromosome and RFP from the other (Fig. 10.4). A unique aspect of this system is that it allows for the generation of labeled cells that are also homozygous for a mutation on the chromosome harboring the reporter alleles. With this technique, named mosaic analysis with double markers (MADM; Zong et al., 2005), mutant cells can be induced that are genetically marked and therefore fate mapped in an otherwise wild-type environment.

Gene ´ mutant

N-ter-GFP

C-ter-RFP

Gene ´ mutant

N-ter-GFP

C-ter-RFP

Gene ´ wild-type

N-ter-RFP

C-ter-GFP

Gene ´ mutant

N-ter-GFP

C-ter-GFP

Gene ´ wild-type

N-ter-RFP

C-ter-RFP

Gene ´ wild-type

N-ter-RFP

C-ter-GFP

Gene ´ mutant

N-ter-GFP

C-ter-RFP

Gene ´ wild-type

N-ter-RFP

C-ter-GFP

Gene ´ wild-type

N-ter-RFP

C-ter-RFP

Gene ´ mutant

N-ter-GFP

C-ter-GFP

−/−

Scenario 1 (X segregation)

Replication Gene ´ mutant

N-ter-GFP

C-ter-RFP

Gene ´ mutant

N-ter-GFP

C-ter-RFP

Gene ´ wild-type

N-ter-RFP

C-ter-GFP

Gene ´ wild-type

N-ter-RFP

C-ter-GFP

+/+

Recombination Gene ´ mutant

N-ter-GFP

C-ter-RFP

Gene ´ wild-type

N-ter-RFP

C-ter-RFP

Gene ´ mutant

N-ter-GFP

C-ter-GFP

Gene ´ wild-type

N-ter-RFP

C-ter-GFP

+/−

Scenario 2 (Z segregation) +/−

Figure 10.4 Schematic representation of a reporter allele for interchromosomal recombination (MADM; Zong et al., 2005). In MADM reporter allele, the loxP sites are on two homologous chromosomes. The sequences of the reporter genes are complementary, each containing either the N-terminal coding sequence of GFP and the C-terminal coding sequence of RFP or the N-terminal coding sequence of RFP and the C-terminal coding sequence of GFP. Recombination between the loxP sites restores the full coding sequences for both genes. The advantage of such a reporter allele is that they can be used to simultaneously mutate and mark cells, if a mutated gene is on the same chromosome as the reporter. Marking and mutating cells requires replication of DNA and depends on the segregation of the chromatids. The possible outcomes of chromatid segregation are represented on the right, with the genotype and labeling of the cells that are generated. In the X segregation scenario (the two recombined chromatids segregate to different cells), two labeled cells are generated: a mutant GFP-expressing cell and a wild-type RFP-expressing cell. In the Y segregation scenario, the two recombined chromatids segregate in the same cell and the cells generated are heterozygous for the mutation, one expresses both GFP and RFP and is represented in yellow and the other is not labeled.

Genetic Fate Mapping

163

2.2. Genetic fate mapping methods With the modifications made in the SSRs and a variety of different reporter lines, the applications of genetic fate mapping have been broadened. In this section, we describe the basic uses of each type of genetic fate mapping. 2.2.1. Cumulative genetic fate mapping Cumulative genetic fate mapping involves the use of an SSR (e.g., Cre) that is constitutively active. The approach is called ‘‘cumulative’’ because all the cells that at any point during development expressed a sufficient amount of Cre to induce recombination of the loxP sites in the reporter allele, as well as all their descendants, will be genetically marked. During development, any additional cell that begins to express Cre will add its contribution to the total number of cells that are ultimately marked (Fig. 10.1A). The specificity of potential cell labeling (spatial and temporal) in cumulative genetic fate mapping therefore comes entirely from the regulatory sequence chosen to drive SSR expression. Since promoter activity can change during development, it often is not possible to use this approach for definitive prospective fate mapping of a particular group of cells or cell lineage. All labeled cells therefore arise from the additive (cumulative) contribution of all the cells that at any previous time point expressed sufficient SSR to induce recombination of the reporter allele. 2.2.2. Genetic inducible fate mapping In GIFM studies, the timing of the SSR-mediated recombination event (marking of cells) is controlled (i.e., induced by ligand), and limited to a discrete period of time (Fig. 10.1B). GIFM therefore utilizes one of the modified nuR hormone-binding domain SSR-fusion proteins that allow for temporal control over the site-specific recombination activity by administration of a specific ligand. Recombination occurs only over an approximately 24-h time period (see below for details), and recombination is less efficient than with constitutive SSRs. Although this leads to mosaic marking of the population of cells expressing the SSR, the general pattern of marking is reproducible between animals. This approach therefore allows for prospective fate mapping, since the initial population of marked cells can be definitively identified 24–48 h following administration of the ligand and then their fate determined. If the timing of marking is important for a study, and/or an ability to definitively identify the marked population, then GIFM is the approach of choice. 2.2.3. Genetic inducible clonal analysis In clonal analysis, the goal is to achieve a level of recombination low enough such that the labeled cells observed can be identified as coming from a single recombined cell (a clone) (Fig. 10.1C). Genetic retrospective clonal analysis

164

Emilie Legue´ and Alexandra L. Joyner

using spontaneous recombination provided new and extremely useful insights about cell behaviors. However, the approach does not allow for temporal control of the marking of the initial cell that produces the clone (Petit et al., 2005). Genetic inducible clonal analysis is an adaptation of the GIFM method, as clonality is achieved by decreasing the dose of ligand injected. In genetic inducible clonal analysis, the approach is partly prospective in the sense that the population from which, and the time when the initial cell is marked can be determined. However, it is also retrospective, since the exact position of the initially labeled cell within the population of cells that could potentially be marked is not known, and therefore must be inferred retrospectively based on analyzing a library of different clones induced at the same time point. 2.2.4. Intersectional genetic fate mapping For intersectional genetic fate mapping, mice must carry three genetically modified alleles: one expressing Cre (or CreER), one expressing Flp (or FlpER), and a reporter allele with two tandem reporters in which the first is both preceded by a loxP (or FRT) flanked stop sequence and surrounded by FRT (or loxP) sites. Although the approach requires that an addition allele be bred onto the experimental mice, the advantage is that more precise marking can be achieved since expression of the second reporter protein depends on the expression of both Cre and Flp in a cell. The approach is also referred to as combinatorial fate mapping, since a combination of Cre and Flp is used for marking (Awatramani et al., 2003; Dymecki and Kim, 2007) (see Chapter 11 for details).

2.3. Important considerations in designing all types of genetic fate mapping studies Initiating a genetic fate mapping study involves the use of at least two transgenic mouse lines. Since generation of transgenic lines is expensive and time-consuming, it is important to carefully choose or design the optimal mouse lines for each study. We first discuss general considerations relevant to all genetic fate mapping studies, and then describe ones specific to using GIFM and genetic inducible clonal analysis approaches, since more sophisticated tools are needed, and as a consequence, more parameters must be taken into account when designing an experiment. 2.3.1. Choice of promoters to drive the expression of a reporter allele A key aspect of genetic fate mapping is knowing whether all cells that have undergone recombination, as well as all their descendants, can be visualized (i.e., expression a marker protein at sufficient levels). Depending on the

Genetic Fate Mapping

165

question being addressed, the stages at which fate-mapped cells are observed, and the method of visualizing the marker protein, either a ubiquitous or restricted promoter can be used to drive the expression of the reporter allele. The use of a ubiquitously expressed promoter, such as the endogenous ROSA promoter (Zambrowicz et al., 1997), allows for all marked cells to be visualized. It is important to note, however, that there may be a few cell types that do not express the ROSA gene, and that all cell types do not express the ROSA gene at the same level. An effective reporter allele for the neuroscience community was generated by targeting a reporter construct into the Tau gene that is expressed only in differentiated neurons (Hippenmeyer et al., 2005). The reporter has the advantage that it expresses both myristoylated GFP (Laux et al., 2000), which marks the axons and dendrites, and nuclear localized b-galactosidase that marks the location of the cell body. Since restricted promoters often have a dynamic pattern of expression and are not always expressed throughout development and into the adult, the ubiquitous promoters have the advantage of providing visualization of all labeled cells at any stage, during development and in the adult. Instead of utilizing gene targeting to engineer reporter alleles that conditionally express a reporter from endogenous loci with known expression patterns, versions of the CAG promoter (Miyazaki et al., 1989; Niwa et al., 1991) have been used to make transgenic mice and multiple lines then screened for a line with broad expression (Akagi et al., 1997). The CAG promoter confers a high level of expression of the reporter gene in many cell types after SSR recombination. Since the complete expression of any of the transgenebased reporter alleles has not been published, it is necessary to test that the reporter expresses in the cells of interest. This can be achieved by deleting the stop sequence in the transgenic reporter in the germ line and observing expression of the reporter in offspring that have the stop sequence removed in all cells (Lallemand et al., 1998). An alternative is to use lines made by gene targeting in which a reporter construct with the CAG promoter has been targeted into the ROSA locus, as these reporter lines have widespread expression, that appears to be high in the postnatal nervous system (Zong et al., 2005). For most studies, ubiquitous reporter alleles are the best, since all the progeny of the marked population of cells can be visualized, regardless of the organ or the cell type they contribute to. However, in some cases, for instance when a study is focused on one cell type, it can be advantageous to use a cell type specific reporter allele in order to limit marking to one cell type. For example, in some cases the marking of multiple cell types in an organ can confound the visualization of the cell type of interest. Finally, the level of expression driven by the promoter in the reporter allele must be taken into account, since the ability to visualize a reporter protein depends on the sensitivity of the reporter detection technique. Different detection techniques are therefore chosen depending on the level of protein expressed. Relatively low levels of expression can be tolerated when fixed

166

Emilie Legue´ and Alexandra L. Joyner

tissues are being analyzed, since detection techniques with high sensitivity can be used, such as immunohistochemical staining that involves an enzymatic reaction that allows for amplification of the signal. However, for live imaging of fate-mapped cells, since direct visualization of fluorescent reporters is used, it is crucial to use promoters that have a high level of expression. 2.3.2. Choice of promoters to drive SSR expression The promoter that drives the SSR determines the population of cells that can be fate mapped. Spatial and/or cell lineage control of the SSR is achieved by choosing the appropriate promoter to drive its expression in the desired population of cells. Of great utility in designing new genetic fate mapping studies, many mouse lines have already been established and are freely available. In order to facilitate the search for existing Cre or CreERT lines, databases have been created: http://nagy.mshri.on.ca/cre/index.php http://www.informatics.jax.org/recombinase.shtml (mouse genome informatics) More general databases can also be used to search for patterns of expression that match the desired population to be fate-mapped. http://www.ncbi.nlm.nih.gov/projects/gensat/ (brain) http://www.brain-map.org/ (brain) http://www.informatics.jax.org/expression.shtml (embryo) http://www.emouseatlas.org/emage/home.php (embryo) Before starting a genetic fate mapping study, it is essential to characterize the pattern of expression of the SSR in the population of cells of interest, and the degree to which the population is marked with the reporter allele of choice (functional assay). To fate map a cell population that expresses a specific gene, it is important to use the line that most faithfully reproduces the pattern of the endogenous gene. Knock-in alleles are normally the best tools for achieving expression that is the same as an endogenous gene. However, for many other applications, transgenes can be an advantage if they express in a subset of the normal expression domain, or have ectopic expression in a cell population of interest. The use of a transgenic line rather than a targeted line requires a more thorough characterization of SSR expression and function to ensure the desired cell types are marked. A possible complication of targeted knock-in alleles, however, is that the allele is often a mutant allele and therefore could have a subtle phenotype in the heterozygous state, especially when combined with other heterozygous mutations, for example, in the reporter allele. Another parameter to test is the level of expression of the SSR. A high level of expression ensures robust labeling of the cell population of interest (i.e., reproducible marking of a high percentage of the cell population).

Genetic Fate Mapping

167

However, for some GIFM applications such as clonal analysis, low levels of expression of the SSR are preferred, since a low efficiency of recombination and no recombination in the absence of ligand is required (see below). 2.3.3. Kinetics of recombination Three key components of the kinetics of site-specific recombination and marking of cells that must be taken into consideration are the time necessary for (1) the SSR to induce recombination, (2) the reporter gene to subsequently be transcribed and translated, and (3) the reporter protein to accumulate sufficiently to be detected. There is usually an approximately 12h delay between the appearance of SSR transcripts and detection of reporter protein (Nakamura et al., 2006). However, this can vary depending on the level of expression of the SSR and reporter allele used, and the sensitivity of detection of the reporter protein. As a good approximation, in cumulative fate mapping studies, the first population of SSR-expressing cells is marked around 12 h after the onset of expression of the SSR. For studies using GIFM and clonal analysis, additional key components are added to the kinetics of recombination and are discussed in the following sections. 2.3.4. Cre toxicity There are a number of publications documenting toxicity produced by the presence of Cre (or CreER) in the nucleus. Expression of nuclear-localized Cre in neural progenitors, under the control of Nestin regulatory elements, was found to cause hydrocephaly and microencephaly in mice carrying two copies of the transgene, and thus expressing high levels of Cre (Forni et al., 2006). Suggesting that the phenotype is due to toxic effects of Cre, rather than the site of insertion of the transgene, the group also found that two Nestin–CreER transgenics have a similar phenotype after tamoxifen administration. However, we have used a Nestin–Cre line (Tronche et al., 1999) and not found phenotypes in mice carrying one copy of the transgene (Blaess et al., 2006, 2008; Corrales et al., 2006). Another study documented that the CreER fusion protein expressed from the ubiquitous ROSA allele leads to widespread apoptosis and anemia when tamoxifen is administered during embryogenesis (Naiche and Papaioannou, 2007). These reports and others point to the importance of testing that the level of expression of Cre or CreER does not cause toxicity that can confound interpretation of a genetic fate mapping study, since toxicity appears to be dependent on the level of Cre/CreER protein present in the nucleus (Forni et al., 2006; Naiche and Papaioannou, 2007; Takebayashi et al., 2008). In vitro studies showed that the Cre cytotoxicity is dependent on the endonuclease activity of Cre (Loonstra et al., 2001). The presence of cryptic loxP sites in mammalian genomes are thought to trigger illegitimate Cre activity, leading to double-strand breaks or nicks that are converted into double-strand breaks

168

Emilie Legue´ and Alexandra L. Joyner

after replication (Abremski et al., 1986), which results in reduced proliferation of the cells that carry damaged DNA (Loonstra et al., 2001). 2.3.5. Mouse breeding At least two mouse lines are needed for genetic fate mapping experiments: the SSR line and the reporter line. An ideal breeding scheme is to use males homozygous for the reporter allele that also carry one copy of the SSR transgene and to cross them to Swiss Webster or other outbred female mice. Half of the litter will carry one copy of each transgene, and can be used for fate mapping analysis. Using outbred females that can be purchased when needed aids in generating the maximum number of transgenic mice for analysis, and in minimizing the size of the mouse colony needed for genetic fate mapping studies. Moreover, in the case of GIFM approaches, since many inbred strains of mice are more sensitive to ligand toxicity (see below), it is advantageous to use outbred females when ligand is administered to pregnant females. 2.3.6. Detection of marking After recombination, the reporter gene will be expressed under the control of the promoter in the reporter allele. Marked cells can then be detected by a variety of methods. The technique chosen to reveal labeled cells depends on the reporter gene and aim of the study. If the reporter encodes a fluorescent protein, and the level of expression is high enough, it can be detected directly by using a fluorescence microscope with filters selecting the appropriate excitation and emission wavelengths. Fluorescent protein reporters thus have the advantage that they can be detected in living cells, making them ideal tools for live-imaging of whole embryos or slices of tissues in culture (see Chapter 20 in volume 476). The other techniques for detecting reporter proteins involve fixation of the tissue. If the reporter gene encodes an enzyme (b-galactosidase or alkaline phosphatase), marked cells can be revealed by an enzymatic reaction. All reporters (enzymes and fluorescent proteins) can be detected by antibody staining. Both enzymatic reactions and antibody staining are sensitive techniques that allow for the detection of relatively low levels of protein. Enzymatic reactions have the advantage that the strength of the signal is directly related to the level of protein and can therefore be maximized by changing the length of staining, whereas antibody staining has the advantage that the signal can be amplified by adding additional steps, including enzymatic reactions. 2.3.6.1. Protocol for detection of b-galactosidase activity by X-gal staining Whole embryos or dissected organs are fixed in 4% paraformaldehyde by immersion at 4  C. The time of fixation depends on the size of the sample. A general guideline is that a 12.5-day embryo should be fixed

Genetic Fate Mapping

169

for 30 min or a postnatal day 28 brain for 45 min. Alternatively, adult mice can be fixed by perfusion (see below). Note, over fixation inhibits enzyme activity, therefore the length of fixation is critical. However, some fixation is necessary to preserve the tissue for sectioning, thus the right balance must be found. The sample is then rinsed twice in PBS at 4  C for 10 min. For whole-mount staining, the tissue is incubated in X-gal reaction solution (PBS, 0.1% NP40, 0.05% deoxycholate, 2 mM MgCl2, 5 mM K3Fe(CN)6, 5 mM K4Fe(CN)6, 0.5 mg/ml, 5-bromo-4-chloro-3-indolyl-b-D-galactoside) for 6–48 h at 30  C. Incubation at 30  C, rather than at 37  C can reduce the background staining from endogenous enzyme (Bonnerot and Nicolas, 1993). The reaction is stopped when the staining is the desired level by rinsing the samples twice in PBS for 10 min. Samples can then be preserved in PBS with azide (0.05%), or 1% PFA. For staining of cryosections (10–20 m), the tissue is immersed in sterile 30% sucrose in PBS at 4  C overnight, and then briefly immersed in OCT compound (TissueTek) before freezing in cold 2-methylbutane (a beaker containing 2-methylbutane is placed in liquid nitrogen) for 2–4 min. After the tissue blocks are sectioned, the slides are placed in PBS at room temperature for 5 min, rinsed twice in X-gal washing solution (PBS, 0.1% NP40, 0.05% deoxycholate, 2 mM MgCl2) for 10 min at 4  C and incubated at 37  C (or 30  C) in X-gal reaction solution for 2–48 h. 2.3.6.2. Protocol for antibody staining For antibody staining, postnatal animals are fixed by intracardiac perfusion, before dissection of organs, with cold PBS first to flush out the blood and then cold 4% PFA in PBS to achieve efficient fixation. Equal volumes of PBS and 4% PFA are perfused. The volume used for perfusion depends on the size of the mouse. Approximately 10–15 ml is used for neonates and 50–60 ml for adults. After dissection of organs, the samples are postfixed in 4% PFA for at least 2 h at 4  C and then prepared for cryosectioning as above. For antibody staining of embryos, they can be immersion fixed in 4% PFA in PBS at 4  C. The length of time the tissue is left in fixation at 4  C depends on the antibody. Also, some antibodies only detect the antigen in paraffin embedded tissue, or when the tissue is fixed with a different fixative such as alcohol. The best conditions must be determined for each antibody. After sectioning, the cryosections are rinsed in PBS, postfixed in 4% PFA for 5 min at room temperature and rinsed twice in PBS. Two types of antibody staining can be used: fluorescent staining for the reporter protein, which has the advantage of allowing for double staining with a second marker and/or DAPI staining to highlight the nuclei, whereas diaminobenzidine (DAB) stained sections have the advantage that they can be examined under bright field microscopy, therefore avoiding the problem of bleaching and/or dimming of the fluorescent signal, since the staining is permanent. For DAB staining that relies on the

170

Emilie Legue´ and Alexandra L. Joyner

peroxidase enzymatic reaction, the endogenous peroxidases must be inhibited by incubating the sections in H2O2 (0.3% in PBS) for 20 min at room temperature, and then rinsing them twice for 10 min in PBS. For fluorescent and DAB staining, the sections are incubated in blocking solution (10% normal serum, 0.25% Triton) for 1 h at room temperature. They are then incubated in primary antibody solution (blocking solution þ primary antibody). For example, a rabbit-anti-GFP antibody can be used (Invitrogen A11122, 1:2000 for DAB or fluorescent staining) or a rabbit-anti-b-galactosidase antibody (ICN 55976, 1:20,000 for DAB, 1:1000 for fluorescent staining) depending on the reporter gene present in the reporter allele and incubated overnight at 4  C. The sections are then rinsed three times in PBS for 10 min at room temperature, and then incubated with the secondary antibody. For fluorescent staining, the secondary antibody is coupled to a fluorophore (Invitrogen) or coupled to biotin (Vector). For DAB staining, the secondary antibody can be directly coupled to the peroxidase (Dakocytomation) or coupled to biotin (Vector). The biotinylated secondary antibody adds an extra step of amplification of the signal. The secondary antibody is incubated for 2 h at room temperature, the concentration of the secondary antibody typically ranges from 1:300 to 1:500. The sections are rinsed three times in PBS. For fluorescent staining using secondary antibodies coupled to biotin, the sections are then incubated with Avidin coupled to a fluorophore for 1 h, and then three times in PBS. Fluorescent slides are coverslipped in Fluoro-Gel (Electron Microscopy Sciences). For DAB staining using a secondary antibody coupled to Biotin, the sections are incubated with Avidin and biotin coupled to peroxidase (Vectastain ABC kit, Vector) for 1 h and then rinsed three times 10 min in PBS. For DAB staining, the sections are then incubated at room temperature with the peroxidase substrate (DAB) for 3–10 min, until the staining is sufficiently strong. The reaction is stopped by rinsing the sections three times in PBS for 10 min at room temperature. DAB is prepared immediately before use (one tablet of 10 mg (Sigma-Aldrich, D5905) in 40 ml of PBS þ 10 l of 30% H2O2). DAB is highly carcinogenous, therefore gloves should be worn when handling DAB, and it must be disposed of according to safety guidelines: for example, DAB waste can be inactivated in permanganate solution (3% KMnO4, 2% Na2CO3 in H2O).

2.4. Special consideration when designing GIFM experiments 2.4.1. Choice of promoters for the inducible SSR and the reporter allele that influence recombination efficiency Since the pattern of expression driven by a specific promoter and regulatory elements often varies during development, it is necessary to assess the expression and function of the inducible SSR at all the time points of ligand

Genetic Fate Mapping

171

administration (induction). Importantly, the general pattern of expression for a given time point is usually consistent between mice, at least with robust regulatory elements driving expression of the SSR. However, the efficiency of recombination with inducible SSRs can vary between animals, even when the same dose of ligand is administered. Thus, for quantitative studies that attempt to compare the percentage of cells labeled at different time points the variability must be taken into consideration, and if possible, the numbers normalized to an internal standard at each time point. Alternatively, the relative proportions of different types of cells that are marked can be compared. Moreover, the efficiency of recombination depends on both the SSR line and the reporter line. Not only is the level of expression of the SSR critical but also the accessibility of the DNA containing the SSR sites can differ between reporter alleles, and possibly in different cell types of a mouse or embryo. It is therefore important to assess the efficiency of recombination for each combination of SSR and reporter line, and at each time point analyzed, since the level of SSR and/or the accessibility of the recombination sites can change over time. For example, we find that with the same SSR line, the R26lox-stoplacZ line (Soriano, 1999) recombines more efficiently than the R26lox-stop-YFP line (Srinivas et al., 2001), and the Taulox-stop-mGFP-IRES-nlacZ reporter (Hippenmeyer et al., 2005) recombines more efficiently than the R26lox-stoplacZ line, at least in neurons (unpublished observations). The dose and the number of administrations of ligand can be adjusted, in order to achieve an ideal (or maximal) level of recombination with a minimum of toxicity (see below). With a high dose of ligand, some combinations of alleles have been reported to achieve marking in 80–90% of the cells in the domain where the SSR is expressed (En2-CreER knockin (Sato and Joyner, 2009; Sgaier et al., 2005); Hoxb6-CreER transgene (Nguyen et al., 2009; Zhu et al., 2008)), whereas other combinations of alleles can achieve as little as 1–5% (our unpublished results). The precise design of the transgene or knock-in allele can dramatically influence the level of SSR expression, and hence the efficiency of cell marking. For example, the presence of neo in a targeted allele can lower expression (Bai and Joyner, 2001; unpublished results), or the type of 30 UTR sequences and polyadenylation sequence used (Kakoki et al., 2004; unpublished observations) can modify the expression level. Thus, it is important to carefully design the transgene or gene targeting construct for an inducible SSR prior to embarking on a GIFM study. 2.4.2. Ligand administration The main technical difference between cumulative genetic fate mapping with a constitutive SSR and GIFM using an inducible SSR-fusion protein is that the latter requires ligand administration.

172

Emilie Legue´ and Alexandra L. Joyner

2.4.2.1. Preparation of the ligand For CreER or FlpER, the ligand is 4-OHT; however, 4-OHT is expensive and very difficult to get into solution. The nonhydroxylated form of tamoxifen is metabolized by the liver into 4-OHT, and is easier to get into solution and much cheaper than 4-OHT. Tamoxifen is therefore used in most in vivo studies in mice. Tamoxifen should be diluted in fresh corn oil. The stock solution is typically 20 mg/ml. Tamoxifen powder (Sigma-Aldrich, T5648-1G) can be first partially dissolved in prewarmed corn oil (e.g., Sigma-Aldrich, C8267-500ML) by vortexing. To dissolve the tamoxifen completely, the solution can then either be sonicated at 60 Hz for approximately 30 min or heated at 37  C overnight. For CrePR the ligand is RU-486. It is suspended in a 0.25% carboxymethyl cellulose, 0.5% Tween 80 solution at a concentration of 50 mg/ml (Casper et al., 2007; Takeuchi et al., 2005). Alternatively, other antiprogestins such as Org 31376 or Org 31806 (N. V. Organon Scientific Development Group) can be used and dissolved in fresh sesame oil at a concentration of 6.5 g/ml (Tsujita et al., 1999). 2.4.2.2. Routes of administration of ligand Tamoxifen dissolved in corn oil can be administered orally by gavage, or by intraperitoneal (i.p.) or subcutaneous (s.c.) injection. The same range of concentrations is used in all three routes. Before administration, the animal should be weighed to adjust the dose administered. For adult inductions and embryonic inductions, gavage or i.p. injections are the most common routes of injection. In the case of embryonic inductions, the pregnant female receives the tamoxifen. For early postnatal stages (P0–P12), s.c. injections of a volume ranging from 30 to 50 ml or gavage with gavaging needles specific for mouse pups (UNO, the Netherlands) are used. S.c. injection is the easiest and safest way to administer tamoxifen at these early stages since the chance of unintentionally damaging essential organs is less than when using i.p. injection. Gavaging neonates is a delicate operation because of their small size and the fragility of their esophagus. Note, however, that the physical properties of the oil make it leak from the point of s.c. injection and thus the volume must be adjusted accordingly. Restraining the pup from moving by holding it still between two fingers and applying pressure on the point of injection helps for achieving a more accurate and reproducible injected volume. The use of a Hamilton syringes with a fine needle is also recommended. 2.4.2.3. Volumes injected The volume injected depends on the desired dose and the weight of the animal. For reproducibility, we recommend that volumes of at least 100 ml are used for i.p. injections and gavage. Before administration, the animals should be weighed and the volume to be injected calculated. If the volume is too small, the concentration of the ligand should be adjusted and the volume recalculated accordingly. Up to

Genetic Fate Mapping

173

500ml can be used to gavage an adult animal and up to 1 ml for an i.p. injection. For s.c. injections in neonates, it is recommended that volumes of approximately 50 ml or less if using a Hamilton syringe are used. 2.4.3. Kinetics of labeling In GIFM studies, it is important to identify the time window during which the recombination occurs. The kinetics of labeling for GIFM studies consist of several components in addition to the time it takes for (1) the SSR to induce recombination, (2) the reporter gene to subsequently be transcribed and translated, and (3) the reporter protein to accumulate sufficiently to be detected. These components are (4) the time for the ligand to be metabolized, in the case of tamoxifen, and to reach the cells and bind to the receptor part of the SSR-fusion protein, (5) the time for the fusion protein to be translocated into the nucleus, and (6) the time during which tamoxifen is present in the mouse tissue of interest and bound to the inducible SSR (in an active state). The time between ligand administration and the first appearance of labeled cells has been described in several papers for CreER and FlpeER (Hayashi and McMahon, 2002; Hunter et al., 2005; Nakamura et al., 2006). Using CreER, the first labeled cells are detected approximately 8 h after tamoxifen administration, but there is a delay of about 12 h before recombination reaches a significant level (Hayashi and McMahon, 2002; Nakamura et al., 2006). With FlpeER, the first recombined cells appear around 12 h after tamoxifen administration (Hunter et al., 2005). The time during which tamoxifen is able to induce CreER to be active in cells initiating new expression of the inducible SSR is approximately 24 h (Kimmel et al., 2000; Nakamura et al., 2006) after administration. All the lengths of time described above are approximate, and can vary depending on the dose of tamoxifen injected, the type of SSR-fusion protein used, and on the level of expression of the SSR. One study detected CreER in the nucleus up to 36 h after tamoxifen administration (Hayashi and McMahon, 2002). The higher the dose, the longer tamoxifen is likely to be present in mouse tissues and able to induce site-specific recombination. When performing embryonic inductions, whether or not the embryo is at a stage when feto-placental circulation has been established might also be a consideration (Nakamura et al., 2006), since drug delivery and clearance of tamoxifen could be different in embryos before and after the establishment of feto-placental circulation. Therefore, the kinetics of recombination might be different depending on the stage that the ligand is administered. The kinetics of labeling also differs depending on the route of administration. Reporter protein expression was found to be more rapid following i.p. injection compared to oral gavage (within 8–12 h compared to 12–18 h) and the peak of activity to occur earlier with i.p. injection (Nguyen et al., 2009).

174

Emilie Legue´ and Alexandra L. Joyner

2.4.4. Toxicity of ligands The use of high doses of synthetic antagonists of steroids such as RU-486 and tamoxifen can induce toxic effects, especially in pregnant females. We have found a high rate of loss of pregnancy 4–6 days after administration of tamoxifen, and an even greater loss of whole litters around the time of birth. When a high dose of tamoxifen is administered (125 g/g of pregnant mouse) between E8.5 and E10.5, we have found in some experiments that as few as 5% of the litters survive past birth (unpublished results). The toxic effect of tamoxifen injection can be partially counteracted by coadministration of progesterone. However, progesterone administration interferes with birth, so it can only be used when embryonic time points are of interest for final analysis. Up to 200 g of tamoxifen/g of mouse can be injected to P2 animals without an observable toxic effect. In adults, up to five daily repeated injections of 250 g/g of mouse (our unpublished results), or up to 16 injections of 145 g/g of mouse every other day (Balordi and Fishell, 2007) can be tolerated. The genetic background of the mouse influences the severity of the toxicity. When possible, outbred mice should be used. Also, lowering the dose of tamoxifen to 50 g/g of mouse can increase survival of litters after birth to 30%. For clonal analysis, a dose of 1 g/g of pregnant mouse does not cause toxic effects (unpublished results). 2.4.5. Initial population Since the expression pattern of the SSR-fusion protein often varies over time, the population of cells that is initially labeled must be determined for each time point of the study. The advantage of GIFM is that the initially marked and defined cells are the only cells that will undergo recombination by the action of the inducible SSR (Fig. 10.1B). To characterize the initial population, the tissue of interest should be collected 24–48 h after the last administration of the ligand. In order to characterize the type of cells within a tissue that are marked in a GIFM study, expression of the reporter gene in the initial population can be compared to cell type specific markers, or spatially restricted markers that provide spatial information. If the goal of the experiment is to fate map cells expressing a particular gene, then the initial population of labeled cells can be compared to the domain of expression of the endogenous gene from which the regulatory sequences were taken to express the inducible SSR. Finally, to test how faithfully the marked cells reflect expression of the inducible SSR, the initial population of labeled cells can be compared to Cre or Flp expression. However, we have found that most commercial antibodies are not effective at detecting Cre, and no Flp antibody is currently available, therefore RNA in situ hybridization analysis of sections or whole-mount embryos/organs must be performed. Comparison of SSR expression, to endogenous gene expression and reporter expression on adjacent sections is the ideal approach.

Genetic Fate Mapping

175

In some cases, the population of cells expressing the inducible SSR is so dynamic that the timing of active recombination can be after the time during which a cell transcribes the inducible SSR. In such a case, some of the labeled population of cells observed 24–36 h after ligand administration will no longer be expressing the SSR (Machold and Fishell, 2005). Finally, in experiments in which a cell-type-specific promoter is used instead of a ubiquitous promoter to drive expression of the reporter protein, it is important to be aware that the initially marked population might not express the reporter gene. It is therefore necessary to use a ubiquitous promoter to determine the initial population of cells that is labeled.

2.5. Special consideration when designing genetic inducible clonal analysis experiments In clonal analysis studies, it is crucial to be able to demonstrate that the labeled cells are indeed clonally related and induced by the administration of the ligand. Two essential parameters have to be tested to validate the use of a particular combination of inducible SSR and reporter alleles for clonal analysis: the absence of labeling of cells in the absence of ligand, and a low enough level of recombination in the presence of ligand to ensure that the labeled groups of cells actually arose from a single cell (are clones). 2.5.1. Choice of SSR and reporter lines Contrary to most genetic fate mapping studies, in clonal analysis, a low efficiency of recombination is an advantage since the labeling must be a rare event. Therefore, lines with low expression levels of an inducible SSR can be used. We and others have found FlpeER to be much less efficient than CreER, thus FlpeER can be very useful for clonal analysis. The first generation CreERT protein (Feil et al., 1996) also induces less recombination that later versions (e.g., CreERT2; Indra et al., 1999), and therefore might be useful for clonal analysis. Another factor that can limit recombination efficiency is the accessibility of the recombination sites in the reporter allele to the SSR. The reporter lines that are less efficiently recombined can therefore be an advantage. Another clonal analysis strategy is to use the interchromosomal reporter alleles (MADM system, see below). In this system, the recombination efficiency is extremely low, allowing for rare labeling events that generate clones (Espinosa and Luo, 2008; Zong et al., 2005). Finally, the use of the Brainbow reporter allele (Livet et al., 2007), and similar next-generation ubiquitous reporters that randomly express multiple different reporter proteins, should be a great advance to clonal analysis.

176

Emilie Legue´ and Alexandra L. Joyner

2.5.2. No recombination in the absence of ligand The first step in developing a clonal analysis protocol is to test whether sitespecific recombination leads to labeling of cells in the absence of ligand with the chosen inducible SSR and reporter lines. Most of the combinations between a CreER line and a reporter line we have tested display some labeled cells in the absence of tamoxifen, that is, in the absence of induction. We term such labeling events as ‘‘noninduced labeling.’’ If the level of noninduced labeling is too high, the allele combination cannot be used for a clonal analysis. To determinate the level of noninduced labeling, a detailed and systematic analysis of reporter expression in the adult organs of interest from several mice that carry both the inducible SSR and reporter alleles must be carried out after administration of oil alone, or in the absence of any treatment. Since the acceptable level of noninduced labeling can be higher for GIFM studies than clonal analysis, it is important to thoroughly test the inducible SSR and reporter allele combination before undertaking a clonal analysis, even with lines that have already been used for GIFM. When using a specific inducible SSR line, it is also important to test the level of noninduced labeling with all the reporter alleles that will be included in the study, since the combination of a given SSR line and different reporter alleles will not necessarily generate the same level of noninduced labeling (unpublished observation). Therefore, the choice of reporter allele is also dictated by the SSR line. 2.5.3. Titration of the dose of ligand administered The optimal dose of ligand must be determined experimentally for each inducible SSR and reporter allele combination at each time point of ligand administration. This is achieved by decreasing the dose of ligand and observing the level of labeling at the time point the final analysis will be carried out. Serial dilutions of ligand should be prepared and tested. Again, it can be beneficial to use an inducible SSR line that is quite inefficient, or a reporter allele with loxP sites that are more difficult to recombine, or a combination of both that ultimately has an inefficient response to ligand administration. If too low a dose of ligand is needed, the dose injected could be close to the threshold necessary to bind to the nuR-modified binding domain and lead to variable labeling results, or no labeling at one dose and too much labeling at the next higher does (unpublished observations). The optimal dose of ligand depends on the organ analyzed. If the pool of cells at the time of induction is small, the probability of hitting a single cell in the pool is low; therefore the appropriate dose for clonal labeling is probably higher than for a larger pool of cells. In addition, it depends on the shape of an organ and behavior of the cells (e.g., the degree of cell mixing) as to how spread out each clone must be to ensure each group of cells is a clone. For

Genetic Fate Mapping

177

example, in order to study cell clones during hair follicle regeneration, since there are millions of hair follicles and no mixing between follicles, a dose that labels one in 10 hair follicles is ideal (Legue´ and Nicolas, 2005). However, there is no way to predict for a particular pair of alleles what the optimal dose will be. It must to be determined experimentally, first grossly and then refined to obtain the most promising dose. A good estimate of doses to first test are 100, 10, and 1 g/g mouse.

3. Future Applications: Combining Genetic Fate Mapping with Mutant Analysis Genetic fate mapping allows the fate of cell populations, or individual cells to be followed in wild-type mice. To understand how cell behaviors are regulated by genes, the next step is to follow the fate of mutant cells, either in a whole mutant mice or in a mosaic analysis in which a mutation is introduced into a subset of cells that are also genetically marked.

3.1. Intrachromosomal recombination: Using existing conditional loxP containing alleles To induce mutant cells by intrachromosomal recombination, conditional mutant alleles made by gene targeting that contain critical exon sequences flanked by two loxP sites (floxed allele) could be used. To date, for such mosaic studies, an SSR line has been combined with a floxed allele (homozygous or over a null mutation at the same locus) and a reporter allele. However, recombination between the two loxP sites in the conditional allele as well as the two loxP sites in the reporter allele does not always occur in the same cell, especially when using inducible SSRs, and by definition in clonal analysis studies. Such an approach therefore leads to the production of three distinct mosaic populations of cells within a mouse: (i) labeled mutant cells, (ii) unlabeled mutant cells, and (iii) labeled wild-type cells. Thus, when using two floxed alleles one cannot assume that all labeled cells are mutant and conversely that all mutant cells are labeled. It is therefore essential to develop new tools to follow the fate of mutant cells created using the many available floxed conditional alleles.

3.2. Interchromosomal recombination: Mosaic analysis with double makers The MADM system offers a way to circumvent the problem using floxed alleles and a separate marker allele, by ensuring that all mutant cells are labeled (Zong et al., 2005). A limitation, however, is that the mutant (often

178

Emilie Legue´ and Alexandra L. Joyner

a null allele) allele of the gene to be studied (gene x) must be on the same chromosome as the MADM reporter alleles. In MADM studies the mutant mouse line for gene x is bred to one of the MADM reporter lines in order to generate a mouse with a chromosome containing the mutant allele of gene x and one MADM reporter line (Fig. 10.4). These mice are then bred to a mouse line that has the complementary MADM allele and a wild-type version of gene x to produce embryos that carry one MADM reporter allele and the mutant allele of gene x on one chromosome and the other MADM reporter allele and a wild-type allele of gene x on the homologous chromosome. After DNA replication in cells, each chromosome will have two copies (one on each chromatid) of its reporter allele and of either the mutant or the wild-type gene x allele. Upon Cre recombination, four chromosomal configurations are generated: the two parental configurations plus two recombined configurations, one with a full-length GFP and on the same chromosome as the mutant allele of gene x and one with a full-length RFP and a wild-type allele of gene x. Depending on the segregation of the chromatids during mitosis, two scenarios are possible. One scenario generates two cells that are labeled, one that expresses GFP and that is homozygous mutant for gene x and one that expresses RFP and is wild-type for gene x. The other scenario generates one cell that expresses both GFP and RFP, and a cell that is not labeled. Each cell in this scenario is heterozygous for gene x. By studying the products of recombination events that generate complementary GFP and RFP clones (mutant and wild type), the behaviors of mutant cells in an otherwise wild-type tissue can be analyzed. The approach also provides a direct comparison of cell behaviors between wild-type and mutant cells that are generated and labeled concomitantly. Two limiting factors of this system are that (1) cell division is needed to generate the recombination, and thus labeling and mutagenesis only can occur in dividing cells expressing active Cre, and (2) the mutated gene of interest must be on the same chromosome that carries the markers.

REFERENCES Abremski, K., et al. (1986). Bacteriophage P1 Cre-loxP site-specific recombination. Sitespecific DNA topoisomerase activity of the Cre recombination protein. J. Biol. Chem. 261, 391–396. Akagi, K., et al. (1997). Cre-mediated somatic site-specific recombination in mice. Nucleic Acids Res. 25, 1766–1773. Andrews, B. J., et al. (1985). The FLP recombinase of the 2 micron circle DNA of yeast: Interaction with its target sequences. Cell 40, 795–803. Austin, S., et al. (1981). A novel role for site-specific recombination in maintenance of bacterial replicons. Cell 25, 729–736. Awatramani, R., et al. (2003). Cryptic boundaries in roof plate and choroid plexus identified by intersectional gene activation. Nat. Genet. 35, 70–75.

Genetic Fate Mapping

179

Bai, C. B., and Joyner, A. L. (2001). Gli1 can rescue the in vivo function of Gli2. Development 128, 5161–5172. Balordi, F., and Fishell, G. (2007). Mosaic removal of hedgehog signaling in the adult SVZ reveals that the residual wild-type stem cells have a limited capacity for self-renewal. J. Neurosci. 27, 14248–14259. Blaess, S., et al. (2006). Sonic hedgehog regulates Gli activator and repressor functions with spatial and temporal precision in the mid/hindbrain region. Development 133, 1799–1809. Blaess, S., et al. (2008). Gli3 coordinates three-dimensional patterning and growth of the tectum and cerebellum by integrating Shh and Fgf8 signaling. Development 135, 2093–2103. Bonnerot, C. and Nicolas, J. F. (1993). Application of LacZ gene fusions to postimplantation development. Methods Enzymol. 225, 451–469. Buchholz, F., et al. (1996). Different thermostabilities of FLP and Cre recombinases: Implications for applied site-specific recombination. Nucleic Acids Res. 24, 4256–4262. Buchholz, F., et al. (1998). Improved properties of FLP recombinase evolved by cycling mutagenesis. Nat. Biotechnol. 16, 657–662. Casper, K. B., et al. (2007). Characterization of astrocyte-specific conditional knockouts. Genesis 45, 292–299. Chudakov, D. M., et al. (2005). Fluorescent proteins as a toolkit for in vivo imaging. Trends Biotechnol. 23, 605–613. Corrales, J. D., et al. (2006). The level of sonic hedgehog signaling regulates the complexity of cerebellar foliation. Development 133, 1811–1821. Dymecki, S. M., and Kim, J. C. (2007). Molecular neuroanatomy’s ‘‘Three Gs’’: A primer. Neuron 54, 17–34. Espinosa, J. S., and Luo, L. (2008). Timing neurogenesis and differentiation: Insights from quantitative clonal analyses of cerebellar granule cells. J. Neurosci. 28, 2301–2312. Farley, F. W., et al. (2000). Widespread recombinase expression using FLPeR (flipper) mice. Genesis 28, 106–110. Feil, R., et al. (1996). Ligand-activated site-specific recombination in mice. Proc. Natl. Acad. Sci. USA 93, 10887–10890. Feil, R., et al. (1997). Regulation of Cre recombinase activity by mutated estrogen receptor ligand-binding domains. Biochem. Biophys. Res. Commun. 237, 752–757. Fields-Berry, S. C., et al. (1992). A recombinant retrovirus encoding alkaline phosphatase confirms clonal boundary assignment in lineage analysis of murine retina. Proc. Natl. Acad. Sci. USA 89, 693–697. Forni, P. E., et al. (2006). High levels of Cre expression in neuronal progenitors cause defects in brain development leading to microencephaly and hydrocephaly. J. Neurosci. 26, 9593–9602. Golden, J. A., et al. (1995). Construction and characterization of a highly complex retroviral library for lineage analysis. Proc. Natl. Acad. Sci. USA 92, 5704–5708. Hayashi, S., and McMahon, A. P. (2002). Efficient recombination in diverse tissues by a tamoxifen-inducible form of Cre: A tool for temporally regulated gene activation/ inactivation in the mouse. Dev. Biol. 244, 305–318. Hippenmeyer, S., et al. (2005). A developmental switch in the response of DRG neurons to ETS transcription factor signaling. PLoS Biol. 3, e159. Hoess, R. H., et al. (1982). P1 site-specific recombination: Nucleotide sequence of the recombining sites. Proc. Natl. Acad. Sci. USA 79, 3398–3402. Hunter, N. L., et al. (2005). Ligand-activated Flpe for temporally regulated gene modifications. Genesis 41, 99–109. Indra, A. K., et al. (1999). Temporally-controlled site-specific mutagenesis in the basal layer of the epidermis: Comparison of the recombinase activity of the tamoxifen-inducible Cre-ER(T) and Cre-ER(T2) recombinases. Nucleic Acids Res. 27, 4324–4327.

180

Emilie Legue´ and Alexandra L. Joyner

Joyner, A. L., and Zervas, M. (2006). Genetic inducible fate mapping in mouse: Establishing genetic lineages and defining genetic neuroanatomy in the nervous system. Dev. Dyn. 235, 2376–2385. Kakoki, M., et al. (2004). Altering the expression in mice of genes by modifying their 30 regions. Dev. Cell. 6, 597–606. Kimmel, R. A., et al. (2000). Two lineage boundaries coordinate vertebrate apical ectodermal ridge formation. Genes Dev. 14, 1377–1389. Kolb, A. F. (2001). Selection-marker-free modification of the murine beta-casein gene using a lox2272 [correction of lox2722] site. Anal. Biochem. 290, 260–271. Lallemand, Y., et al. (1998). Maternally expressed PGK-Cre transgene as a tool for early and uniform activation of the Cre site-specific recombinase. Transgenic Res. 7, 105–112. Laux, T., et al. (2000). GAP43, MARCKS, and CAP23 modulate PI(4, 5)P(2) at plasmalemmal rafts, and regulate cell cortex actin dynamics through a common mechanism. J. Cell Biol. 149, 1455–1472. Lawson, K. A. (1999). Fate mapping the mouse embryo. Int. J. Dev. Biol. 43, 773–775. Legue´, E., and Nicolas, J. F. (2005). Hair follicle renewal: Organization of stem cells in the matrix and the role of stereotyped lineages and behaviors. Development 132, 4143–4154. Livet, J., et al. (2007). Transgenic strategies for combinatorial expression of fluorescent proteins in the nervous system. Nature 450, 56–62. Lobe, C. G., et al. (1999). Z/AP, a double reporter for cre-mediated recombination. Dev. Biol. 208, 281–292. Loonstra, A., et al. (2001). Growth inhibition and DNA damage induced by Cre recombinase in mammalian cells. Proc. Natl. Acad. Sci. USA 98, 9209–9214. Luche, H., et al. (2007). Faithful activation of an extra-bright red fluorescent protein in ‘‘knock-in’’ Cre-reporter mice ideally suited for lineage tracing studies. Eur. J. Immunol. 37, 43–53. Machold, R., and Fishell, G. (2005). Math1 is expressed in temporally discrete pools of cerebellar rhombic-lip neural progenitors. Neuron 48, 17–24. Metzger, D., and Chambon, P. (2001). Site- and time-specific gene targeting in the mouse. Methods 24, 71–80. Miyazaki, J., et al. (1989). Expression vector system based on the chicken beta-actin promoter directs efficient production of interleukin-5. Gene 79, 269–277. Naiche, L. A., and Papaioannou, V. E. (2007). Cre activity causes widespread apoptosis and lethal anemia during embryonic development. Genesis 45, 768–775. Nakamura, E., et al. (2006). Kinetics of tamoxifen-regulated Cre activity in mice using a cartilage-specific CreER(T) to assay temporal activity windows along the proximodistal limb skeleton. Dev. Dyn. 235, 2603–2612. Nguyen, M. T., et al. (2009). Tamoxifen-dependent, inducible Hoxb6CreERT recombinase function in lateral plate and limb mesoderm, CNS isthmic organizer, posterior trunk neural crest, hindgut, and tailbud. Dev. Dyn. 238, 467–474. Niwa, H., et al. (1991). Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene 108, 193–199. Novak, A., et al. (2000). Z/EG, a double reporter mouse line that expresses enhanced green fluorescent protein upon Cre-mediated excision. Genesis 28, 147–155. Pakhomov, A. A., and Martynov, V. I. (2008). GFP family: Structural insights into spectral tuning. Chem. Biol. 15, 755–764. Petit, A. C., Legue´, E. and Nicolas J. F. (2005). Methods in clonal analysis and applications. Reprod. Nutr. Dev. 45, 321–339. Raymond, C. S., and Soriano, P. (2007). High-efficiency FLP and PhiC31 site-specific recombination in mammalian cells. PLoS One 2, e162. Sanes, J. R., et al. (1986). Use of a recombinant retrovirus to study post-implantation cell lineage in mouse embryos. EMBO J. 5, 3133–3142.

Genetic Fate Mapping

181

Sato, T., and Joyner, A. L. (2009). The duration of Fgf8 isthmic organizer expression is key to patterning different tectal-isthmo-cerebellum structures. Development 136, 3617–3626. Sauer, B. (1993). Manipulation of transgenes by site-specific recombination: Use of Cre recombinase. Methods Enzymol. 225, 890–900. Schaft, J., et al. (2001). Efficient FLP recombination in mouse ES cells and oocytes. Genesis 31, 6–10. Sgaier, S. K., et al. (2005). Morphogenetic and cellular movements that shape the mouse cerebellum; insights from genetic fate mapping. Neuron 45, 27–40. Soriano, P. (1999). Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat. Genet. 21, 70–71. Srinivas, S., et al. (2001). Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev. Biol. 1, 4. Takebayashi, H., et al. (2008). Tamoxifen modulates apoptosis in multiple modes of action in CreER mice. Genesis 46, 775–781. Takeuchi, T., et al. (2005). Control of synaptic connection by glutamate receptor delta2 in the adult cerebellum. J. Neurosci. 25, 2146–2156. Tronche, F., et al. (1999). Disruption of the glucocorticoid receptor gene in the nervous system results in reduced anxiety. Nat. Genet. 23, 99–103. Tsujita, M., et al. (1999). Cerebellar granule cell-specific and inducible expression of Cre recombinase in the mouse. J. Neurosci. 19, 10318–10323. Vegeto, E., et al. (1992). The mechanism of RU486 antagonism is dependent on the conformation of the carboxy-terminal tail of the human progesterone receptor. Cell 69, 703–713. Wu, Y., et al. (2009). High-efficient FLPo deleter mice in C57BL/6J background. PLoS One 4, e8054. Zambrowicz, B. P., et al. (1997). Disruption of overlapping transcripts in the ROSA beta geo 26 gene trap strain leads to widespread expression of beta-galactosidase in mouse embryos and hematopoietic cells. Proc. Natl. Acad. Sci. USA 94, 3789–3794. Zhu, J., et al. (2008). Uncoupling Sonic hedgehog control of pattern and expansion of the developing limb bud. Dev. Cell. 14, 624–632. Zong, H., et al. (2005). Mosaic analysis with double markers in mice. Cell 121, 479–492.

C H A P T E R

E L E V E N

Mapping Cell Fate and Function Using Recombinase-Based Intersectional Strategies Susan M. Dymecki, Russell S. Ray, and Jun C. Kim Contents 1. Introduction 2. Accessing Mouse Embryonic Cells In Utero for Tracer Molecule ‘‘Delivery’’ Using Transgenesis and Site-Specific DNA Recombination 3. Improving Cell-Subtype Selectivity in Genetic Fate Maps Using a Dual-Recombinase Intersectional Method 3.1. Flp variants for use in combination with Cre 3.2. Target indicator transgenes 4. Transgenes Enabling Subtractive as well as Intersectional Genetic Fate Mapping 5. Exploiting Different Reporter Molecules to Reveal Different Features of Mapped Cell Populations 6. 3 for 1 7. Intersectional Transgene Activation Reaches from Mapping Cell Fate to Mapping Cell Function 8. Methods and Materials 8.1. Comparing recombinase versus endogenous driver gene expression by in situ hybridization 8.2. Comparing the spatiotemporal profile of recombinase expression to that of actual recombinase activity 8.3. Identifying progenitor cells and their descendants by enzymatic or immunofluorescence detection of the lineage tracer molecules b-gal or GFP 8.4. Codetection of subtractive and intersectional lineage tracer molecules 9. Concluding Remarks Acknowledgments References

Department of Genetics, Harvard Medical School, Boston, Massachusetts, USA Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77011-7

#

184

185 189 190 190 193 195 196 196 199 199 204

205 208 208 209 209

2010 Elsevier Inc. All rights reserved.

183

184

Susan M. Dymecki et al.

Abstract Cell types are typically defined by expression of a unique combination of genes, rather than a single gene. Intersectional methods therefore become crucial to selectively access these cells for higher resolution fate mapping and functional manipulations. Here, we discuss one such intersectional method. Two recombinase systems (Cre/loxP and Flp/FRT) work together to remove a double STOP cassette and thereby activate expression of a target transgene solely in cells defined by a particular pairwise combination of driver genes. Depending on the nature of the target transgene, this strategy can be used to deliver cell-lineage tracers, sensors, and/or effector molecules to highly selective cell types in vivo. In this chapter, we discuss concepts, reagents, and methods underlying this intersectional approach and encourage consideration of various intersectional and binary methods for accessing uniquely defined cell subsets in the mouse.

1. Introduction For any map—whether it be a road map, a genetic cell-fate map, or a functional neuronal circuitry map—improving resolution improves accuracy of the mapped relationships and thus the usefulness of the information conveyed. For murine cell-fate or cell-function maps, one means to improve resolution is by directing expression of the cell-lineage tracer or cell effector molecule, respectively, to smaller more uniquely defined subsets of cells. This allows a tighter correspondence to be drawn, for example, between a specific progenitor cell subtype and the kinds of progeny cells it produces, or between a specific neuron subtype in the adult brain and its contribution to a specific animal behavior. In this chapter, we discuss a general approach for driving conditional expression of a reporter or effector molecule with high cell-subtype selectivity in the mouse—called intersectional transgene activation (Awatramani et al., 2003; Branda and Dymecki, 2004; Farago et al., 2006; Jensen et al., 2008; Kim et al., 2009)—and we detail how this strategy can be applied to enhance cell-subtype resolution in murine genetic fate maps and in functional neuronal circuitry maps. In brief, order-of-magnitude improvements in map resolution can be achieved simply through increasing, by one, the conditions on which cell subtypes are selected for genetic manipulation: instead of selecting cells based on their capacity to drive the expression of one particular gene, as is the case for most conventional transgenics, they are now selected based on a capacity to drive expression of pairwise gene combinations. In other words, transgeneencoded tracer or effector molecule expression is directed exclusively to cells residing within the intersection of the patterns driven by two independent promoter/enhancer fragments (envision the partial overlap of two circles in a Venn diagram). Here, we discuss concepts, reagents, and

Intersectional Genetic Fate Mapping

185

methods underlying this ‘‘intersectional’’ approach, beginning first by introducing the basics of genetic fate mapping (previously reviewed in Branda and Dymecki, 2004; Dymecki and Kim, 2007; Dymecki et al., 2002; Joyner and Zervas, 2006) and the recombinase enzymes involved and follow with implementation of currently available intersectional fate mapping and effector alleles. Relevant methods and materials are detailed in Section 8 and are applicable across many kinds of recombinase-based genetic experiments.

2. Accessing Mouse Embryonic Cells In Utero for Tracer Molecule ‘‘Delivery’’ Using Transgenesis and Site-Specific DNA Recombination Genetic fate mapping, being noninvasive and transgene-based, has opened wide the door to fate mapping studies in mice. Hindering murine studies in the past has been the access barrier of in utero development, making it difficult, if not impossible, to inject mouse embryos with lineage tracers (Cepko et al., 1990; Galileo et al., 1990; Keller, 1975; Walsh and Cepko, 1988; Wetts and Fraser, 1988) or cell grafts (Le Douarin, 1982) without also disturbing development—the very process under investigation. No longer though is physical access in utero required for lineage tracer delivery. Instead, genetic access can be exploited. Through transgenesis and the clever deployment of genetically encoded site-specific recombinase enzymes and reporter molecules, lineage tracer expression can be directed (‘‘delivered’’) to discrete cells in the developing mouse embryo—noninvasively, reproducibly, and conveniently. The fate of cells and/or their progeny cells can then be explored and their contributions to different tissues examined. Gaining genetic access to mouse cells, whether embryonic or adult, involves transgenesis, the incorporation of an exogenous DNA construct—a transgene—into the mouse genome. Transgenes are typically designed to express a specific genetically encoded molecule, that is, a reporter or effector molecule, in a cell-restricted fashion. Cell-type restriction is determined largely by the gene promoter and enhancer elements incorporated into the transgene; these elements bestow the needed genetic access to specific cell types, and they do so noninvasively and reproducibly. In practice, cell-type specific enhancers can be isolated and directly incorporated into a transgene to drive restricted expression of the desired reporter- or effector-encoding sequence (Nagy et al., 2003); alternatively, the reporter- or effector-encoding sequence can be inserted into a bacterial artificial chromosome (BAC) that contains the relevant constellation of enhancer sequences (to generate a BAC transgene) (Gong et al., 2007;

186

Susan M. Dymecki et al.

Heintz, 2001; Johansson et al., 2010; Lee et al., 2001; Muyrers et al., 2001); or the reporter- or effector-encoding sequence can be introduced into a capable endogenous gene locus, for example, by homologous recombination in embryonic stem cells (a knockin transgene) thereby co-opting the complete repertoire of DNA regulatory elements relevant to that gene expression profile (Nagy et al., 2003). For genetic fate mapping, these various transgenesis approaches are exploited to two effects. One effect is to drive expression of a genetically encoded site-specific recombinase in a cell-type specific fashion. The site-specific recombinase is typically either Cre recombinase (causes recombination of the bacteriophage P1 genome) or the high-efficiency Flp recombinase variants (Flp named for its ability to ‘‘flip’’ a DNA segment in Saccharomyces cerevisiae, variants discussed below) (reviewed in Branda and Dymecki, 2004; Dymecki, 2000; Dymecki and Kim, 2007; Joyner and Zervas, 2006; Lewandoski, 2001; Nagy, 2000; Stark et al., 1992). The other purpose for which transgenesis is exploited is to create a reporterencoding transgene—one that is configured in a functionally ‘‘off’’ state but which can be switched ‘‘on’’ for constitutive reporter expression by Cre- or Flp-mediated excisional recombination (Fig. 11.1A (a–c)) (Akagi et al., 1997; Awatramani et al., 2001; Branda and Dymecki, 2004; Farago et al., 2006; Jensen et al., 2008; Lobe et al., 1999; Madisen et al., 2010; Novak et al., 2000; Rodriguez and Dymecki, 2000; Soriano, 1999; Srinivas et al., 2001; Zinyk et al., 1998). The product of the first transgene mentioned above (the recombinase ‘‘driver’’ transgene; Fig. 11.1A(b)) activates, in a cell-type restricted manner, functional expression of the second transgene (the reporter- or effector-encoding transgene; Fig. 11.1A(a, c)). Activation is achieved through recombinase-mediated excision of an ‘‘off’’ or ‘‘STOP’’ portion of the reporter transgene. Transcription (cartooned as squiggle arrow in Fig. 11.1), now no longer blocked by ‘‘STOP’’ sequences, proceeds through the reporter-encoding region of the transgene enabling reporter synthesis and cell labeling. Use of broadly active promoter and enhancer sequences (BAP in Fig. 11.1) in this transgene (by contrast to the cell-type specific elements used to drive the recombinase transgene) allow for most cell types to be marked by reporter expression. This is an important feature because all daughter cell types will inherit the recombined (‘‘activated’’) reporter transgene (Fig. 11.1A(c, d)) from their parental progenitor cell and ideally all daughter cells will be labeled, and thus become trackable, regardless of the final differentiated fate assumed and independent of any further recombinase action—reporter expression is effectively switched-on and stays on. In this way, reporter molecule expression has been turned into an indelible cell-lineage tracer (Fig. 11.1A(c, d)). Thus, a simple site-specific DNA excision reaction and two transgenes—a cell-restricted recombinase transgene and an ‘‘activatable’’ reporter transgene with the potential to mark virtually any cell type—partner to ‘‘deliver’’ a cell-lineage tracer to a

187

Intersectional Genetic Fate Mapping

Single-recombinase based genetic fate mapping

A

(e)

(a) BAP

mCherry

mCherry

BAP

eGFP

(f)

(b)

Gene A::Cre

Gene A::Cre

(g)

(c) BAP

(d)

Dual-recombinase based genetic fate mapping

B

BAP

mCherry

eGFP

mCherry

(h) Gene B ::FlpO

(i) BAP

eGFP

(j) n

latio

opu nal p

ectio

Inters

Subtractive population

Figure 11.1 Single- versus dual-recombinase-based genetic fate mapping. (A) Illustration of how site-specific recombination (typically mediated by either Cre/loxP or Flp/FRT) can be used to study the deployment of progenitor cells and their descendants during development. We refer to this as single-recombinase-based genetic fate mapping; only one recombinase system is employed, as cartooned here for Cre/loxP, but the general method holds similarly for Flp/FRT. (a) A Cre-responsive indicator transgene comprised of a broadly active promoter and set of cis-regulatory sequences (BAP) separated from reporter-encoding sequence (e.g., mCherry) by loxP (blue triangle)-flanked STOP sequences (red outlined hexagon). Transcription terminates at the STOP, reporter is not expressed, cells are not marked. (b) cis-Regulatory sequences from hypothetical Gene A are employed in the GeneA::cre transgene to restrict Cre expression to Gene A-expressing progenitor cells (blue stripes in gray cartoon of embryonic brain). (c) Recombined indicator transgene present in the GeneA::creexpressing cells; upon Cre-excision of the loxP-flanked STOP sequences mCherry is expressed and marks the Gene A cell population (red stripes in gray cartoon of embryonic brain). Ideally, activation of reporter molecule expression (e.g., mCherry) is permanent, and all cells descended from the GeneA::cre-expressing progenitors will

188

Susan M. Dymecki et al.

targeted cell population in the embryo. Daughter cell fates can then be tracked to later embryonic or adult stages. Physical access is not required. In utero development is not disturbed. Reporter molecule levels are not diluted by repeated cell divisions. Because the initially labeled cell population in genetic fate mapping is defined and selected based on genetic criteria—by the gene or enhancer activity used to drive recombinase expression—advantages beyond noninvasiveness are rendered. Great precision and reproducibility in lineage tracer delivery can be achieved even in the face of a targeted cell population that is highly distributed and intermingled among other cells. Further, the resultant map, by linking an expressed gene to a particular cell lineage and fate, can, in some instances, pave the way toward identifying gene products that may play an important role in the specification and differentiation of that particular cell type. As well, the effects of various gene mutations on cell fate can be ascertained directly—genetic fate maps can be generated in mutant tissues, making it possible to test how certain gene modifications alter the migratory routes or ultimate fates of specific cells. Depending on the nature of the gene regulatory elements used to drive recombinase (Cre or Flp variant) expression, the initially labeled cell

continue expressing the reporter, thereby marking a genetic lineage as it contributes to different brain regions during development. Descendant cells are depicted in (d) as red dots or cartooned red neurons in the enlargement. (B) Illustration of how the Cre/loxP and Flp/FRT systems can be used together to remove a double STOP cassette and thereby activate expression of a target indicator transgene solely in cells defined by a particular pairwise combination of driver genes. (e) An example of a dual-recombinaseresponsive indicator allele comprised of a broadly active promoter and set of cisregulatory sequences (BAP) separated from GFP encoding sequence by two STOP cassettes: the first flanked by loxP sites (blue triangle) and the second flanked by FRT sites (yellow semiovals)—this latter cassette also contains mCherry encoding sequence. (f) As in (b), cis-regulatory sequences from hypothetical Gene A are employed in the GeneA::cre transgene to restrict Cre expression to Gene A-expressing progenitor cells (blue stripes in gray cartoon of embryonic brain). (g) Cre-recombined indicator transgene, driving reporter mCherry expression in Gene A::cre-expressing cells (red stripes in gray cartoon of embryonic brain). The remaining STOP sequences prevent GFP expression. (h) cis-Regulatory sequences from hypothetical Gene B are used in the GeneB::Flpo transgene to restrict Flpo protein (yellow cross) to Gene B-expressing progenitor cells (yellow areas in cartoon of embryonic brain). (i) Fully recombined indicator transgene present in cells having undergone both Cre and Flpo excisional recombination, thus marking by GFP expression those progenitor cells lying at the intersection of Gene A and Gene B (small green patches in cartoon of embryonic hindbrain). Descendant cells—the intersectional population—are depicted in ( j) as green dots or cartooned as green neurons in the enlargement. In this way, smaller more uniquely defined subsets of cells can be marked and studied. In this cartooned indicator transgene, the Cre-only cells express mCherry and are cartooned here in red; we refer to these cells as the subtractive population. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this chapter.)

Intersectional Genetic Fate Mapping

189

population can be either highly specific or (confusingly) inclusive of multiple progenitor cell subtypes (Fig. 11.1A). While the ideal is to use as driver a gene that offers adequate cell-type restriction in expression, such choices may not always be available or exist for the cell population under study. Indeed, in many cases the expression pattern of a single gene spreads over multiple progenitor cell territories across different regions of the nervous system. It can be difficult, if not impossible, to assign unambiguously particular daughter cell fates to specific subsets of progenitor cells. In other words, a single gene expression pattern often is too broad, spatially and temporally, to exclusively resolve the desired progenitor cell pool of interest. Prior work on a given gene may have identified separate enhancer modules each responsible for a discrete portion of the overall expression profile, such that an individual enhancer can be used in isolation to drive and limit recombinase expression to a smaller subset of cells. Such information, though, is available for a relatively small fraction of genes, albeit growing. An alternative approach toward improving cell-subtype selectivity, and thus fate map resolution, is to employ an intersectional strategy where cell labeling (reporter activation) occurs only in those cells situated in the overlap of the patterns driven by two genes (Fig. 11.1B) (Awatramani et al., 2003; Dymecki and Kim, 2007; Farago et al., 2006; Jensen et al., 2008). In this way, lineage resolution can be improved substantially.

3. Improving Cell-Subtype Selectivity in Genetic Fate Maps Using a Dual-Recombinase Intersectional Method For intersectional genetic fate mapping, two site-specific recombinases rather than one are required to activate a reporter transgene (Fig. 11.1B) (Awatramani et al., 2003; Farago et al., 2006; Dymecki and Kim, 2007; Jensen et al., 2008; Miyoshi et al., 2010; Yamamoto et al., 2009). Two STOP cassettes, disrupting the reporter transgene, must be removed: one STOP cassette is loxP-flanked and removable by Cre, the other is FRTflanked and removable by Flpe (Fig. 11.1B). Thus only cells contained in the overlap between the expression patterns driven by the Cre and the Flpe transgenes are selected for fate mapping. This cell population is referred to as the intersectional population (Awatramani et al., 2003; Farago et al., 2006) (Fig. 11.1B(i) green GFPþ population). Their descendant cells too are marked (e.g., by GFP expression) by virtue of inheriting the ‘‘activated’’ dual-recombined and thus constitutively expressed reporter transgene. Next, we discuss practical considerations surrounding recombinase driver and target reporter-encoding transgenes as relates to intersectional fate mapping. We also describe how a second population, called the

190

Susan M. Dymecki et al.

‘‘subtractive’’ population (Farago et al., 2006; Jensen et al., 2008), can be marked simultaneous to the intersectional population.

3.1. Flp variants for use in combination with Cre Cre and Flp are the principle site-specific recombinases employed to date in murine intersectional approaches. Worth noting are the different variants of Flp recombinase that can be used in mice. They are referred to as wild-type Flp (Flp-wt), low-activity Flp (FlpL), enhanced Flp (Flpe), and optimized Flpe (referred to as Flpo). The latter two, because they exhibit high efficiency, are recommended for most fate mapping applications. The enhanced Flpe contains four point mutations that together confer increased protein thermostability resulting in up to a 10-fold increase in recombinase activity as compared to Flpwt (Buchholz et al., 1996, 1998; Rodriguez et al., 2000). We have generated multiple Flpe transgenics (conventional and BAC) harboring excellent recombinase efficiency, capturing most if not all cells within a gene expression domain (Awatramani et al., 2003; Farago et al., 2006; Jensen et al., 2008; Landsberg et al., 2005; Ray and Dymecki, 2009; Rodriguez et al., 2000). Recently, Flpe has been codon-optimized (Raymond and Soriano, 2007), successfully tackling the goal of boosting Flpe activity in mouse embryonic stem cells—a cell population for which even Flpe, despite its excellent activity in somatic (Awatramani et al., 2003; Farago et al., 2006; Jensen et al., 2008; Landsberg et al., 2005) and germ cells (Rodriguez et al., 2000), has proven less effective than Cre (Schaft et al., 2001). Indeed, Flpo appears to be about five times more effective than Flpe (Kranz et al., 2010; Wu et al., 2009). Flpo expression vectors (Raymond and Soriano, 2007) are readily available from AddGene (www.addgene.org; plasmid 13793, pPGKFLPopA). Whichever recombinases are employed, it is critical to establish the extent to which each recombinase expression mirrors spatiotemporally the expression profile of the chosen driver gene or driver enhancer elements. This is because unexpected recombinase expression could lead to erroneous inclusion of cells into the fate map that are actually unrelated genetically, while unexpectedly limited recombinase expression could lead to erroneous exclusion of cells that are in fact related genetically. Relevant methods to assess recombinase and endogensous driver gene expression are presented below.

3.2. Target indicator transgenes While the intersectional selection of cells for study exploits cell type-restricted expression of both Cre and Flpe (or Flpo) and their expression overlap, actual labeling and fate mapping of the selected cell population involves dualrecombinase-mediated rearrangement of a reporter-encoding transgene into a configuration that sustains constitutive reporter expression. The three key elements of this target transgene from 30 to 50 are (1) a gene encoding the

Intersectional Genetic Fate Mapping

191

desired intersectional reporter (e.g., GFP); (2) two STOP cassettes, each capable of preventing reporter expression; and (3) promoter/enhancer sequences used to drive reporter expression following recombination. We first consider elements two and three and later touch on element one. STOP cassettes effective at blocking transcription are critical for achieving conditional cell-subtype selective ‘‘delivery’’ of the intersectional reporter molecule. Reporter expression must be effectively ‘‘off’’ until both STOP cassettes are removed, at which point reporter expression switches ‘‘on’’ in the cell population of interest. Thus each of the STOP cassettes, by themselves, must be able to effectively stop transcription. We have characterized three STOP cassettes and list them here in descending order of efficiency (J. C. Kim and S. M. Dymecki, unpublished data): (1) the lox2 cassette (Sauer, 1993), containing SV40 intron and polyadenylation (pA) signal sequences, a gratuitous ATG translation start, and 50 splice donor signal; (2) a concatemer of SV40 pA sequences (Awatramani et al., 2001; Farago et al., 2006; Lobe et al., 1999; Soriano, 1999); and (3) a concatemer of bovine growth hormone pA sequences (Awatramani et al., 2001; Farago et al., 2006). We have observed the tightest off-to-on switch upon employing the first two cassettes in our intersectional alleles (Table 11.1), and therefore recommend their preferential use. In practice, it is essential that the unrecombined reporter transgene (Fig. 11.1B(e)), as well as transgene versions in which only one of the two STOP cassettes remains, be examined for any leaky unwanted reporter expression in tissues of interest. If leaky expression is present yet left undiscovered, these marked (recombinase-independent) cells will be included erroneously in the fate map. Sensitive methods for detecting reporter molecule expression are presented in Section 8. Ideally, the target reporter transgene should have the potential to drive detectable levels of reporter molecule in all types of cells regardless of their identity or stage, thus offering the capacity to map, following dual recombination, all progeny cell types arising from a given progenitor population. No daughter cell types should be missed because of low, or no, reporter transcription. This is a tall order, though, and one yet to be fully achieved. But a number of promoter/enhancer combinations come close. One strategy that we (Farago et al., 2006; Jensen et al., 2008; Kim et al., 2009) and others (Madisen et al., 2010; Miyoshi et al., 2010; Zong et al., 2005) have employed uses the chicken b-actin promoter, cytomegalovirus enhancer sequence (when paired together, they are referred to as CAG; Niwa et al., 1991), and regulatory sequences from the endogenous mouse Gt(ROSA)26Sor (R26) locus (Zambrowicz et al., 1997). This R26/CAG partnership (we refer to as RC in our intersectional transgenes presented in Table 11.1) can offer, depending on cell type, improved levels of expression by comparison to either R26 or CAG alone, especially in postnatal brain. Other exploited loci include tau (Kramer et al., 2006) for nervous system studies, and colA1 (Beard et al., 2006). To enhance mRNA transcript stability and thus possibly enhance reporter

Table 11.1 Dual Cre/Flp-repsonsive intersectional alleles and single recombinase-responsive derivatives. Derivative Flp-only responsive allele

Cre/Flp intersectional allele

Derivative Cre-only responsive allele

RC::Frepe BAP

mCherry

eGFP

Subtractive domain: mCherry; Intersectional domain: eGFP

BAP

eGFP

BAP

Background: None; Flp domain: eGFP

mCherry

eGFP

Background: mCherry; Cre domain: eGFP

RC::Fela eGFP

BAP

nlslacZ

Subtractive domain: eGFP; Intersectional domain: nlslacZ

nlslacZ

BAP

BAP

Background: None; Flp domain: nlslacZ

eGFP

nlslacZ

Ires-eGFP

PLAP

PLAP

BAP

BAP

Background: None; Flp domain: PLAP

Subtractive domain: eGFP; Intersectional domain: PLAP

Ires-eGFP

PLAP

BAP

eGFP

Subtractive domain: None; Intersectional domain: eGFP

eGFP

BAP

Background: None; Flp domain: eGFP

Cre domain: eGFP Yamamoto et al., (2009)

R26

nlslacZ

BAP

eGFP

Subtractive domain: nlslacZ; Intersectional domain: eGFP

nlslacZ

eGFP

Background: nlslacZ; Flp domain: eGFP

BAP

eGFP

Cre domain: eGFP

RC::PFwe BAP

nlslacZ

WGAIres-eGFP

Subtractive domain: nlslacZ; Intersectional domain: WGA, eGFP

BAP

nlslacZ

WGA Ires-eGFP

Background: nlslacZ; Flp domain: WGA, eGFP

BAP

WGA Ires-eGFP

mCherry

eGFP-TeTxLC

Subtractive domain: mCherry; Intersectional domain: eGFP-TeTxLC

BAP

mCherry

eGFP-TeTxLC

Background: mCherry; Flp domain: eGFP-TeTxLC

Cre/Flp modular vectors RC::PF base BAP

MCS

Farago et al., (2007)

Cre domain: WGA, eGFP

RC::PFTox BAP

Awatramani et al., (2003)

Miiyoshi et al., (2010)

eGFP

NGZ

BAP

Jensen et al., (2008)

Cre domain: PLAP

RCE BAP

Seri, B. and Dymecki, S., unpublished

Background: eGFP; Cre domain: nlslacZ

R26::Flap BAP

References

BAP

eGFP- TeTxLC

Kim et al., (2009)

Cre domain: eGFP-TeTxLC Ray, R., Kim, J.C. and Dymecki, S., unpublished

RC::FP base MCS

MCS

Subtractive domain activated by Cre; Intersectional domain activated by Flp

BAP

MCS

MCS

MCS

Subtractive domain activated by Flp; Intersectional domain activated by Cre

Intersectional Genetic Fate Mapping

193

expression, the woodchuck hepatitis virus posttranscriptional regulatory element (WPRE) has also been employed (Madisen et al., 2010). Thus, the target reporter transgene ideally employs regulatory elements with the broadest range and highest levels of activity; this contrasts with the Cre- and Flpe-encoding transgenes that strive for maximal cell-type restriction. In practice, the actual range of cell types that can be marked by a given target reporter transgene must be assessed empirically. An approximation of scope can be gathered upon analyzing tissue from an animal in which the target reporter transgene has been partnered with broadly active Cre and Flpe (or Flpo) transgenes simultaneously. Reporter transgene ‘‘activation’’ through removal of both STOP cassettes should occur broadly, resulting in most, if not all, cells in the mouse harboring the fully recombined reporter transgene. Cell types under study can then be assayed for robustness of reporter expression, thereby providing key information regarding reporter sensitivity and utility in the cell lineages of interest, and whether certain cell types fall outside the realm of fate mapping with a particular reporter allele because of limited coverage of reporter expression. This is a critical validation step, requisite before proceeding with actual fate mapping. A similar alternative would be to analyze reporter expression following germ line transmission of the fully recombined indicator transgene. Methods for reporter detection are presented in Section 8. With regard to target reporter transgenes, we prefer the designation ‘‘indicator’’ transgene or allele, and in this case an ‘‘intersectional indicator’’ transgene. This is to distinguish it from more conventional, constitutively driven (nonrecombinase dependent) promoter::reporter transgenes and to emphasize that the target transgene, through reporter expression, serves to ‘‘indicate’’ or provide a permanent record of all earlier occurring recombination events— reporter expression does not equate with present recombinase expression but rather recombinase activity at some point in the history of that cell. In other words, these genetic fate maps are cumulative because they include any cell that has ever, past or present, expressed both Cre and Flp. Of course single recombinase-based fate maps are also cumulative. This is not the case for most promoter::reporter transgenes; although persistence of a reporter molecule beyond the actual window of its transcription can sometimes be used effectively as a short-term lineage tracer.

4. Transgenes Enabling Subtractive as well as Intersectional Genetic Fate Mapping A typical intersectional indicator allele can be configured as cartooned in Figs. 11.1B and 11.2A, where the intersectional reporter (e.g., GFP)encoding sequence is separated from R26/CAG transcriptional regulatory and initiation elements by the presence of two STOP cassettes, the first loxP-flanked and the second FRT-flanked. Advantageously, a second cell

194

Susan M. Dymecki et al.

A

B PF BAP

P

Stop P

F

FP n bgal Stop F

GFP

BAP

F Stop F

geneA::Flpe

geneA::Flpe

geneB::cre

geneB::cre

geneA::Flpe, geneB::cre, intersectional indicator

GFP+ intersectional population b-gal+ subtractive population

Initial indicator marking

geneA::Flpe, geneB::cre, intersectional indicator

P

n bgal Stop P

GFP

GFP+ intersectional population b-gal+ subtractive population

Initial indicator marking

Figure 11.2 Illustration of intersectional and subtractive populations and the latter dependency on STOP-cassette order. In the ‘‘PF’’ configured allele, the loxP-flanked STOP cassette precedes the FRT-flanked cassette (left panel), while the reciprocal order characterizes the ‘‘FP’’ configuration (right panel). Shown are schematics of the neural tube (gray cylinder), with the expression domain for hypothetical gene A and Flpe recombinase (yellow) restricting along the dorsoventral (DV) axis but extending along the anteroposterior (AP) axis (top row); in contrast, the expression domain for gene B (pink) restricts along the AP axis but extends along the DV axis (middle row). When geneA::Flpe and geneB::cre are coupled with a PF dual-recombinase-responsive indicator allele (bottom row, left), cells expressing cre and Flpe activate production of GFP (green domain, intersectional population) while cells expressing only cre activate production of nb-gal (blue domain, subtractive population). When geneA::Flpe and geneB::cre are coupled with an FP configured allele (bottom row, right), cells expressing cre and Flpe still activate production of GFP in the same intersectional population (green domain) but now cells expressing only Flpe (rather than cre) activate production of nb-gal (blue domain, new subtractive population). (Reproduced from Dymecki and Kim, 2007 with permission from Elsevier Science.) (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

population lying outside the intersectional subset can be mapped simultaneously simply by incorporating a second reporter gene, such as one encoding mCherry within the second cassette upstream of its cognate STOP sequence (Fig. 11.1B(e–g)). This scenario allows not only for tracking the highly selective intersectional Cre/Flp population by GFP but also the remaining Cre/non-Flp cells by mCherry. We refer to the mCherry cells as the ‘‘subtractive’’ population (red cells cartooned in Fig. 11.1B or blue cells cartooned in Fig. 11.2) because they are what remain when Cre/ Flp-intersecting cells are subtracted from the Cre-only expressing population. For example, we have used allele RC::PFwe (Table 11.1) to map the

Intersectional Genetic Fate Mapping

195

fate of an entire rhombomere (composed of the intersectional GFPþ cells and subtractive b-galþ cells) while being able to distinguish the unique fates assumed by cells arising specifically from the dorsal-most region of that rhombomere (the intersectional GFPþ population) (Fig. 11.2A) (Farago et al., 2006). We refer to this ‘‘style’’ of intersectional transgene as being in the ‘‘PF configuration’’ because the loxP-flanked STOP cassette precedes the FRT-flanked cassette (Fig. 11.2A). We have generated additional intersectional alleles in which the subtractive population is the reverse (Table 11.1), the Flpe/non-Cre population (e.g., when employing intersectional alleles RC::Fela (Jensen et al., 2008) and RC::Frepe (R. Brust, S. Seri and S. M. Dymecki, unpublished data). This is achieved simply by swapping the order of the floxed and flrted STOP cassettes (Fig. 11.2B): the first (50 situated) cassette is now FRT-flanked and the second is loxP-flanked while also harboring the second reporter-encoding sequence. We refer to this style as the ‘‘FP configuration’’ and have used this strategy to fate map the entire serotonergic neuron population (e.g., Pet1::Flpe-mapped population) while being able to distinguish those subtypes arising from different rhombomeres (rhombomere-specific::Cre/Pet1::Flpe intersectional population) ( Jensen et al., 2008).

5. Exploiting Different Reporter Molecules to Reveal Different Features of Mapped Cell Populations Depending on the type of reporter molecule employed for genetic fate mapping, whether it is intersectional or single-recombinase based, different features of the mapped cell population may be uncovered in addition to their genetic history. Nuclear localized versions of reporters (e.g., nGFP or nb-gal) can allow visualization of individual cells in a highly sensitive way, often not achievable by cytoplasmic reporters when cells are tightly clustered. On the other hand, cytoplasmic or membrane-localized reporters (e.g., farnesylated or myristylated GFP or alkaline phosphatase) help resolve cell morphology, including axonal projections by virtue of either filling or outlining cells. Endogenously fluorescing protein reporters (e.g., GFP) can offer the further possibility of visualizing dynamic changes in live cell morphology and position and can permit electrophysiology in cultured brain slices or explants. As long as it is genetically encoded, virtually any newly developed reporter can be incorporated into intersectional fate mapping alleles; indeed, we have generated a modular base vector system designed for rapid insertion of such gene sequences (Table 11.1).

196

Susan M. Dymecki et al.

6. 3 for 1 In addition to providing improved cell-subtype selectivity, intersectional alleles offer a further practical advantage. Three different mouse lines can be generated from one initial transgene construction and strain generation: the intersectional indicator mouse line, but also two derivative single recombinase responsive lines (Table 11.1). The latter two are readily generated through germ line deletion of either the loxP- or FRT-flanked cassette from the original intersectional allele (Farago et al., 2006). A number of effective germ line deleter lines are available for derivative strain generation (Bai et al., 2002; Farley et al., 2000; Kanki et al., 2006; Kranz et al., 2010; Meyers et al., 1998; Rodriguez et al., 2000; Schwenk et al., 1995; Wu et al., 2009). Thus, dual-recombinase intersectional alleles offer high cell-subtype selectivity when needed, while also offering all the features of a single recombinase-based approach; they therefore can be considered as a standard approach in developing future indicator or effector alleles.

7. Intersectional Transgene Activation Reaches from Mapping Cell Fate to Mapping Cell Function Genetic fate maps are indeed advancing our understanding of mammalian development and disease, ushering in new molecular views of tissue architecture and anatomy and their underlying construction. Importantly, genetic fate maps are also templating an entirely new class of tools where various cell functions can be probed in addition to cell fate. This is achieved by modifying the intersectional indicator allele to now include a genetically encoded effector molecule, for example, a molecule capable of perturbing cell activity in a cell autonomous way such that cell function can be inferred through studying the behavioral consequences of this disruption. One representative and exciting class of effector molecules are neuromodulators, allowing one to control the activity of discrete neural circuits in vivo as a means to assess circuit roles in particular behaviors—an example we touch on more below. In practice, effector-encoding sequence can be inserted along with intersectional reporter-encoding sequence in the form of a fusion- or a bicistronic-gene so that the manipulated cells remain distinguishable by reporter expression (e.g., RC::PFtox (Kim et al., 2009) or RC::PFwe (Farago et al., 2006), respectively). Alternatively, the intersectional reporter sequence can be swapped for effector sequence; in this case,

Intersectional Genetic Fate Mapping

197

it is important to be able to visualize the manipulated cells through direct detection of the effector molecule itself, for example, through immunodetection of an effector molecule epitope or of an epitope tag engineered into the effector molecule (see Section 8). Similar to intersectional genetic fate mapping, the selection of cells for intersectional perturbation (through effector molecule expression) is by way of the overlap in expression between the employed Cre and Flpe (or Flpo) driver transgenes. Thus the resolution by which functions can be mapped to very specific cell subtypes remains high. As for any genetic fate mapping experiment, it is important in cell-function mapping to: (1) determine the full range of cells accessed by the Cre and Flpe drivers employed (methods to do this are presented below); (2) evaluate expression of the effector molecule or reporter as proxy (general immunodetection methods are presented below); and (3) assess whether the intersectional allele is able to drive effector molecule amounts necessary to mediate the desired perturbation—for example, if R26/CAG sequences are being employed, it needs to be determined if they are sufficiently active in the cell population of interest to achieve functional effector molecule levels. This will of course depend on the targeted cell type and the nature and potency of the effector molecule itself. An exciting group of effector molecules is that encompassing genetically encoded neuromodulators. Importantly, many are well suited to the above dual-recombinase-based approach for their in vivo delivery—thereby offering intersectional precision in cell-subtype selection for perturbation. Three general classes of genetically encoded neuromodulators are currently being developed and applied to manipulate neuron activity in the mouse: exocyto-, pharmaco- (also referred to as chemico-), and optogenetic— each have been reviewed recently (Airan et al., 2007; Conklin et al., 2008; Dymecki and Kim, 2007; Luo et al., 2008; Pei et al., 2008; Schneider et al., 2008; Zhang et al., 2007) and thus presented in limited detail here. Of note, the necessary levels of expression for each appear to be different (not surprising given their different modes of action), as are their offered features: modulation of chemical versus electrical neurotransmission, inducibility, reversibility, modulation kinetics, and so on. We have generated dual-recombinase-based intersectional alleles that exploit effectors from the first two classes. One such intersectional effector allele of the exocyto-genetic class, called RC::PFtox (Kim et al., 2009), has the potential to suppress vesicular neurotransmitter release (exocytosis) by conditional expression of the light chain from tetanus toxin (tox). Tox cleaves, in a cell autonomous fashion, VAMP2/synaptobrevin, a synaptic vesicle protein required for synaptic vesicle docking and evoked neurotransmitter release (Schiavo et al., 2000). Tox is quite potent (Schiavo et al., 2000) and has been used with great success in many model organisms, including fruit flies (Martin et al., 2002; Sweeney et al., 1995) and mice (Nakashiba et al.,

198

Susan M. Dymecki et al.

2008; Yamamoto et al., 2003). We have shown the RC::PFtox effector allele to drive functional levels of tox expression in a wide range of neuron types, and demonstrated its suitability for various behavioral (e.g., motor and cognitive) studies (Kim et al., 2009). Pharmaco (or chemico)-genetic approaches generally involve exogenous ligand–receptor combinations, where, ideally, the receptor-expressing neurons become uniquely and exclusively responsive to an exogenous drug (Alexander et al., 2009; Arenkiel et al., 2008; Armbruster et al., 2007; Conklin et al., 2008; Lechner et al., 2002; Lerchner et al., 2007; Nichols and Roth, 2009; Pei et al., 2008; Slimko et al., 2002; Tan et al., 2006; Tan et al., 2008; Wehr et al., 2009). Two commonly used effectors in this class are the vanilloid receptor (TRPV1) channel (Arenkiel et al., 2008) and G-protein coupled receptors (hM1D–hM4D) (Alexander et al., 2009; Armbruster et al., 2007; Conklin et al., 2008; Nichols and Roth, 2009; Pei et al., 2008) that have been engineered so that they no longer bind endogenous ligand and instead bind with high affinity exogenous ligand. TRPV1 is activated upon application of capsaicin and results in cation current and action potential firing. The modified G-protein coupled receptors are activated by clozapine-N-oxide (CNO), leading to a range of effects depending on the downstream signaling consequences—for example, hM4D induced neuronal silencing via Gi-coupled activation of inward rectifying potassium channels (GiRK channels) resulting in neuron hyperpolarization and inhibition of action potential firing (Armbruster et al., 2007). These approaches have the potential to be reversible as well as inducible—and thus have the potential to be very powerful in that repeated assays can be performed overtime on the same animal. We are generating intersectional effector alleles harboring various of these exogenous receptors. The third general approach for neuromodulation—the optogenetic approach—involves use of light-activated channels and pumps for neuron stimulation and inhibition (Airan et al., 2007; Han and Boyden, 2007; Han et al., 2009; Schneider et al., 2008; Zhang et al., 2006, 2007). They offer inducible and reversible activity modulation on millisecond time scales and have the potential to resolve specific connections between cells as opposed to all connections made by that kind of cell. Opticogenetic approaches require activation by light, as their name indicates. This property has both pros and cons—it can enable very precise regionalized channel activation, but can be a formidable obstacle depending on the location and degree of distribution of the neurons under study and the type of physiology and/or behavior needing to be assessed. Expression levels appear to be more demanding here and consequently the R26/ CAG approach may be suboptimal for various neuron types; other promoter/enhancer combinations, including viral strategies, have been met with excellent success.

Intersectional Genetic Fate Mapping

199

8. Methods and Materials 8.1. Comparing recombinase versus endogenous driver gene expression by in situ hybridization It is critical to establish the extent to which recombinase expression mirrors spatiotemporally the expression profile of the chosen driver gene or driver enhancer elements. This is because unexpected recombinase expression could lead to erroneous inclusion of cells into the fate map that are unrelated genetically. We therefore recommend performing in situ hybridizations on adjacent tissue sections prepared from a developmental series of staged transgenic embryos (e.g., embryos collected at 24–48 h intervals until birth) as well as a range of postnatal animals. Adjacent sections are prepared and probed for either recombinase or driver gene mRNA by in situ hybridization, followed by high-magnification imaging to compare expression domains. Further, we recommend comparing recombinase expression (mRNA and/or protein) to that of actual recombinase activity (assessed by reporter detection). Methods employed are detailed below, with the necessary materials and reagents listed at the end of each section. 8.1.1. Preparing tissue from prenatal mice for in situ hybridization 1. Dissect embryos in chilled diethyl pyrocarbonate–phosphate-buffered saline (DEPC–PBS) and remove extraembryonic membranes. Rinse with cold DEPC–PBS to remove any residual blood. 2. Replace DEPC–PBS with 4% paraformaldehyde (PFA) fix solution and gently rock at 4  C for 4–6 h. 3. Wash three times with PBS for 5 min each to remove residual fixative. 4. To preserve tissue architecture, soak tissues in 30% sucrose/PBS solution, rocking gently at 4  C overnight or until tissue sinks. 5. Carefully roll tissues in an aliquot of water-soluble embedding compound (e.g., Optimal Cutting Temperature (OCT) compound; Sakura Fineteck inc.) to remove residual solution before embedding. 6. Submerge and position tissue in an embedding mold-containing embedding media. Gently push out air bubbles with forceps, and immediately freeze by floating mold in a dry ice/ethanol bath until embedding media turn from clear to white. 7. Carefully remove mold from dry ice/ethanol bath and wipe off excess ethanol. Wrap mold tightly in aluminum foil and place in an airtight plastic bag. Store at 80  C until use. 8. Cryosection tissue and mount on charged slides (e.g., SuperfrostÒ Plus microscope slides). Air-dry slides for 1 h. Store sections in sealed box at 80  C.

200 

Susan M. Dymecki et al.

Materials/reagents

1. PBS: 137 mM NaCl, 2.6 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.4. Autoclave and store at room temperature. 2. DEPC–PBS: Add 1 ml of DEPC to 1 l of PBS, stir overnight in fume hood, autoclave, and store at room temperature. Note that DEPC is a suspected carcinogen and should be handled with care. 3. 4% PFA fix solution: Prepare fresh by diluting 16% PFA/PBS stock solution with PBS. 16% PFA stock solution: In a fume hood, dissolve 160 g PFA in 800 ml of distilled water stirring at 60  C. Add 1 N NaOH dropwise until solution clears. Add 100 ml of 10 PBS. Cool and adjust pH to 7.4. Add distilled water to 1 l and filter. Aliquot into 50 ml tubes and store at 20  C. 4. 30% sucrose/PBS solution: 30% sucrose (w/v) in PBS. Filter (0.2 mm), sterilize, and store at 4  C. 5. Embedding media: OCT (VWR, Cat. no. 100498-158) or TBS (Triangle Biomedical Sciences, Cat. no. H-TFM). 6. SuperfrostÒ Plus microscope slides (VWR scientific, Cat. no. 483II-703) 8.1.2. Preparing tissue from postnatal mice for in situ hybridization 1. Perfuse animal transcardially with ice-cold PBS solution followed by ice-cold 4% PFA fix solution. 2. Dissect tissue of interest and incubate in 4% PFA fix solution, rocking gently at 4  C overnight. 3. Rinse with ice-cold PBS at 4  C to remove residual fixative. 4. Follow above method steps 4–8. 8.1.3. Synthesis of digoxigenin-labeled riboprobes for in situ detection of mRNA—Endogenous or recombinase transcripts 1. To prepare DNA template (1 mg/ml) for in vitro transcription, cut plasmid DNA containing cDNA of interest (e.g., Cre or Flpe) downstream of T3, T7, or SP6 promoter by appropriate restriction enzyme. To minimize probe contamination with vector sequence, use a restriction enzyme that cuts adjacent to the cDNA sequences on the side distal to the promoter. As long as the restriction enzyme does not cut between promoter and cDNA sequences, the restriction site need not be unique. 2. Plasmid DNA should contain a promoter and cDNA sequences in an orientation that will generate antisense RNA during the in vitro transcription reaction. After restriction digestion, purify linearized DNA using preferred methods. 3. To generate approximately 10 mg of a given riboprobe, assemble the following reagents: 2 ml of 10 transcription buffer, 2 ml of 10NTP

Intersectional Genetic Fate Mapping

4. 5. 6. 7. 8.

9.



201

labeling mixture, 1 ml of linearized plasmid (1 mg/ml), 1 ml of RNase inhibitor (40 U/ml), 1 ml of T3, T7, or SP6 RNA polymerase (20 U/ml), add DEPC–water up to 20 ml. Incubate at 37  C for 2 h. To digest template DNA, add 1 ml of DNase I (RNase-free, 10 U/ml) and incubate for 25 min at 37  C. Add 2 ml of 0.5 M EDTA to stop the polymerase reaction. Add 100 ml of TE, 10 ml of 4 M LiCl, and 300 ml of ice-cold 100% ethanol. Mix well and incubate at 20  C for 2 h to overnight. Spin at 10,000g at 4  C for 15 min in a microcentrifuge. Discard supernatant while ensuring that the riboprobe pellet remains in the tube. Wash pellet with ice-cold 70% ethanol and air-dry for 5 min. Resuspend pellet in 100 ml of TE to achieve a final concentration of approximately 0.1 mg/ml. If the labeled probe is not used immediately, store the probe solution at 80  C. Avoid repeated freezing and thawing of the probe. To more precisely estimate riboprobe concentration, run 1 ml of riboprobe on a 1% agarose gel containing 0.5 mg/ml ethidium bromide beside an RNA standard of known concentration. Compare band intensities to estimate the amount of riboprobe synthesized. To avoid RNase contamination, all electrophoresis apparatus should be cleaned thoroughly with DEPC-treated water and treated with reagents containing RNase inhibitor. We found the commercially available reagent such as RNaseZap (Ambion, Cat. no. 9780-9784) works efficiently. Materials/reagents

1. DEPC–water: Add 1 ml of DEPC to 1 l of distilled water, stir overnight in the fume hood, autoclave, and store at room temperature. 2. 10 transcription buffer (Roche, Cat. no. 1-465-384): 400 mM Tris–HCl, pH 8.0, 60 mM MgCl2, 100 mM dithioerythritol, 20 mM spermidine, 100 mM NaCl. 3. 10 NTP labeling mixture (Roche, Cat. no. 1-277-073): 10 mM ATP, 10 mM CTP, 10 mM GTP, 6.5 mM UTP, and 3.5 mM digoxigeninUTP in Tris–HCl, pH 7.5. 4. RNase inhibitor (Roche, Cat. no. 799-017). 5. T3, T7, or SP6 RNA polymerase (Roche, Cat. no. 1-031-163, 881-767, or 810-274, respectively). 6. DNase I (RNase-free) (Roche, Cat. no. 776-785). 7. 0.5 M EDTA: Add 186.1 g of EDTA (disodium, dihydrate) to 800 ml of distilled water and adjust pH to 8.0 by adding NaOH pellets (about 20 g). Autoclave and store at room temperature. 8. TE: 50 mM Tris–HCl, 1 mM EDTA, pH 8.0. Autoclave and store at room temperature. 9. Digoxigenin-labeled RNA standards (Roche, Cat. no. 1-585-746).

202

Susan M. Dymecki et al.

8.1.4. RNA detection on tissue sections by in situ hybridization 1. Remove tissue slides from  80  C and allow them to come to room temperature. 2. Transfer slides into DEPC–PBS. Incubate 5 min at room temperature. 3. Transfer slides into 1 mg/ml proteinase K in DEPC–PBS (prepared from freshly thawed 10 mg/ml proteinase K stock). Incubate 5 min at room temperature. Optimal concentration of proteinase K and incubation time should be empirically decided depending on thickness of tissue. 4. To quench proteinase K activity, transfer slides into glycine/DEPC– PBS solution. Incubate 5 min at room temperature. 5. Transfer slides into fresh DEPC–PBS. Incubate 5 min at room temperature. 6. Transfer slides into 0.25% acetic anhydride in 0.1 M triethanolamine. The acetic anhydride should be added to the triethanolamine solution just before use. Incubate 10 min at room temperature. 7. Transfer slides into fresh DEPC–PBS. Incubate 5 min at room temperature. 8. Prehybridization: Transfer slides into prewarmed (65  C) prehybridization solution. Incubate 15 min at 65  C. Prehybridization solution can be made ahead of time and stored at 20  C. 9. Hybridization: Place aliquot of probe (0.1 mg per slide) in microcentrifuge tube and put in 85  C heat block for 3 min. Transfer to ice immediately. Add hybridization buffer (100 ml per slide) directly to the tube containing riboprobe and keep it at 65  C until needed. Apply 100 ml of hybridization buffer containing 0.5–1 mg/ml digoxigeninlabeled riboprobe to each slide and cover with siliconized coverslip. Hybridize in sealed humid chamber. This can be achieved by placing Kim Wipes soaked in 5 SSC at the bottom of chamber. Incubate at 65  C overnight. If working with multiple probes, change gloves between each probe to avoid contamination. Avoid placing multiple slides with different probes in the same box. 10. Transfer slides into prewarmed (65  C) 5 SSC. Incubate 5 min at room temperature and dip until coverslips fall off. Do not pull off coverslips since damage to the tissue section can be incurred. 11. Transfer slides into prewarmed (65  C) wash buffer I. Incubate 15 min at 65  C. Repeat twice using fresh prewarmed (65  C) wash buffer I. 12. Transfer slides into fresh prewarmed (65  C) wash buffer I:wash buffer II (1:1 ratio). Incubate 10 min at 65  C. 13. Transfer slides into prewarmed (65  C) wash buffer II. Incubate 5 min at 65  C. Repeat twice using fresh prewarmed (65  C) wash buffer II. 14. Transfer slides into RNase A containing wash buffer II (RNase A 25 mg/ml). Incubate for 30 min at 37  C.

Intersectional Genetic Fate Mapping

203

15. Transfer slides into prewarmed (65  C) wash buffer II. Incubate 5 min at 65  C. 16. Transfer slides into TBS-T. Incubate 10 min at room temperature. Repeat twice using fresh TBS-T. 17. Apply 0.5 ml of TBS-T containing 10% heat-inactivated lamb serum to each slide. Incubate in humid chamber for 30 min at room temperature. 18. Apply 0.5 ml of antibody (alkaline phosphatase-conjugated antidigoxigenin antibody) diluted at 1:5000 in TBS-T to each slide. Incubate in humid chamber for 2 h at room temperature or overnight at 4  C. 19. Transfer slides into TBS-T. Incubate 5 min at room temperature. 20. Transfer slides into fresh TBS-T. Incubate 15 min at room temperature. Repeat four times using fresh TBS-T. 21. Transfer slides into NTMT. Incubate 5 min at room temperature. Repeat twice using fresh NTMT. 22. Apply 1 M of freshly prepared alkaline phosphatase staining solution to each slide. Incubate in humid chamber in the dark at room temperature until desired color development is achieved (a few hours to a few days depending on the abundance of the mRNA). Change alkaline phosphatase staining solution every day. 23. Stop reaction by transferring slides into 20 mM EDTA in PBS. 24. Rinse slides in PBS for 10 min at room temperature. 25. Remove excess PBS, apply mounting media, and place a glass coverslip carefully. 26. Air-dry slides until mounting medium has solidified. Slides should be stored in the dark to prevent additional color development. 

Materials/reagents 1. DEPC–PBS: Add 1 ml of DEPC to 1 l of PBS, stir overnight in fume hood, autoclave, and store at room temperature. 2. Proteinase K stock solution (10 mg/ml): Add 0.1 g of proteinase K to 10 ml of TE. Store in aliquots at 20  C. 3. Glycine/DEPC–PBS solution: 2 mg/ml glycine in DEPC–PBS. Prepare just before use. 4. 0.25% acetic anhydride in 0.1 M triethanolamine: Add 625 ml of acetic anhydride in 250 ml of 0.1 M triethanolamine. 5. Prehybridization/hybridization solution: 50% formamide, 5 SSC, 50 mg/ml yeast tRNA, 1% SDS, and 50 mg/ml heparin in DEPC–water. Store at 20  C. 6. Siliconized coverslip: Treat coverslips as follows. Dip 15 times in 3% silicon in chloroform (v/v). Repeat with fresh 3% silicon in chloroform (v/v). Dip 15 times in 100% ethanol. Repeat twice with fresh 100% ethanol. Air-dry under hood.

204

Susan M. Dymecki et al.

7. 5 SSC solution: Dilute from 20 stock solution with DEPC-treated water. 20 SSC stock solution: 3 M NaCl, 0.3 M sodium citrate in DEPC–water. Adjust to pH 4.5 with 0.6 M citric acid. 8. Wash buffer I: 50% formamide, 5 SSC, 1% SDS in DEPC–water. 9. Wash buffer II: 0.5 M NaCl, 10 mM Tris–HCl, pH 7.5, 0.1% polyoxyethylene–sorbitan monolaurate (Tween-20) in DEPC–water. 10. Wash buffer III: 50% formamide, 2 SSC in DEPC–water. 11. 1 TBS-T: Dilute from 10 stock solution with distilled water. 10 TBS-T: 1.4 M NaCl, 27 mM KCl, 250 mM Tris–HCl, pH 7.5, and 10% Tween-20. Store at room temperature. 12. Heat-inactivated sheep serum (Gibco, Cat. no. 16070-096). 13. Antidigoxigenin antibody conjugated to alkaline phosphatase (Roche, Cat. no. 1-093-274). 14. NTMT: 100 mM NaCl, 100 mM Tris–HCl, pH 9.5, 50 mM MgCl2, 1% Tween-20. 15. 4-Nitro blue tetrazolium chloride (NBT) (Roche, Cat. no. 1383-213). 16. X-phoshate/5-bromo-4-chloro-3-indolyl-phosphate (BCIP) (Roche, Cat. no. 1-383-221). 17. Alkaline phosphatase staining solution: Add 4.5 ml of NBT and 3.5 ml of BCIP to 1 ml of NTMT. Prepare immediately before use. 18. Mounting medium: Aqua-Poly/Mount (Polyscience, Cat. no. 18606).

8.2. Comparing the spatiotemporal profile of recombinase expression to that of actual recombinase activity As indicated above, it is important to establish the extent to which indicator transgene recombination (actual reporter activation) matches the initial recombinase driver gene expression profile (mRNA and/or protein)— later they will diverge because the reporter expression is cumulatively and permanently tracking all cells that ever in their history expressed the driver gene, whereas the driver gene expression is transient in a set of embryonic cells, for example. In other words, it is important to determine whether all or only a part of an initial gene expression domain is being fate mapped. This becomes of particular concern if the driver gene exhibits a gradient in its expression; it is then possible that the lowest expressers in the gradient may be missed because adequate levels of recombinase are not achieved (reviewed in Dymecki and Kim, 2007). Thus, in double transgenic (recombinase; indicator) mice, it is important to check whether cells positive for recombinase mRNA are also positive for recombinase activity as reflected in reporter molecule expression. Sets of serial sections are obtained and analyzed for recombinase mRNA on one set and reporter molecule on the other (e.g., X-gal detection for b-gal activity).

Intersectional Genetic Fate Mapping

205

1. Prepare embryonic tissue as described above. 2. Cryosection and collect adjacent tissue sections into serial sets. 3. Perform in situ hybridization on one set of sections as described above using riboprobe that recognizes recombinase mRNA and perform X-gal staining (detailed below) or reporter immunodetection (detailed below) with the other set of adjacent sections. It is also suitable to coimmunodetect recombinase and reporter protein (see below). Antibodies to detect Cre are commercially available (Covance, Cat no. MMS-106p) (Tsien et al., 1996); antibodies to detect Flp protein in tissue section remain lacking.

8.3. Identifying progenitor cells and their descendants by enzymatic or immunofluorescence detection of the lineage tracer molecules b-gal or GFP For intersectional genetic fate mapping, both a Cre- and a Flpe-encoding transgene are combined with a dual-recombinase responsive indicator allele of choice (PF or FP variations, see Fig. 11.2), thus generating triple transgenic animals (harboring cre, Flpe, and intersectional indicator alleles). Breeding strategies to combine these alleles vary depending on reproductive rate of each strain and on any recombinase activity in the germ line. A typical breeding scheme, for example, might involve first generating cre; Flpe double transgenics, which are then crossed to indicator homozygotes. Triple transgenic progeny are identified by PCR genotyping, followed by tissue harvest, preparation, and analyses to locate marked descendant cells by fluorescence or enzymatic detection of indicator molecules. 8.3.1. Tissue preparation from prenatal mice for X-gal detection of b-gal 1. Dissect embryos in ice-cold PBS, removing extraembryonic membranes. 2. Wash with cold PBS to remove any residual blood. 3. Replace PBS with 2% PFA fix solution and gently rock at 4  C for 4–6 h. Overfixation can affect b-gal activity. Optimal fixation time may be empirically determined. 4. Wash three times with PBS for 5 min each to remove residual fixative. 5. To preserve tissue architecture, soak tissues in 30% sucrose/PBS solution, rocking gently at 4  C overnight or until tissue sinks. 6. Carefully roll tissues in an aliquot of embedding media (e.g., OCT) to remove residual solution before embedding. 7. Submerge and position tissue in an embedding mold-containing embedding media. Gently push out air bubbles with forceps, and immediately

206

Susan M. Dymecki et al.

freeze by floating mold in a dry ice/ethanol bath until embedding media turn from clear to white. 8. Carefully remove mold from dry ice/ethanol bath and wipe off excess ethanol. Wrap mold tightly in aluminum foil and place in an airtight plastic bag. Store at 80  C until use. 9. Cryosection tissue and mount on charged slides (e.g., SuperfrostÒ Plus microscope slides). Air-dry slides for 1 h. Store sections in sealed box at 80  C. 

Materials/reagents

1. PBS: 137 mM NaCl, 2.6 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.4. Autoclave and store at room temperature. 2. 4% PFA fix solution: Prepare fresh by diluting 16% PFA/PBS stock solution with PBS. 16% PFA stock solution: In a fume hood, dissolve 160 g PFA in 800 ml of distilled water stirring at 60  C. Add 1 N NaOH dropwise until solution clears. Add 100 ml of 10 PBS. Cool and pH to 7.4. Add distilled water to 1 l and filter. Aliquot into 50 ml tubes and store at  20  C. 3. 30% sucrose/PBS solution: 30% sucrose (w/v) in PBS. Filter-sterilize (0.2 mm) and store at 4  C. 4. Embedding media: OCT (VWR, Cat. no. 100498-158), TBS (Triangle Biomedical Sciences, Cat. no. H-TFM). 5. SuperfrostÒ Plus microscope slides (VWR scientific, Cat. no. 483II-703) 6. Mounting medium: Aqua-Poly/Mount (Polyscience, Cat. no. 18606) 8.3.2. Tissue preparation from postnatal mice for X-gal detection of b-gal 1. Perfuse animal transcardially with ice-cold PBS followed by ice-cold 4% PFA fix solution. 2. Dissect tissue of interest and incubate in 4% PFA fix solution, rocking gently at 4  C overnight. Overfixation can affect b-galactosidase activity. Optimal fixation time may be empirically determined. 3. Rinse with ice-cold PBS at 4  C to remove residual fixative. 4. Follow steps 5–8 above. 8.3.3. Detecting b-gal activity on tissue section by X-gal histochemistry 1. Collect and process tissue as described above. 2. Remove the tissue slides from  80  C and allow them to come to room temperature.

Intersectional Genetic Fate Mapping

207

3. Transfer slides into PBS. Incubate 5 min at room temperature. Repeat with fresh PBS. 4. Transfer slides into X-gal staining solution and incubate at 37  C in the dark until color develops to the desired extent (2–48 h). X-gal staining solution can be reused several times if stored at 4  C and protected from light. 5. Transfer slides into PBS. Incubate 5 min at room temperature. Repeat with fresh PBS. 6. Remove excess PBS, apply mounting media, and place a glass coverslip carefully. 7. Air-dry slides until mounting medium is solidified. 

Materials/reagents

1. PBS: 137 mM NaCl, 2.6 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.4. Autoclave and store at room temperature. 2. X-gal staining solution: 5 mM K3[Fe(CN)6], 5 mM K4[Fe(CN)6], 1 mg/ml X-gal, 0.01% sodium deoxycholate, 0.02% Nonidet P-40 (NP-40), 2 mM MgCl2 in PBS. 3. Mounting medium: Aqua-Poly/Mount (Polyscience, Cat. no. 18606).

8.3.4. Detecting lineage tracer molecule (e.g., b-gal or GFP) by immunofluorescence 1. Collect and process tissue as described above. 2. Remove the tissue slides from 80  C and allow them to come to room temperature. 3. Transfer slides into PBS. Incubate 5 min at room temperature. 4. Transfer slides into PBS-T. Incubate 5 min at room temperature. Note that detection of membrane-bound antigens may be affected by Triton X-100 treatment. 5. Blocking: Apply 5% normal serum (use serum of species that secondary antibody is raised in) in PBS-T and incubate in humid chamber at room temperature for 1 h. 6. Transfer slides into PBS-T. Incubate 5 min at room temperature. 7. Apply primary antibody (e.g., anti-GFP or b-gal antibody) diluted in PBS-T to each slide and incubate in humid chamber at 4  C overnight. We found it is very important to determine the optimal titer of primary antibody by titration. For example, if the manufacturer suggests a dilution of 1:500, it is useful to test a range such as 1:250, 1:500, 1:1000, 1:5000, and 1:20,000. 8. Transfer slides into PBS-T. Incubate 5 min at room temperature. Repeat twice with fresh PBS-T.

208

Susan M. Dymecki et al.

9. Apply fluorophore-conjugated secondary antibody diluted in PBS-T to each slide and incubate in humid chamber at room temperature for 1 h. 10. Transfer slides into PBS. Incubate 5 min at room temperature. Repeat twice with fresh PBS. 11. If necessary, stain with 40 ,6-diamidino-2-phenyindole (DAPI) by incubating slides in DAPI staining solution for 5 min at room temperature. DAPI staining solution can be stored as 10 stock at 4  C. 12. Transfer slides into PBS. Incubate 5 min at room temperature. 13. Remove excess PBS, apply mounting media, and place a glass coverslip carefully. 14. Air-dry slides until mounting medium is solidified. Store slides at 4  C in the dark. 

Materials/reagents

1. PBS: 137 mM NaCl, 2.6 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.4. Autoclave and store at room temperature. 2. PBS-T: 0.05% Triton X-100 in PBS. 3. DAPI staining solution: 300 nM DAPI in PBS. DAPI staining solution can be stored as 10 stock at 4  C. 4. Mounting medium: Aqua-Poly/Mount (Polyscience, Cat. no. 18606).

8.4. Codetection of subtractive and intersectional lineage tracer molecules Two different antigens (e.g., GFP and mCherry) can be detected simultaneously by adding two different primary antibodies in the same solution. However, we found that in many cases, more sensitive detection is achievable by a sequential detection approach: 1. Detect the first lineage tracer molecule (either subtractive or intersectional) as described above. 2. Continue with detecting the second lineage tracer molecule by next following the above method steps 5–14.

9. Concluding Remarks Our goal for this chapter has been to introduce how the site-specific recombinases, Cre and Flpe (or Flpo) can be used together along with a responsive target transgene to effectively ‘‘deliver’’ virtually any genetically encoded molecule to highly selective cell subsets in the mouse—and to

Intersectional Genetic Fate Mapping

209

achieve this noninvasively, reproducibly, and with versatility. Emphasis has been placed on exploiting intersectional genetics to improve the resolution by which cells are selected for tracking or manipulation, so as to enable drawing tighter correspondences, for example, between a type of progenitor cell, the genes it expresses, and the kinds of daughter cells it produces, or between a type of neuron and the behaviors it underlies. The better resolved the population, the more useful is the generated map, whether concerning progenitor–progeny cell relationships or neural circuitry underlying behavior. There are of course many other ways to achieve combinatorial selectivity and we encourage the reader to fully explore all possibilities.

ACKNOWLEDGMENTS This work was supported by grants from the National Institutes of Health and the Harvard Neurodiscovery Center. J. C. K. has been a fellow of the Foundation for Fighting Blindness of Canada. We thank members of the Dymecki lab for critical reading of the manuscript and valuable suggestions.

REFERENCES Airan, R., Hu, E., Vijaykumar, R., Roy, M., Meltzer, L., and Deisseroth, K. (2007). Integration of light-controlled neuronal firing and fast circuit imaging. Curr. Opin. Neurobiol. 17, 587–592. Akagi, K., Sandig, V., Vooijs, M., Van der Valk, M., Giovannini, M., Strauss, M., and Berns, A. (1997). Cre-mediated somatic site-specific recombination in mice. Nucleic Acids Res. 25, 1766–1773. Alexander, G. M., Rogan, S. C., Abbas, A. I., Armbruster, B. N., Pei, Y., Allen, J. A., Nonneman, R. J., Hartmann, J., Moy, S. S., Nicolelis, M. A., McNamara, J. O., and Roth, B. L. (2009). Remote control of neuronal activity in transgenic mice expressing evolved G protein-coupled receptors. Neuron 63, 27–39. Arenkiel, B. R., Klein, M. E., Davison, I. G., Katz, L. C., and Ehlers, M. D. (2008). Genetic control of neuronal activity in mice conditionally expressing TRPV1. Nat. Methods 5, 299–302. Armbruster, B. N., Li, X., Pausch, M. H., Herlitze, S., and Roth, B. L. (2007). Evolving the lock to fit the key to create a family of G protein-coupled receptors potently activated by an inert ligand. Proc. Natl. Acad. Sci. USA 104, 5163–5168. Awatramani, R., Soriano, P., Mai, J. J., and Dymecki, S. M. (2001). A FLP indicator strain, R26::FRAP, driving alkaline phosphatase expression from the ROSA26 locus. Nat. Genet. 29, 257–259. Awatramani, R., Soriano, P., Rodriguez, C., Mai, J. J., and Dymecki, S. M. (2003). Cryptic boundaries in roof plate and choroid plexus identified by intersectional gene activation. Nat. Genet. 35, 70–75. Bai, C. B., Auerbach, W., Lee, J. S., Stephen, D., and Joyner, A. L. (2002). Gli2, but not Gli1, is required for initial Shh signaling and ectopic activation of the Shh pathway. Development 129, 4753–4761.

210

Susan M. Dymecki et al.

Beard, C., Hochedlinger, K., Plath, K., Wutz, A., and Jaenisch, R. (2006). Efficient method to generate single-copy transgenic mice by site-specific integration in embryonic stem cells. Genesis 44, 23–28. Branda, C., and Dymecki, S. M. (2004). Talking about a revolution: The impact of sitespecific recombinases on genetic analyses in mice. Dev. Cell 6, 7–28. Buchholz, F., Ringrose, L., Angrand, P.-O., Rossi, F., and Stewart, A. F. (1996). Different thermostabilities of FLP and Cre recombinases: Implications for applied site-specific recombination. Nucleic Acids Res. 24, 4256–4262. Buchholz, F., Angrand, P. O., and Stewart, A. F. (1998). Improved properties of FLP recombinase evolved by cycling mutagenesis. Nat. Biotechnol. 16, 657–662. Cepko, C. L., Austin, C. P., Walsh, C., Ryder, E. F., Halliday, A., and Fields-Berry, S. (1990). Studies of cortical development using retrovirus vectors. Cold Spring Harb. Symp. Quant. Biol. 55, 265–278. Conklin, B. R., Hsiao, E. C., Claeysen, S., Dumuis, A., Srinivasan, S., Forsayeth, J. R., Guettier, J. M., Chang, W. C., Pei, Y., McCarthy, K. D., Nissenson, R. A., Wess, J., et al. (2008). Engineering GPCR signaling pathways with RASSLs. Nat. Methods 5, 673–678. Dymecki, S. M. (2000). Site-specific recombination in cells and mice. In ‘‘Gene Targeting: A Practical Approach,’’ (A. L. Joyner, ed.) pp. 37–99. Oxford University Press, Oxford. Dymecki, S. M., and Kim, J. C. (2007). Molecular neuroanatomy’s ‘‘Three Gs’’: A primer. Neuron 54, 17–34. Dymecki, S. M., Rodriguez, C. I., and Awatramani, R. B. (2002). Switching on lineage tracers using site-specific recombination. In ‘‘Embryoni Stem Cells: Methods and Protocols,’’ (K. Turksen, ed.). Humana Press, New Jersey. Farago, A. F., Awatramani, R. B., and Dymecki, S. M. (2006). Assembly of the brainstem cochlear nuclear complex is revealed by intersectional and subtractive genetic fate maps. Neuron 50, 205–218. Farley, F. W., Soriano, P., Steffen, L. S., and Dymecki, S. M. (2000). Widespread recombinase expression using FLPeR (flipper) mice. Genesis 28, 106–110. Galileo, D. S., Gray, G. E., Owens, G. C., Majors, J., and Sanes, J. R. (1990). Neurons and glia arise from a common progenitor in chicken optic tectum: Demonstration with two retroviruses and cell type-specific antibodies. Proc. Natl. Acad. Sci. USA 87, 458–462. Gong, S., Doughty, M., Harbaugh, C. R., Cummins, A., Hatten, M. E., Heintz, N., and Gerfen, C. R. (2007). Targeting Cre recombinase to specific neuron populations with bacterial artifical chromosome constructs. J. Neurosci. 27, 9817–9823. Han, X., and Boyden, E. S. (2007). Multiple-color optical activation, silencing, and desynchronization of neural activity, with single-spike temporal resolution. PLoS One 2, e299. Han, X., Qian, X., Bernstein, J. G., Zhou, H. H., Franzesi, G. T., Stern, P., Bronson, R. T., Graybiel, A. M., Desimone, R., and Boyden, E. S. (2009). Millisecond-timescale optical control of neural dynamics in the nonhuman primate brain. Neuron 62, 191–198. Heintz, N. (2001). BAC to the future: The use of bac transgenic mice for neuroscience research. Nat. Rev. Neurosci. 2, 861–870. Jensen, P., Farago, A. F., Awatramani, R. B., Scott, M. M., Deneris, E. S., and Dymecki, S. M. (2008). Redefining the serotonergic system by genetic lineage. Nat. Neurosci. 11, 417–419. Johansson, T., Broll, I., Frenz, T., Hemmers, S., Becher, B., Zeilhofer, H. U., and Buch, T. (2010). Building a zoo of mice for genetic analyses: A comprehensive protocol for the rapid generation of BAC transgenic mice. Genesis. 48, 264–280. Joyner, A., and Zervas, M. (2006). Genetic inducible fate mapping in mouse: Establishing genetic lineages and defining genetic neuroanatomy in the nervous system. Dev. Dyn. 235, 2376–2385.

Intersectional Genetic Fate Mapping

211

Kanki, H., Suzuki, H., and Itohara, S. (2006). High-efficiency CAG-FLPe deleter mice in C57BL/6J background. Exp. Anim. 55, 137–141. Keller, R. E. (1975). Vital dye mapping of the gastrula and neuraula of Xenopus laevis. I. Prospective areas and morphogenetic movements of the superficial layer. Dev. Biol. 42, 222–241. Kim, J. C., Cook, M. N., Carey, M. R., Shen, C., Regehr, W. G., and Dymecki, S. M. (2009). Linking genetically defined neurons to behavior through a broadly applicable silencing allele. Neuron 63, 305–315. Kramer, I., Sigrist, M., de Nooij, J. C., Taniuchi, I., Jessell, T. M., and Arber, S. (2006). A role for Runx transcription factor signaling in dorsal root ganglion sensory neuron diversification. Neuron 49, 379–393. Kranz, A., Fu, J., Duerschke, K., Weidlich, S., Naumann, R., Stewart, A. F., and Anastassiadis, K. (2010). An Improved Flp Deleter Mouse in C57Bl/6 Based on Flpo Recombinase. Genesis. May 4. [Epub ahead of print]. Landsberg, R. L., Awatramani, R. B., Hunter, N. L., Farago, A. F., DiPietrantonio, H. J., Rodriguez, C. I., and Dymecki, S. M. (2005). Hindbrain rhombic lip is comprised of discrete progenitor cell populations allocated by Pax6. Neuron 48, 933–947. Lechner, H. A., Lein, E. S., and Callaway, E. M. (2002). A genetic method for selective and quickly reversible silencing of Mammalian neurons. J. Neurosci. 22, 5287–5290. Le Douarin, N. (1982). The Neural Crest. Cambridge University Press, Cambridge. Lee, E. C., Yu, D., Martinez de Velasco, J., Tessarollo, L., Swing, D. A., Court, D. L., Jenkins, N. A., and Copeland, N. G. (2001). A highly efficient Escherichia coli-based chromosome engineering system adapted for recombinogenic targeting and subcloning of BAC DNA. Genomics 73, 56–65. Lerchner, W., Xiao, C., Nashmi, R., Slimko, E. M., van Trigt, L., Lester, H. A., and Anderson, D. J. (2007). Reversible silencing of neuronal excitability in behaving mice by a genetically targeted, ivermectin-gated Cl channel. Neuron 54, 35–49. Lewandoski, M. (2001). Conditional control of gene expression in the mouse. Nat. Rev. Genet. 2, 743–755. Lobe, C. G., Koop, K. E., Kreppner, W., Lomeli, H., Gertsenstein, M., and Nagy, A. (1999). Z/AP, a double reporter for Cre-mediated recombination. Dev. Biol. 208, 281–292. Luo, L., Callaway, E. M., and Svoboda, K. (2008). Genetic dissection of neural circuits. Neuron 57, 634–660. Madisen, L., Zwingman, T. A., Sunkin, S. M., Oh, S. W., Zariwala, H. A., Gu, H., Ng, L. L., Palmiter, R. D., Hawrylycz, M. J., Jones, A. R., Lein, E. S., and Zeng, H. (2010). A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat. Neurosci. 13, 133–140. Martin, J. R., Keller, A., and Sweeney, S. T. (2002). Targeted expression of tetanus toxin: A new tool to study the neurobiology of behavior. Adv. Genet. 47, 1–47. Meyers, E. N., Lewandoski, M., and Martin, G. R. (1998). An Fgf8 mutant allelic series generated by Cre- and Flp-mediated recombination. Nat. Genet. 18, 136–141. Miyoshi, G., Hjerling-Leffler, J., Karayannis, T., Sousa, V. H., Butt, S. J., Battiste, J., Johnson, J. E., Machold, R. P., and Fishell, G. (2010). Genetic fate mapping reveals that the caudal ganglionic eminence produces a large and diverse population of superficial cortical interneurons. J. Neurosci. 30, 1582–1594. Muyrers, J. P., Zhang, Y., and Stewart, A. F. (2001). Techniques: Recombinogenic engineering—New options for cloning and manipulating DNA. Trends Biochem. Sci. 26, 325–331. Nagy, A. (2000). Cre recombinase: The universal reagent for genome tailoring. Genesis 26, 99–109.

212

Susan M. Dymecki et al.

Nagy, A., Gertsenstein, M., Vintersten, K., and Behringer, R. (2003). Manipulating the Mouse Embryo: A Laboratory Manual. Cold Spring Harbor Laboratory Press, New York. Nakashiba, T., Young, J. Z., McHugh, T. J., Buhl, D. L., and Tonegawa, S. (2008). Transgenic inhibition on synaptic transmission reveals role of CA3 output in hippocampal learning. Science 319, 1260–1264. Nichols, C. D., and Roth, B. L. (2009). Engineered G-protein coupled receptors are powerful tools to investigate biological processes and behaviors. Front. Mol. Neurosci. 2, 16. Niwa, H., Yamamura, K., and Miyazaki, J. (1991). Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene 108, 193–200. Novak, A., Guo, C., Yang, W., Nagy, A., and Lobe, C. G. (2000). Z/EG, a double reporter mouse line that expresses enhanced green fluorescent protein upon Cre-mediated excision. Genesis 28, 147–155. Pei, Y., Rogan, S. C., Yan, F., and Roth, B. L. (2008). Engineered GPCRs as tools to modulate signal transduction. Physiology (Bethesda) 23, 313–321. Ray, R. S., and Dymecki, S. M. (2009). Rautenlippe Redux—Toward a unified view of the precerebellar rhombic lip. Curr. Opin. Cell Biol. 21, 741–747. Raymond, C. S., and Soriano, P. (2007). High-efficiency FLP and PhiC31 site-specific recombination in mammalian cells. PLoS One 2, e162. Rodriguez, C. I., and Dymecki, S. M. (2000). Origin of the precerebellar system. Neuron 27, 475–486. Rodriguez, C. I., Buchholz, F., Galloway, J., Sequerra, R., Kasper, J., Ayala, R., Stewart, A. F., and Dymecki, S. M. (2000). High-efficiency deleter mice show that FLPe is an alternative to Cre-loxP. Nat. Genet. 25, 139–140. Sauer, B. (1993). Manipulation of transgenes by site-specific recombination: Use of Cre recombinase. (P. M. Wassarman and M. L. DePamphilis, eds.), Vol. 225, pp. 890–900. Academic Press, Inc., San Diego. Schaft, J., Ashery-Padan, R., van der Hoeven, F., Gruss, P., and Stewart, A. F. (2001). Efficient FLP recombination in mouse ES cells and oocytes. Genesis 31, 6–10. Schiavo, G., Matteoli, M., and Montecucco, C. (2000). Neurotoxins affecting neuroexocytosis. Physiol. Rev. 80, 717–766. Schneider, M. B., Gradinaru, V., Zhang, F., and Deisseroth, K. (2008). Controlling neuronal activity. Am. J. Psychiatry 165, 562. Schwenk, F., Baron, U., and Rajewsky, K. (1995). A cre-transgenic mouse strain for the ubiquitous deletion of loxP-flanked gene segments including deletion in germ cells. Nucleic Acids Res. 23, 5080–5081. Slimko, E. M., McKinney, S., Anderson, D. J., Davidson, N., and Lester, H. A. (2002). Selective electrical silencing of mammalian neurons in vitro by the use of invertebrate ligand-gated chloride channels. J. Neurosci. 22, 7373–7379. Soriano, P. (1999). Generalized lacZ expression with ROSA26 Cre reporter strain. Nat. Genet. 21, 70–71. Srinivas, S., Watanabe, T., Lin, C.-S., William, C. M., Tanabe, Y., Jessell, T. M., and Costantini, F. (2001). Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev. Biol. 1, 4. Stark, W. M., Boocock, M. R., and Sherratt, D. J. (1992). Catalysis by site-specific recombinases. Trends Genet. 8, 432–439. Sweeney, S. T., Broadie, K., Keane, J., Niemann, H., and O’Kane, C. J. (1995). Targeted expression of tetanus toxin light chain in Drosophila specifically eliminates synaptic transmission and causes behavioral defects. Neuron 14, 341–351. Tan, E. M., Yamaguchi, Y., Horwitz, G. D., Gosgnach, S., Lein, E. S., Goulding, M., Albright, T. D., and Callaway, E. M. (2006). Selective and quickly reversible inactivation

Intersectional Genetic Fate Mapping

213

of mammalian neurons in vivo using the Drosophila allatostatin receptor. Neuron 51, 157– 170. Tan, W., Janczewski, W. A., Yang, P., Shao, X. M., Callaway, E. M., and Feldman, J. L. (2008). Silencing preBotzinger complex somatostatin-expressing neurons induces persistent apnea in awake rat. Nat. Neurosci. 11, 538–540. Tsien, J. Z., Chen, D. F., Gerber, D., Tom, C., Mercer, E. H., Anderson, D. J., Mayford, M., Kandel, E. R., and Tonegawa, S. (1996). Subregion- and cell typerestricted gene knockout in mouse brain. Cell 87, 1317–1326. Walsh, C., and Cepko, C. L. (1988). Clonally related cortical cells show several migration patterns. Science 241, 1342–1345. Wehr, M., Hostick, U., Kyweriga, M., Tan, A., Weible, A. P., Wu, H., Wu, W., Callaway, E. M., and Kentros, C. (2009). Transgenic silencing of neurons in the mammalian brain by expression of the allatostatin receptor (AlstR). J. Neurophysiol. 102, 2554–2562. Wetts, R. F., and Fraser, S. E. (1988). Multipotent precursors can give rise to all major cell types of the frog retina. Science 239, 1142–1145. Wu, Y., Wang, C., Sun, H., LeRoith, D., and Yakar, S. (2009). High-efficient FLPo deleter mice in C57BL/6J background. PLoS One 4, e8054. Yamamoto, M., Wada, N., Kitabatake, Y., Watanabe, D., Anzai, M., Yokoyama, M., Teranishi, Y., and Nakanishi, S. (2003). Reversible suppression of glutamatergic neurotransmission of cerebellar granule cells in vivo by genetically manipulated expression of tetanus neurotoxin light chain. J. Neurosci. 23, 6759–6767. Yamamoto, M., Shook, N. A., Kanisicak, O., Yamamoto, S., Wosczyna, M. N., Camp, J. R., and Goldhamer, D. J. (2009). A multifunctional reporter mouse line for Cre- and FLP-dependent lineage analysis. Genesis 47, 107–114. Zambrowicz, B. P., Imamoto, A., Fiering, S., Herzenberg, L. A., Kerr, W. G., and Soriano, P. (1997). Disruption of overlapping transcripts in the ROSA ßgeo 26 gene trap strain leads to widespread expression of b-galactosidase in mouse embryos and hematopoietic cells. Proc. Natl. Acad. Sci. USA 94, 3789–3794. Zhang, F., Wang, L. P., Boyden, E., and Deisseroth, K. (2006). Channelrhodopsin-2 and optical control of excitable cells. Nat. Methods 10, 785–792. Zhang, F., Aravanis, A. M., Adamantidis, A., de Lecea, L., and Deisseroth, K. (2007). Circuit-breakers: Optical technologies for probing neural signals and systems. Nat. Rev. Neurosci. 8, 577–581. Zinyk, D., Mercer, E. H., Harris, E., Anderson, D. J., and Joyner, A. L. (1998). Fate mapping of the mouse midbrain–hindbrain constriction using a site-specific recombination system. Curr. Biol. 8, 665–668. Zong, H., Espinosa, J. S., Su, H. H., Muzumdar, M. D., and Luo, L. (2005). Mosaic analysis with double markers in mice. Cell 121, 479–492.

C H A P T E R

T W E LV E

Genome-Wide Forward Genetic Screens in Mouse ES Cells Meng Amy Li, Stephen J. Pettitt, Kosuke Yusa, and Allan Bradley Contents 1. Introduction 2. Strategies for Genome-Wide Mutagenesis 2.1. Choice of mutagen 2.2. Designs for insertional mutagens 2.3. Methods and protocols for genome-wide insertional mutagenesis 3. Using Blm-Deficient ES Cells to Conduct Recessive Genetic Screens 3.1. Heterozygote-to-homozygote conversion by mitotic recombination 3.2. Inducible Blm-deficient system 3.3. Protocol for the generation of homozygous mutant library 4. Mutant Validation 4.1. Mutant identification based on splinkerette-PCR method 4.2. Validation of mutant genes 4.3. Genetic rescue to validate the genotype–phenotype causality 5. Concluding Remarks Acknowledgments References

218 220 220 225 226 229 229 231 231 232 233 235 236 238 239 240

Abstract Mouse embryonic stem (ES) cells are an attractive model system for investigating mammalian biology. Their relatively stable genome and high amenability to genome modification enables the generation of large-scale mutant libraries, which can be subsequently used for phenotype-driven genetic screens. While retroviral vectors have traditionally been used to generate insertional mutations in ES cells, their severe distribution-bias in the mammalian genome substantially limits genome-wide mutagenesis. The recent development of the DNA transposon piggyBac offers an efficient and highly versatile alternative for

Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, United Kingdom Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77012-9

# 2010 by Allan Bradley. Published by Elsevier Inc. All rights reserved.

217

218

Meng Amy Li et al.

achieving more unbiased mutagenesis. Furthermore, heterozygous mutations created by insertional mutagens can be converted in parallel to homozygosity by using Blm-deficient ES cells, allowing genome-wide loss-of-function screens to be conducted. In this chapter, we describe the principles underpinning genetic screens in mouse ES cells with examples of previously successful screens. Protocols are provided for piggyBac transposon-mediated mutagenesis, production of the corresponding homozygous mutants in a Blm-deficient genetic background, and methods for mapping and validation of mutations recovered from screens of such libraries.

1. Introduction Forward genetic screens enable nonhypothesis-driven discovery of gene function(s). They have been successfully conducted in a variety of experimental model organisms such as bacteria, yeast, fruit flies, and nematodes for dissecting numerous biological pathways. However, forward genetic screens in mammalian systems are more challenging to design and implement because of the complexity of the mammalian genome, long generation times, and limitations of the methods used for mutagenesis. The mouse was the first mammalian model organism in which the reverse genetic approach was developed to study specific genes. This became possible because of the development of mouse embryonic stem (ES) cells which can be cultured in vitro while still maintaining their potential to contribute to the development of all adult somatic cells and the germ line of mice (Evans and Kaufman, 1981). In addition, the development of homologous recombination techniques and their application to the ES-cell genome (Koller and Smithies, 1989; Thomas and Capecchi, 1987) enabled the generation of mice with directed modifications of endogenous genes (Schwartzberg et al., 1989; Snouwaert et al., 1992; Zijlstra et al., 1989). Currently, international consortia are systematically generating ES cells with targeted mutations in every gene (http://www.knockoutmouse.org/ about/komp). This gene-indexed mutant library, eventually covering every protein-coding gene in the mouse genome, is a very powerful resource for addressing gene function in both hypothesis-driven and nonhypothesisdriven manners. This targeted mutagenesis approach is described elsewhere in this volume. Despite the significant progress with targeted mutagenesis and the reliable establishment of alleles generated in ES cells into mice, these can only be examined one gene at a time. Even in centers with large organized efforts, the throughput is limited to a 100 or at most 200 different genes each year. The parallel generation of thousands of mutations in mammalian cell lines provides an alternative but equally important technology to address

Genome-Wide Forward Genetic Screens in Mouse ES Cells

219

gene function on a larger scale. Mammalian cell lines offer the advantage of incorporating screen-specific requirements for unbiased gene discovery of genes involved in specific biological pathways. Cell lines have substantially shorter discovery time scales and a significantly lower cost basis than wholeanimal mutagenesis. In addition, cell lines are simpler than whole animals due to the homogeneity of cellular phenotype and defined culture conditions which limit variation from environmental factors. Pluripotent mouse ES cell lines possess several unique features that make them particularly attractive as model systems. They can grow indefinitely and retain both their pluripotency and genomic stability through periods of extended culture. ES cells are also highly amenable to multiple rounds of sophisticated genetic manipulations without the loss of pluripotency. If required ES cells can be promoted to differentiate into an array of different cell types in vitro or they can generate a full spectrum of differentiated cell types in a chimera in vivo (Bradley et al., 1984). Forward genetic screens can be designed to recover dominant (or gain-of-function) and recessive (or loss-of-function) mutations. Dominant screens are configured to isolate genes that show the phenotype of interest when overexpressed or expressed in an unregulated manner. There are many examples of dominant screens; some of the more recent include those that have identified pluripotency factors. Chambers et al. (2003) conducted a screen using a mouse ES cell cDNA library and identified Nanog as a key factor in maintaining ES cell pluripotency. In 2006, Takahashi and Yamanaka (2006) described their groundbreaking gain-of-function screen for ES-cell-specific factors that can reprogram somatic cells to ES-like pluripotent cells, termed induced pluripotent stem (iPS) cells. Following similar thinking for cell-fate reprogramming, another group has used a dominant screen to identify factors that can lead to transdifferentiation between two different somatic cell types without going through the pluripotent ES-cell stage (Vierbuchen et al., 2010). These examples illustrate how overexpression of exogenous factors can cause gain-of-function phenotypes. These screens have relied on the phenotype conversion by the activity of single genes from a genome-wide library or by combinations of genes from an expression library with a limited repertoire. Recessive screens are designed to isolate genes showing a phenotype of interest when inactivated. In a mammalian genome, such screens are challenged by the diploid nature of the genome, since usually both copies of a gene need to be inactivated to evoke a phenotypic change. In cell culture, the probability of biallelic inactivation by two independent events is extremely low. If ES cells are used, single-allele-mutated ES cells can be used to derive heterozygote mice through germ-line transmission and intercrossing of the F1 generation is required to obtain homozygote mutant animals. This process is slow and costly; therefore, strategies to directly obtain homozygote ES-cell mutants have been developed. Several methods

220

Meng Amy Li et al.

have been developed which exploit the naturally occurring phenomenon of loss of heterozygosity (LOH) to recover rare homozygous mutant cells from heterozygous cells (Guo, Wang and Bradley, 2004; Lefebvre et al., 2001; Mortensen et al., 1992; Yusa et al., 2004). The rate of spontaneous LOH is actually quite low, thus this system was initially only practical on a gene-bygene basis. The use of Blm-deficient ES cells removed this limitation. The 20-fold increase in the rate of LOH in Blm-deficient ES cells greatly enhances the rate of segregating homozygous mutants from heterozygote counterparts (Guo and Bradley, 2004; Yusa et al., 2004). Using this genetic background together with efficient mutagenesis, recessive genetic screens have been successfully conducted in ES cells. These screens cover a variety of biological pathways and include retroviral resistance, RNAi processing, and toxin resistance screens (Trombly et al., 2009; Wang and Bradley, 2007; Wang et al., Unpublished). The main steps involved in forward genetic screens are shown in Fig. 12.1: 1. Genome-wide mutagenesis and enrichment for the mutants by selection. 2. Pooling of mutants to generate sectored mutant libraries. 2B. Culture for sufficient time to segregate homozygous mutants using Blm-deficient ES cells (for recessive genetic screens only). 3. Phenotypic screens or selection. 4. Mutation identification. 5. Genetic and functional validation of the mutants. The chapter is organized in three parts. Section 2 describes the tools and strategies for a genome-wide mutagenesis in mammalian systems, which are applicable to both dominant and recessive genetic screens. We describe DNA transposon piggyBac-based insertional mutagenesis, since in our view this is the most powerful and versatile tool available. Section 4 focuses on utilizing the Blm-deficient ES cells for the generation of homozygous mutant library. Section 5 describes the methods for mutant identification and validation.

2. Strategies for Genome-Wide Mutagenesis 2.1. Choice of mutagen In a forward genetic screen, the function of a gene can be assigned to a specific biological process by analyzing the phenotypic consequences when the gene’s activity is altered. There are three main categories of agents which can be used in ES cells as well as in animals, namely chemical, physical, and biological mutagens. Each of them has its own characteristics

221

Genome-Wide Forward Genetic Screens in Mouse ES Cells

Mutagenesis Gain-of-function screen

Loss-of-function screen

Mutant generation and enrichment by selection

Mutant generation and enrichment by selection

Mutant pooling

Mutant pooling

Homozygote conversion by LOH

Phenotypic screening

Phenotypic screening

Mutant locus identification

Mutant locus identification Homozygosity validation

Functional validation

Functional validation

Figure 12.1 The experimental steps involved in forward genetic screens in mouse ES cells.

for the nature of mutations, efficiency of mutagenesis, and genome coverage. A summary of different mutagens are shown in Table 12.1. 2.1.1. Chemical agents Chemical agents such as N-ethyl-N-nitrosourea (ENU) and ethylmethanesulfonate (EMS) have been efficiently used to create mutations in mouse ES cells (Chen et al., 2000; Munroe et al., 2000). The main drawback of such mutagens is the difficulty in identifying the causal mutation in the complex mammalian genome. The main method for the identification of the causal mutation is by sequencing or complementation. With the rapid development in whole-genome, exon-, and RNA-sequencing

Table 12.1 Comparison of different classes of mutagens Mutagen type

Mutagenesis characteristics

Chemical mutagen (ENU, EMS)


Efficient mutagen
Difficult to control number of mutations
Generate point mutations, small insertions, and deletions per cell
Mostly cause loss-of-function, but can be gain-of
Difficult to isolate causal mutation

function mutation

Physical mutagen (gamma ray)


Unbiased genome-wide mutagenesis
Efficient mutagen
Generate large regions of deletions, amplifications, and

rearrangements


Cause both gain- and loss-of-function mutations
Unbiased genome-wide mutagenesis

Retrovirus (MuLV)
Efficient mutagen
Insertional mutagenesis
Can achieve both gain-of-function and loss-of-function mutations depending on the design
Titratable copy number per cell
Molecular tag for mutant identification
Highly efficient mutagen DNA transposon
Insertional mutagenesis (piggyBac)
Can achieve both gain-of-function and loss-of-function mutations depending on the design
Large cargo capacity
Molecular tag for mutant identification
Easy to achieve genetic reversion without footprint


Large genomic regions being altered
No control over the size and type of

genomic changes


Laborious process to isolate causal

mutation
Severe hot- and cold-integration spots in

mammalian genome


Limited cargo size
Prone to silencing in ES cells


Possible biased ‘‘hot-spots’’ for genomic

integrations


Possible ‘‘Local hopping’’ effect
Transposition efficiency is locus- and

methylation-status dependent

Genome-Wide Forward Genetic Screens in Mouse ES Cells

223

technologies, the bottleneck of mutant identification may be improved and become cost-effective in the near future. 2.1.2. Physical agents Physical agents such as gamma-ray irradiation have been used to efficiently generate genome-wide mutations in mouse ES cells (Schimenti et al., 2000; You et al., 1997). The mutations are typically large deletions, duplications, amplifications, and other rearrangements, causing both loss- and gain-offunction mutations. The advantages of large alterations are that the genome can be covered with relatively a small number of mutants. However, identifying casual gene–phenotype relationship is difficult, because of the large number of genes affected in each clone. Strategies such as comparative genomic hybridization (CGH) arrays can be used to locate the regions of alterations in the genome. Causal regions can be further narrowed down by identifying the commonly altered region in independent cell lines and followed by complementation assays to reintroduce the genes within the region to rescue the phenotype. 2.1.3. Biological agents Retroviral and DNA transposons are commonly used as recombinant vectors to integrate mutagens into the host genome. These vectors are flexible and can accommodate different molecular designs to achieve mutagenesis. Additionally, they serve as molecular tags for the identification of the mutated gene, a significant advantage over chemical and physical mutagenesis. 2.1.3.1. Retroviral vectors Retroviruses have a long history of use for the transduction of mammalian cells. The murine leukemia virus (MuLV)based retroviral vector has been used in recessive screens in ES cells and led to successful identifications of known and novel factors in the DNA mismatch repair, siRNA processing pathway, and the process of retroviral infection (Guo and Bradley, 2004; Trombly et al., 2009; Wang and Bradley, 2007). However, it has become increasingly apparent that retroviral integrations have a severe nonrandom genome distribution, with both ‘‘hot-’’ and ‘‘cold-’’ integration spots in the mammalian genome. The large resource of retroviral gene-trap clones, TIGM OmniBank II, provides a useful dataset for analyzing the retroviral integration patterns (Hansen et al., 2008). The bank possesses over 350,000 ES cell clones, with 10,433 unique genes. However, only 27% of the genes in this resource have been trapped once and the rest of the genes trapped at multiple times. With such highly uneven integration patterns, mutating genes in the retroviral integration ‘‘cold-spots’’ is difficult and requires highly redundant coverage of the genome and many genes will remain untouched.

224

Meng Amy Li et al.

2.1.3.2. DNA transposons DNA transposons are DNA elements that can mobilize in a host genome. Transposition is catalyzed by a transposase enzyme which binds to the inverted terminal repeats and ‘‘cuts and pastes’’ the transposon from one location to another. These unique properties have been harnessed extensively and used to generate a variety of nonautonomous vectors for transgenesis and insertional mutagenesis in a wide range of model organisms. The lack of active DNA transposons in mammals had impeded the application of such technologies in mammalian genomes until over a decade ago. The reactivation of a Tc1-like DNA transposon, named Sleeping Beauty, ignited the development of transposon technologies for use in mammalian systems (Ivics et al., 1997). More recently, the transposon tool-kit for applications in mammalian genomes has been extended by the discovery and development of several other elements from different transposon families (Ivics et al., 2009). piggyBac, isolated from the cabbage moth Trichoplusia ni (Cary et al., 1989), is a highly efficient transposon in both insects and vertebrates. Several direct-comparison studies have suggested that piggyBac is the most efficient transposon among all known DNA transposons in mammalian cells (Liang et al., 2009; Wu et al., 2006). In addition, piggyBac possesses several advantageous properties as an insertional mutagen. Its integrations in ES cells have been shown to be much more random than comparable retroviral vectors (Wang et al., 2008a,b). Even in small libraries of piggyBac-mediated genetrap clones, 8% of the trapped genes were not previously identified in the retroviral-based gene-trap resource OmniBank II (Wang et al., 2008a). Thus, piggyBac integrations provide access to genes which have not been tagged in a more than 20-fold saturated retroviral insertion library. Comparable DNA mismatch repair screens have been conducted using gene-trap libraries constructed with either retroviral or piggyBac vectors. Similar complexity libraries yielded all known mismatch repair genes in the piggyBac-based library whereas just one of the known genes was identified in the retroviral library (Guo and Bradley, 2004; Wang et al., 2008a). Several other properties of piggyBac make it an attractive insertional mutagen. piggyBac-mediated integrations are biased toward actively transcribed regions of the host genome (Ding et al., 2005; Liang et al., 2009), making it advantageous for gene perturbation. piggyBac also has a much large cargo capacity compared with retroviral vectors and the Sleeping Beauty transposon, which allows the accommodation of complex and modular molecular designs. For instance, a combination of reporter genes and expression constructs can be combined into a single vector with ease. Finally, piggyBac has the unique property of excision without leaving footprint, which allows the complete reversion of the mutation to be assessed in phenotypic rescue experiments (Ding et al., 2005; Yusa et al., 2009). In this chapter, we have focused on the use of piggyBac as a means of genome-wide mutagenesis because of these advantages compared with the

225

Genome-Wide Forward Genetic Screens in Mouse ES Cells

other vectors described. However, the methods and protocols provided here can be adapted for other transposon-based insertional mutagens.

2.2. Designs for insertional mutagens The basic types of vector design for both gene inactivation and overexpression are shown in Fig. 12.2. For gain-of-function experiments, a strong exogenous promoter can be engineered to drive the overexpression of A

Gain-of-function Exon

PB3

PGK-Neo

Promoter

SD

PB5

Exon

Exon

B PB3

PGK-Neo Promoter

cDNA1 T2A cDNA2 T2A cDNA3

PB5

Loss-of-function C Exon

PB3

SA

bgeo

PB5

Exon

Exon

Truncated protein joined to bgeo

D Exon

PB3

SA T2A bgal T2A Neo

PB5

Truncated protein

Exon

b-gal

Exon

Neo

E Exon

F

PB3

SApA PGK-Neo T2A TK PB5

Exon

Truncated protein

Neo

SD

Exon

Exon

TK

Dual function Exon

PB3 SApA PGK-Neo Promoter

PB5

Exon

Figure 12.2 Insertional mutagen designs for the gain-of-function and loss-of-function genetic screens, see text for details. Arrows represent transcription initiation sites; gray lines under the exons, transcripts; SA, splice acceptor; pA, polyadenylation signal; SD, splice donor; TK, thymidine kinase; PB5 and PB3, piggyBac 50 - and 30 -inverted terminal repeats; PGK, mouse phosphoglycerate kinase promoter.

226

Meng Amy Li et al.

either a full length or truncated gene product (depending on where the insertional mutagen lands with respect to the transcription unit, Fig. 12.2A). An alternative method is to overexpress genes from a cDNA expression cassette (Fig. 12.2B). In this way, a cDNA library can be screened for the phenotype using either individual genes or groups of genes in combination. For generating loss-of-function mutants, gene-trap cassettes such as SAbgeo are often used to allow gene inactivation through the formation of fusion transcripts between the upstream exon and the reporter (Fig. 12.2C). However, this strategy is restricted to protein-coding genes, which must be expressed in the same reading frame as the selectable marker. An improved version of this vector is to include the viral self-cleaving 2A peptide (Palmenberg et al., 1992; Ryan et al., 1991) to connect the trapped exon with the reporter in a single transcript without the production of fused protein products (Fig. 12.2D). In this way, the reporter function is not compromised by a chimeric fusion with the translated portion from the upstream exon. Designs which separate gene trapping from selection for integration also allow an increase in the mutagenic coverage of the genome by trapping genes with no reading frame restrictions (Fig. 12.2E). This design also extends the mutagenic coverage to genes that are not expressed at the time of mutagenesis. For this reason, if the screens are conducted in different cell types from mutagenesis, for example, in screens where reprogramming or differentiation is part of the phenotypic readout, the strategy for separating out the mutagenic unit from the reporter is desirable. It is also possible to combine both the gene activating and inactivating units together (Fig. 12.2F) to achieve dual functions. Such a strategy has been extensively used in somatic mutagenesis in mice for cancer-gene discovery (Collier et al., 2005). These details are described elsewhere in this volume.

2.3. Methods and protocols for genome-wide insertional mutagenesis When deciding the methods for mutagenesis, the requirements for the copy number of the insertional mutagen per cell must be considered. In most cases, a single-copy per cell is important to facilitate proof that the insertion causes the phenotype. In ES cells, transposon delivery can be achieved by mobilization of the transposon from a transfection of plasmid harboring the transposon or by mobilizing a transposon which has been introduced to the genome by gene targeting. Transfection-based ‘‘vector-to-genome’’ delivery results in relatively unbiased genome-wide integration; however, careful titration of the amount of the plasmid is required to control the copy number (Wang et al., 2008a). Intragenomic mobilization of a targeted single-copy transposon offers the advantage of maintaining a stable copy number of the mutagen. Cells with mobilization events can be enriched based on selection for the

Genome-Wide Forward Genetic Screens in Mouse ES Cells

227

transposon excision from the original donor locus (Liang et al., 2009; Wang et al., 2008b) and reintegration elsewhere in the genome. For genome-wide screens in ES cells, both methods can produce a large number of mutants with single-copy integration per cell. 2.3.1. Protocol for transfection of plasmid harboring the transposon The conditions to achieve predominantly single-copy piggyBac integration have been titrated by adjusting the amount of the plasmid harboring piggyBac transposon while keeping the transposase constant. In this protocol, SA-b geo- or T2A-bgal-T2A-neo-based gene-trap vectors are used (Fig. 12.2C and D). Cotransfection of 1  107 ES cells with 100 ng of transposon donor plasmid and 10 mg of insect-derived transposase expression plasmid can give rise to around 800 gene-trap G418 resistant colonies (Wang et al., 2008a). Over 95% of the colonies have single-copy piggyBac integration. Independent electroporations will provide different mutant repertoires; therefore, the genome-wide coverage can be achieved by making several independent mutant libraries in parallel. If another selection marker scheme is used, for example, the Pgk-neo cassette (Fig. 12.2E) to select for integration events without trapping-mediated reporter expression, a titration experiment should be conducted with different amounts of transposon donor plasmid. The copy number for different conditions can be assessed by Southern blotting using piggyBac terminal repeat as the probe (Wang et al., 2008a). 1. Prepare both transposon donor plasmid and transposase-containing plasmid using commercial column-based preparation to obtain good quality DNA. Dissolve the DNA pellet in sterile Tris–EDTA (TE) buffer. 2. ES cells should be around 80% confluent on the day of electroporation. 3. Two to 4 h before electroporation, replace the medium on the ES cells. 4. Wash the ES cells with PBS twice, add trypsin solution, and incubate at 37  C for 5 min. 5. Add 5 ml of ES cell medium and suspend the cells by vigorous pipetting. Centrifuge the cell suspension at 1100 rpm for 3 min at room temperature. 6. Aspirate the supernatant and resuspend the pellet in PBS. Repeat steps 5 and 6 again. 7. Count the cells and resuspend the cells in PBS to a final concentration of 1.4  107 cells/ml. 8. Mix 0.7 ml of the cell suspension with 100 ng of transposon donor plasmid and 10 mg of piggyBac transposase plasmid. 9. Transfer the mixture to a 0.4 cm electroporation cuvette, and electroporate in the Bio-Rad Gene Pulser at 230 V, 500 mF. Incubate for 5 min at room temperature.

228

Meng Amy Li et al.

10. Transfer the contents to a Falcon tube and mix with ES cell medium. One-tenth of the mixture can be plated on a 90-mm tissue culture plate with feeders, and the rest to another plate. 11. The next day, apply ES cell medium containing 180 mg/ml G418. Change the drug containing medium daily. 12. Eight to 10 days after the electroporation, the colonies can be seen. 13. Stain the plate with 1/10 of the electroporation mixture with 2% (w/v) methylene blue in 75% (v/v) ethanol. The number of colonies can be counted to give an estimate of the mutant library complexity. The colonies of other plate can be pooled for dominant screening or expanded to achieve homozygote conversion in recessive screens (Section 3). 2.3.2. Protocol for using intragenomic mobilization of the preintegrated transposon When the transposon is preintegrated in the genome, transfection of the transposase-containing plasmid mediates the mobilization of the transposon from the original location to elsewhere in the genome. A selection scheme is necessary to eliminate the cells with the transposon still residing in the donor locus. This can be achieved by engineering the transposon into the X-linked Hprt locus. When the transposon is present in Hprt, it disrupts its expression; therefore, such cells are sensitive to HAT (hypoxanthine, aminopterin, and thymine). Upon excision, the function of Hprt is restored; therefore, cells become resistant to HAT. Additionally, a positive selection marker, for example, the gene-trap cassette (Fig. 12.2D), can be used to select for transposon reintegration events. The excision efficiency using Hprt as the donor locus is measured to be around 0.5–1% of total electroporated cells (Liang et al., 2009; Wang et al., 2008b). Steps 1–8 are the same as described above. 9. Mix 25 mg of transposase-containing plasmid with 0.7 ml of the ES cell suspension. 10. Transfer the mixture to a 0.4 cm electroporation cuvette, and electroporate in the Bio-Rad Gene Pulser at 230 V, 500 mF. Incubate for 5 min at room temperature. 11. Transfer the content to a 90-mm tissue culture plate with feeders, and change media daily for the next 3 days. If the cells become confluent during this time, passage the cells. 12. On day 4 after electroporation, trypsinize the ES cells and replate. Use 1  104 cells on one 90-mm tissue culture plate with feeders and 1  106 cells on another. 13. The next day, apply ES cell medium containing 180 mg/ml G418 and HAT. 14. Change selective medium daily for 8–10 days.

Genome-Wide Forward Genetic Screens in Mouse ES Cells

229

15. Stain the 1  104 cells plated. The number of colonies can be counted to give an estimate of the mutant library complexity. The colonies of the other plate can be pooled for dominant screening or expanded to achieve homozygote conversion in recessive screens. Culturing cells without selection for the first 3 days allows the transposase activity to decay before the replating and subsequent selection. If the selection is commenced immediately after the electroporation, the colonies that form will be a mixture of cells with different transposon-integration events since transpositions are still occurring in the dividing daughter cells, making the library-complexity estimation difficult. This will affect the downstream procedure using Blm-deficient ES cells (Section 3). 2.3.3. Mutant library-complexity assessment Library complexity is defined as the number of independent mutants in a library. The number of colonies obtained after the mutagenesis and enrichment (Fig. 12.1) is usually an accurate representation of library complexity, assuming that each colony is clonal and possesses nonredundant integrations. This number is used to estimate the genome coverage of the mutant library and it is also important for estimation of the degree of library expansion required for the heterozygote-to-homozygote conversion step in the Blm-deficient ES-cell system (Section 4). The mouse genome consists of roughly 25,000 genes, around half of which are expressed in ES cells. For example, if a mutant library consists of 10,000 independent mutants using a trap-based method for a loss-of-function screen, in theory, the expressed ES-cell genome is covered once. However, in practice, gene sizes vary significantly and insertional mutagens may not be completely random. Therefore, multiple rounds of screening with 10–20 mutant pools (e.g., 10,000 each) generated independently for each round of screening will provide a more complete coverage of the genome.

3. Using Blm-Deficient ES Cells to Conduct Recessive Genetic Screens 3.1. Heterozygote-to-homozygote conversion by mitotic recombination Patients with the autosomal recessive disorder Bloom syndrome are predisposed to cancers due to the loss-of-function mutations in the BLM gene. BLM encodes a member of the RecQ helicase family, which is highly conserved in evolution and functions to unwind the DNA helix. BLM functions in a multiprotein complex to suppress crossing-over and resolve DNA structures that arise during the process of homology-directed repair

230

Meng Amy Li et al.

(Sung and Klein, 2006). At the cellular level, BLM deficiency shows a characteristic phenotype of a high frequency of sister chromatid exchange and genomic instability. Blm-deficient mouse ES cells also display a genome-wide hyper-recombination phenotype and consequently an elevated LOH rate. This high frequency of crossing-over between homologous nonsister chromosomes and their segregation in Blm-deficient ES cells generates homozygous mutant cells from their heterozygote counterparts. Several successful recessive screens have been conducted in Blm-deficient ES cells using insertional mutagens (Guo and Bradley, 2004; Trombly et al., 2009; Wang and Bradley, 2007; Wang et al., 2008a). Figure 12.3 shows the mechanism of genotype conversion. In the G2 phase of the cell cycle, homologous recombination between nonsister chromatids of homologous chromosomes takes place at a low but detectable frequency, resulting in a ‘‘mitotic chiasma.’’ When a parental cell possesses a heterozygous mutation, two outcomes can be expected following chromosome segregation. In outcome 1, two heterozygous daughter cells appear although one of the daughters has a hybrid chromosome. On the other hand, in outcome 2, one wild-type and one homozygous daughter cell are generated. Therefore, a homozygous mutant arises from the parental cell with a heterozygous mutation. In a wild-type background, the frequency of mitotic recombination is extremely low and measured to be 3.5  10 5 events/cell/generation using Luria–Delbru¨ck fluctuation analysis. The frequency is 4.2  10 4 events/cell/generation in Blm-deficient cells. In other words, one LOH event can be expected in every 2400 divisions, that is, 12 generations. Roughly, when a single cell harboring a heterozygote mutation is expanded to 1  104 cells, a few homozygous mutants are converted in the culture. In practice, the LOH rate differs across the physical length of the chromosomes, with LOH rates higher toward +/−

+/−

−/−

+/+

Segregation

Replication

+/−

Mitotic recombination

Outcome 1

Outcome 2

Figure 12.3 Generation of homozygote mutants by mitotic recombination followed by segregation.

Genome-Wide Forward Genetic Screens in Mouse ES Cells

231

telomeres compared to centromeric ends. Therefore, the genes residing closer to the centromeres may require longer expansion time to obtain homozygote mutants than the genes located close to the telomeres. Prior to expansion, heterozygous mutants can be pooled. It is important to know the library complexity in order to estimate the degree of expansion required to obtain a reasonable number of homozygote mutants in a library. We normally generate 1000 mutants per pool. When this pool is expanded to 1  107 cells in total, which fill a 90-mm dish, each mutant is represented by around 1  104 cells, and a few homozygous mutants are theoretically generated in the culture. For whole-genome mutagenesis, 10 pools should be able to cover nearly all genes expressed in ES cells.

3.2. Inducible Blm-deficient system Despite the successful applications of the Blm-deficient ES cells, there are a few residual practical issues. In these cells, mitotic recombination is taking place continuously. Over prolonged culture, spontaneous recessive mutations may become homozygous. Such mutations interfere with a screen and are extremely difficult to identify. Genes that show synthetic-lethal phenotype with Blm deficiency cannot be screened, although such genes are rare. An alternative system with a conditional Blm allele was generated by knocking in the tet-off cassette upstream of the initiation codon of Blm protein (Yusa et al., 2004). The tet-off cassette consists of a splice acceptor (SA) site, tetracyclinedependent transactivator (tTA), polyadenylation (pA) site, and a tetracycline-responsive element (TRE). Driven by the Blm promoter, tTA proteins are produced which bind to TRE and activate the transcription from the ‘‘ATG’’-containing exon 2 of the endogenous Blm. The addition of doxycycline (Dox) inhibits the binding of tTA to the TRE, thus the transcription of Blm mRNA is repressed. During repression, the cell shows the phenotype equivalent to that seen in Blm-deficient ES cells. After withdrawal of Dox from the culture medium, tTA proteins bind to TRE again and reactivate the Blm expression. Hence, Blm deficiency can be induced only when mitotic recombination is required for the generation of homozygous mutants. During the normal cell expansion and screening, Blm is expressed to maintain the genomic integrity.

3.3. Protocol for the generation of homozygous mutant library 1. Generate genome-wide heterozygote mutations as described in Section 2.3 and calculate the complexity of each mutant pool based on colony number.

232

Meng Amy Li et al.

2. Estimate the cell number required at the end of culture using the following equation: N ¼ C  104  m  4:21

3. 4. 5.

6. 7.

ð12:1Þ

where N is the total number of cells at the end of culture, C is the complexity of the pool, and m is the number of homozygous mutants of each integration required. For example, when a pool contains 1000 heterozygous mutants and at least five homozygous mutants per mutation are required, the mutant pool needs to be expanded until it has reached 11.9  106 cells. Trypsinize the colonies following the protocol in Section 2.3; since colonies are harder to resuspend into single-cell suspension compared to a usual passage, keep dishes in 37  C incubator for 15 min. Add ES-cell medium to quench the trypsinization, resuspend cells, and collect the cells by centrifugation at 1100 rpm for 3 min. Resuspend the cells in ES-cell medium and plate them into a 6-well plate. If Dox is used to regulate Blm, add Dox to the culture medium at 1 g/ml. Alternatively, Dox administration can be initiated just after transposon mutagenesis (step 12 in plasmid-mediated mutagenesis, or step 15 in intragenomic mutagenesis). Keep expanding the culture until cells reach the number calculated in step 2. Withdraw Dox if inducible Blm cells are used. Subject cells to the phenotypic screen.

The heterozygote-to-homozygote conversion is a stochastic process. Therefore, it is useful to culture independent mutant pools, as this will limit the ‘‘jack-pot’’ effect of a single mutant which converts to homozygosity early during pool expansion. Although the design of the phenotypic screens of interests is beyond the scope of this chapter, it is worth noting that the screening method must be sensitive and specific enough to be able to isolate only a few relevant homozygote mutants from a large pool of irrelevant cells.

4. Mutant Validation Once mutant clones have been isolated, the next stage is to identify the insertion site in each clone and then to verify that the mutagen is responsible for the phenotype. Insertional mutagenesis facilitates both steps. Insertional mutagens act as unique tags to identify the mutated loci and a causal relationship between a mutated locus and the phenotype can be established by phenotypic rescue of the mutant by removing the mutation.

233

Genome-Wide Forward Genetic Screens in Mouse ES Cells

4.1. Mutant identification based on splinkerette-PCR method A variety of PCR-based methods, including inverse PCR (Ochman et al., 1988), vectorette PCR (McAleer et al., 2003), and splinkerette PCR (Devon et al., 1995), have been developed for the identification of insertional mutation sites. Splinkerette PCR has become the most widely used method to amplify mutagen–host genomic junctions. Splinkerette PCR is a ligation-mediated PCR method, which utilizes a ‘‘hairpin loop’’ structure within the splinkerette adaptor to overcome the problem of nonspecific amplification that arose through ‘‘end-repair priming’’ (McAleer et al., 2003). The details of the principles and methods of splinkerette PCR are described elsewhere (Uren et al., 2009). Briefly, the genomic DNA is digested with a restriction enzyme and splinkerette adaptors with the compatible cohesive ends are ligated to all the resulting fragments. The ligated fragments are used as the template in the nestedPCRs with primers specific for the vector and the adaptor sequence (Fig. 12.4). Here, we focus on the description of the adaptation of this method for the isolation of piggyBac integration sites. The adaptations of this method for retroviruses and the Sleeping Beauty insertion-site isolation have also been described (Liang et al., 2009; Uren et al., 2009; Wang and Bradley, 2007).

4.1.1. Splinkerette-PCR protocol for isolating piggyBac integration sites 1. Digest 2–3 g of genomic DNA with 4 units of Sau3AI in a 30 l reaction overnight. The digest can be done in 96-well format by preparing DNA directly from cells in gelatinized 96-well plates (Ramirez-Solis et al., 1995). 2. Heat-inactivate the enzyme at 65  C for 20 min.

Knocked-in Blm

E1

SA

tTA

pA

TRE

E2

E3

Blm ATG

Figure 12.4 Illustration of the inducible Blmtet allele. E1, E2, E3, exons 1, 2, and 3 of the Blm gene; SA, splice acceptor; tTA, tetracycline-dependent transactivator; pA, polyadenylation site; TRE, tetracycline-responsive element.

234

Meng Amy Li et al.

3. Prepare the Splinkerette adaptor. Mix 150 pmol of each oligonucleotide (Table 12.2) in 100 l water with 0.5 NEB buffer 2 and heat to 95  C for 10 min to denature. Anneal the oligonucleotides by cooling the mixture slowly to room temperature. Annealed Splinkerette adaptors can be stored at 20  C. 4. Set up ligation reactions as follows: 5 l of digested genomic DNA, 3 l Splinkerette adaptor, 2 l T4 ligase buffer, 1 l T4 ligase, 9 l water. Incubate at 16  C overnight. 5. Inactivate the ligase by heating to 65  C for 10 min. 6. Use 1 l of ligation reaction as the template for the first PCR. Any standard polymerase can be used, for example, Platinum Taq or Thermo Start. Set up a standard 25 l reaction using the primers shown in Table 12.2 and use 30 cycles as follows: 94  C for 30 s, 62  C for 30 s, 72  C for 90 s, followed by a final extension at 72  C for 5 min. 7. Set up a secondary PCR with primers shown in Table 12.2, using 1 l of the primary PCR diluted 1:100 as the template and the same PCR condition as above.

Table 12.2 Splinkerette-PCR primers for piggyBac integration-site isolation

Adaptor linear strand

CGAAGAGTAACCGTTGCTAGGAGAG ACCGTGGCTGAATGAGACTGGTGT CGACACTAGTGG GATCCCACTAGTGTCGACACCAGT CTCTAATTTTTTTTTTCAAAAAAA CGAAGAGTAACCGTTGCTAG GAGAGACC GTGGCTGAATGAGACTGGTGTCGAC

Adaptor hairpin strand with 50 -GATC overhang Adaptor primary-PCR primer Adaptor secondary-PCR primer PB30 -ITRa primary-PCR TAAATAAACCTCGATATACAGACC primer GATAAA PB30 -ITR secondary-PCR ATATACAGACCGATAAAACACAT primer GCGTCAA PB30 -ITR sequencing primer TTTTACGCATGATTATCTTTAAC GTACGTC PB50 -ITR primary-PCR CAAAATCAGTGACACTTACCGCA primer TTGACAA PB50 -ITR secondary-PCR CTTACCGCATTGACAAGCACG primer CCTCACGGG PB50 -ITR sequencing primer TTAGAAAGAGAGAGCAATATTT CAAGAATG a

ITR, piggyBac inverted terminal repeat.

Genome-Wide Forward Genetic Screens in Mouse ES Cells

235

8. Run a small sample of the products on a 2% agarose gel. If the product is unique it can be sequenced directly using the nested sequencing primers (Table 12.2). If multiple products are present, they can be gel purified using a commercially available spin-column kit prior to sequencing. 9. Before mapping, sequences should be processed as follows: (a) Ensure the piggyBac inverted terminal repeat–host genome junction is visible. The piggyBac inverted terminal repeat sequence can be removed. If the sequence read extends into the adaptor sequence, it can also be clipped. (b) Scan for the presence of the restriction site used (GATC for Sau3AI) within the sequence. This can indicate a chimeric molecule formed by ligation of two genomic fragments—in these cases, only the sequence between the piggyBac inverted terminal repeat and the restriction site should be used. A useful web-based tool for these sequence manipulations and for mapping the processed sequences using the SSAHA2 algorithm is available at: http://www.sanger.ac.uk/cgi-bin/teams/team113/imapper.cgi (Kong et al., 2008). 4.1.2. Filtering sister clones An optional sister-clone filtering step may be useful to establish clonal relationship among the mutants which can arise from a clonal expansion of a single mutant in a pooled library. This can reduce duplicative effort in identifying host genome–transposon junction fragments. Such clonal relationship can be easily detected by Southern blotting. DNA from the mutants is digested using an enzyme that cuts within the transposon and a probe internal to the transposon is used to identify transposon–genome junction fragments. Clones with the same insertion will have the same size transposon–genome junction fragment. This analysis will also reveal whether any of the mutant clones harbor multiple insertions, in which case extra care must be taken to determine which integration site is causal.

4.2. Validation of mutant genes For mutants isolated from dominant screens, heterozygote mutants will give rise to the phenotype. Therefore, once the mutagen integration sites are identified, genotype–phenotype causality and functional validation can be carried out. For mutants identified from recessive screens, the mutants can be further confirmed for homozygosity. This is most conveniently done using a three-primer competition PCR. Two of the primers are locus specific and can amplify a wild-type product that spans the transposon insertion site. The third primer is mutagen-specific and amplifies the mutagen–host junction together with one of the locus-specific primers.

236

Meng Amy Li et al.

These three primers are used in equimolar amounts in a conventional PCR reaction. In homozygous mutants, only the product arising from the mutagen–host genome junction should be amplified. Heterozygous clones give rise to both the wild-type and mutant products. Another possibility is to design a locus-specific Southern blotting strategy to discriminate the mutant and wild-type alleles. This has the advantage of not being sensitive to contamination from wild-type DNA arising from the feeder layer or an impure clone of cells. Finally, to prove that the mutagenesis is effective, reverse transcriptase PCR (RT-PCR) with primers hybridizing to the up- or downstream exon and the mutagenic unit can be used to confirm the gene’s perturbation. This is applicable to both dominant and recessive genetic screens.

4.3. Genetic rescue to validate the genotype– phenotype causality Establishing causality between the phenotype and the observed insertional mutagen is important, as spontaneous background mutations arising randomly in culture can be selected for in the screen. Cells with background mutations also contain a mutagen integrated in the genome, which is irrelevant to the phenotype. The best way to formally prove that the phenotype is due to the mutagen is to remove it from the genome, which should revert the phenotype. There are several ways to accomplish this, depending on the mutagen used. Transposons can be remobilized by resupplying the transposase. Excision is precise for piggyBac mobilization, so mutants with insertions even in exons can be fully reverted to their wild-type genotype. The only mechanism to completely remove a retrovirus from the genome is to use homologous recombination, but this requires constructing a targeting vector for each allele being investigated. However, both transposons and retroviral vectors can be designed with a cargo that is removable by sitespecific recombinases such as Cre or Flp. In a retroviral vector, the loxP or FRT site can be conveniently included in the U3 region of the long terminal repeat (LTR). Expression of Cre or Flp mediates the deletion of the cargo from the integrated provirus and leaves a single LTR in the locus, which usually achieves reversion when the mutagen is in an intron. In piggyBac vectors, the loxP or FRT sites are internal to the inverted terminal repeats; in this case, reversion with a recombinase leaves a mini cargo-less transposon in the locus. The efficiency of mutagen removal can vary significantly in a locusdependent manner and selection strategies can enrich the revertants. A negative selectable marker, such as Herpes Simplex Virus thymidine kinase (TK), can be included in the vector to facilitate direct selection of cells that have lost the transposon or cargo (Fig. 12.2E). If such selection scheme is

Genome-Wide Forward Genetic Screens in Mouse ES Cells

237

not available, transient selection using puromycin (Taniguchi et al., 1998) to enrich for cells that have been transiently transfected with the recombinaseor transposase-expression plasmid can enhance the reversion efficiency.

4.3.1. Mutant reversion by transient transfection of transposase or recombinase In situations where the transposon contains a HSV-Tk gene for negative selection: 1. Follow steps 1–7 described in the protocol in Section 2.3 and electroporate 5  106 cells with 20 g of transposase, Cre, or Flp-expression plasmid. 2. Plate cells onto a 90-mm plate in ES-cell medium. Culture cells for 3 days to allow the decay of TK mRNA and protein. 3. Harvest cells by trypsinization, count the cells, and replate 5  105 per 90-mm plate. 4. The next day, apply ES-cell medium containing 200 nM FIAU (1-(2-deoxy-2-fluoro-b-D-arabinofuranosyl)-5-iodouracil). 5. Pick individual clones after 8–10 days. 6. Analyze clones by using the triple-primer competition PCR method described in Section 4.2 to check that the genotype reversion has occurred. 7. Subject cells to phenotypic analysis. When reversion is carried out using piggyBac transposase, the genotype of the revertants is identical to wild-type cells. Therefore, it is important to conduct a control without transposase to ensure that the original clone is not mixed with wild-type cells. 4.3.2. Isolation of revertants with transient puromycin selection for enrichment Transient puromycin selection can be used to enrich cells transfected with the transposase, Cre, or Flp-expression plasmid to enhance the reversion efficiency (Taniguchi et al., 1998). The transposase or recombinase expression plasmid must also contain a puromycin-resistant cassette and the cells subjected to selection should not be puromycin resistant. 1. Electroporate cells with transposase or recombinase expression plasmid as above. 2. Sixteen hours postelectroporation, apply ES-cell medium with puromycin at 1 g/ml and maintain the selection for 48 h. 3. On day 4, harvest the cells by trypsinization, count and replate at a density of 1000 cells per 90-mm plate. 4. Pick colonies 8–10 days later.

238

Meng Amy Li et al.

5. Analyze representative clones by a triple-primer competition PCR method, described in Section 4.2 to check that genotype reversion has occurred. 6. Subject cells to phenotypic analysis. Note that revertants are likely to yield heterozygote reversions in mutants obtained from recessive screens, with the transposon remobilized from just one allele. The heterozygote-reversion efficiency with recombinase Cre or FlpO (Raymond and Soriano, 2007) is around 10–30%, and 5–10% with piggyBac transposase. Heterozygote revertants will not become FIAU resistant and can thus only be identified by single cell cloning and molecular analysis.

5. Concluding Remarks In this chapter, we have outlined the principles, basic tools, and protocols to conduct genome-wide genetic screens in mouse ES cells. The specific design of phenotypic screens is beyond the scope of this chapter. However, sensitivity and specificity of the selection system deployed for the screen is a major factor determining its success. In addition, functional analysis of the mutants is important to elucidate the roles that the causal gene plays in the biological pathway investigated. The methods described here are widely adaptable to other cellular models beyond mouse ES cells.

Appendix Biological materials described in this chapter, which are available upon request. piggyBac transposon-containing plasmids (Wang et al., 2008a):
PBGTV1: piggyBac transposon-containing SA-bgeo-based gene-trap cassette flanked by FRT sites (Fig. 12.2A).
PBGTV2 series: piggyBac transposon-containing a loxP flanked genetrap cassette. The gene-trap cassette consists of an En2 splice acceptor (En2SA) T2A-b-gal-T2A-Neo (Fig. 12.2D). piggyBac transposase-containing plasmids (Cadinanos and Bradley, 2007):
Synthetic version of iPBase: wild-type insect piggyBac transposase.
mPBase: mammalian codon optimized piggyBac transposase. Blm-deficient ES cells:
NN5 cell line: Blm-knockout ES cells (Guo, Wang and Bradley, 2004).
Blmtet/tet ES cell line: conditional Blm ES cells (Yusa et al., 2004) (Fig. 12.5).

239

Genome-Wide Forward Genetic Screens in Mouse ES Cells

X

X

X

X

X

X

Genomic DNA digested with restriction enzyme X Fragment with transposon P HO

OH P

Other fragments P HO

OH P

Adaptor ligation HO

First cycle

Extension from transposon end No extension from linker end

No priming No amplification

Second cycle Amplification from both ends Further cycles Nested PCR

Sequencing

Figure 12.5 The Splinkerette-PCR method. A fragment of genomic DNA containing a piggyBac inverted terminal repeat (black double arrow) is shown. After digestion (cleavage sites marked with X), splinkerette adaptors (shown in gray) are ligated to all fragments. If the fragment contains the transposon end (left), the transposon-specific primer (black arrow) can hybridize and extend in the first round. Extension into the long strand of the adaptor provides the template for the adaptor primer to anneal and extend. In the subsequent cycles of PCR, the transposon-specific and adaptor primers can amplify the transposon–genomic junction. A nested PCR is used to improve specificity.

ACKNOWLEDGMENTS This work is supported by the Wellcome Trust. M. A. L. and S. J. P. are graduate students funded by Wellcome Trust Sanger Institute. K. Y. is funded by postdoctoral fellowship from the Japan Society for the Promotion of Science. We thank George Vassiliou and Roland Rad for useful discussions during the preparation of this manuscript.

240

Meng Amy Li et al.

REFERENCES Bradley, A., Evans, M., Kaufman, M. H., and Robertson, E. (1984). Formation of germ-line chimaeras from embryo-derived teratocarcinoma cell lines. Nature 309, 255–256. Cadinanos, J., and Bradley, A. (2007). Generation of an inducible and optimized piggyBac transposon system. Nucleic Acids Res. 35, e87. Cary, L. C., Goebel, M., Corsaro, B. G., Wang, H. G., Rosen, E., and Fraser, M. J. (1989). Transposon mutagenesis of baculoviruses: Analysis of Trichoplusia ni transposon IFP2 insertions within the FP-locus of nuclear polyhedrosis viruses. Virology 172, 156–169. Chambers, I., Colby, D., Robertson, M., Nichols, J., Lee, S., Tweedie, S., and Smith, A. (2003). Functional expression cloning of Nanog, a pluripotency sustaining factor in embryonic stem cells. Cell 113, 643–655. Chen, Y., Yee, D., Dains, K., Chatterjee, A., Cavalcoli, J., Schneider, E., Om, J., Woychik, R. P., and Magnuson, T. (2000). Genotype-based screen for ENU-induced mutations in mouse embryonic stem cells. Nat. Genet. 24, 314–317. Chu, E. H. (1971). Mammalian cell genetics. 3. Characterization of X-ray-induced forward mutations in Chinese hamster cell cultures. Mutat. Res. 11, 23–34. Collier, L. S., Carlson, C. M., Ravimohan, S., Dupuy, A. J., and Largaespada, D. A. (2005). Cancer gene discovery in solid tumours using transposon-based somatic mutagenesis in the mouse. Nature 436, 272–276. Devon, R. S., Porteous, D. J., and Brookes, A. J. (1995). Splinkerettes—Improved vectorettes for greater efficiency in PCR walking. Nucleic Acids Res. 23, 1644–1645. Ding, S., Wu, X., Li, G., Han, M., Zhuang, Y., and Xu, T. (2005). Efficient transposition of the piggyBac (PB) transposon in mammalian cells and mice. Cell 122, 473–483. Evans, M. J., and Kaufman, M. H. (1981). Establishment in culture of pluripotential cells from mouse embryos. Nature 292, 154–156. Guo, G., Wang, W., and Bradley, A. (2004). Mismatch repair genes identified using genetic screens in Blm-deficient embryonic stem cells. Nature 429, 891–895. Hansen, G. M., Markesich, D. C., Burnett, M. B., Zhu, Q., Dionne, K. M., Richter, L. J., Finnell, R. H., Sands, A. T., Zambrowicz, B. P., and Abuin, A. (2008). Large-scale gene trapping in C57BL/6N mouse embryonic stem cells. Genome Res. 18, 1670–1679. Ivics, Z., Hackett, P. B., Plasterk, R. H., and Izsva´k, Z. (1997). Molecular reconstruction of sleeping beauty, a Tc1-like transposon from fish, and its transposition in human cells. Cell 91, 501–510. Ivics, Z., Li, M. A., Mates, L., Boeke, J. D., Nagy, A., Bradley, A., and Izsvak, Z. (2009). Transposon-mediated genome manipulation in vertebrates. Nat. Methods 6, 415–422. Koller, B. H., and Smithies, O. (1989). Inactivating the beta 2-microglobulin locus in mouse embryonic stem cells by homologous recombination. Proc. Natl. Acad. Sci. USA 86, 8932–8935. Kong, J., Zhu, F., Stalker, J., and Adams, D. J. (2008). iMapper: A web application for the automated analysis and mapping of insertional mutagenesis sequence data against Ensembl genomes. Bioinformatics 24, 2923–2925. Lefebvre, L., Dionne, N., Karaskova, J., Squire, J. A., and Nagy, A. (2001). Selection for transgene homozygosity in embryonic stem cells results in extensive loss of heterozygosity. Nat. Genet. 27, 257–258. Liang, Q., Kong, J., Stalker, J., and Bradley, A. (2009). Chromosomal mobilization and reintegration of Sleeping Beauty and piggyBac transposons. Genesis 47, 404–408. McAleer, M. A., Coffey, A. J., and Dunham, I. (2003). DNA rescue by the vectorette method. Methods Mol. Biol. 226, 393–400. Mortensen, R. M., Conner, D. A., Chao, S., Geisterfer-Lowrance, A. A., and Seidman, J. G. (1992). Production of homozygous mutant ES cells with a single targeting construct. Mol. Cell. Biol. 12, 2391–2395.

Genome-Wide Forward Genetic Screens in Mouse ES Cells

241

Munroe, R. J., Bergstrom, R. A., Zheng, Q. Y., Libby, B., Smith, R., John, S. W. M., Schimenti, K. J., Browning, V. L., and Schimenti, J. C. (2000). Mouse mutants from chemically mutagenized embryonic stem cells. Nat. Genet. 24, 318–321. Ochman, H., Gerber, A. S., and Hartl, D. L. (1988). Genetic applications of an inverse polymerase chain reaction. Genetics 120, 621–623. Palmenberg, A. C., Parks, G. D., Hall, D. J., Ingraham, R. H., Seng, T. W., and Pallai, P. V. (1992). Proteolytic processing of the cardioviral P2 region: Primary 2A/2B cleavage in clone-derived precursors. Virology 190, 754–762. Ramirez-Solis, R., Liu, P., and Bradley, A. (1995). Chromosome engineering in mice. Nature 378, 720–724. Raymond, C. S., and Soriano, P. (2007). High-efficiency FLP and PhiC31 site-specific recombination in mammalian cells. PLoS One 2, e162. Ryan, M. D., King, A. M., and Thomas, G. P. (1991). Cleavage of foot-and-mouth disease virus polyprotein is mediated by residues located within a 19 amino acid sequence. J. Gen. Virol. 72(Pt. 11), 2727–2732. Schimenti, J. C., Libby, B. J., Bergstrom, R. A., Wilson, L. A., Naf, D., Tarantino, L. M., Alavizadeh, A., Lengeling, A., and Bucan, M. (2000). Interdigitated deletion complexes on mouse chromosome 5 induced by irradiation of embryonic stem cells. Genome Res. 10, 1043–1050. Schwartzberg, P. L., Goff, S. P., and Robertson, E. J. (1989). Germ-line transmission of a c-abl mutation produced by targeted gene disruption in ES cells. Science 246, 799–803. Snouwaert, J. N., Brigman, K. K., Latour, A. M., Malouf, N. N., Boucher, R. C., Smithies, O., and Koller, B. H. (1992). An animal model for cystic fibrosis made by gene targeting. Science 257, 1083–1088. Sung, P., and Klein, H. (2006). Mechanism of homologous recombination: Mediators and helicases take on regulatory functions. Nat. Rev. Mol. Cell Biol. 7, 739–750. Takahashi, K., and Yamanaka, S. (2006). Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663–676. Taniguchi, M., Sanbo, M., Watanabe, S., Naruse, I., Mishina, M., and Yagi, T. (1998). Efficient production of Cre-mediated site-directed recombinants through the utilization of the puromycin resistance gene, pac: A transient gene-integration marker for ES cells. Nucleic Acids Res. 26, 679–680. Thomas, K. R., and Capecchi, M. R. (1987). Site-directed mutagenesis by gene targeting in mouse embryo-derived stem cells. Cell 51, 503–512. Trombly, M. I., Su, H., and Wang, X. (2009). A genetic screen for components of the mammalian RNA interference pathway in Bloom-deficient mouse embryonic stem cells. Nucleic Acids Res. 37, e34. Uren, A. G., Mikkers, H., Kool, J., van der Weyden, L., Lund, A. H., Wilson, C. H., Rance, R., Jonkers, J., van Lohuizen, M., Berns, A., and Adams, D. J. (2009). A highthroughput splinkerette-PCR method for the isolation and sequencing of retroviral insertion sites. Nat. Protoc. 4, 789–798. Urlaub, G., Mitchell, P. J., Kas, E., Chasin, L. A., Funanage, V. L., Myoda, T. T., and Hamlin, J. (1986). Effect of gamma rays at the dihydrofolate reductase locus: Deletions and inversions. Somat. Cell Mol. Genet. 12, 555–566. Vierbuchen, T., Ostermeier, A., Pang, Z. P., Kokubu, Y., Sudhof, T. C., and Wernig, M. (2010). Direct conversion of fibroblasts to functional neurons by defined factors. Nature 463, 1035–1041. Wang, W., and Bradley, A. (2007). A recessive genetic screen for host factors required for retroviral infection in a library of insertionally mutated Blm-deficient embryonic stem cells. Genome Biol. 8, R48. Wang, W., Bradley, A., and Huang, Y. (2008a). A piggyBac transposon-based genome wide library of insertionally mutated Blm deficient murine ES cells. Genome Res. 19, 667–673.

242

Meng Amy Li et al.

Wang, W., Lin, C., Lu, D., Ning, Z., Cox, T., Melvin, D., Wang, X., Bradley, A., and Liu, P. (2008b). Chromosomal transposition of PiggyBac in mouse embryonic stem cells. Proc. Natl. Acad. Sci. USA 105, 9290–9295. Wu, S. C., Meir, Y. J., Coates, C. J., Handler, A. M., Pelczar, P., Moisyadi, S., and Kaminski, J. M. (2006). piggyBac is a flexible and highly active transposon as compared to sleeping beauty, Tol2, and Mos1 in mammalian cells. Proc. Natl. Acad. Sci. USA 103, 15008–15013. You, Y., Bergstrom, R., Klemm, M., Lederman, B., Nelson, H., Ticknor, C., Jaenisch, R., and Schimenti, J. (1997). Chromosomal deletion complexes in mice by radiation of embryonic stem cells. Nat. Genet. 15, 285–288. Yusa, K., Horie, K., Kondoh, G., Kouno, M., Maeda, Y., Kinoshita, T., and Takeda, J. (2004). Genome-wide phenotype analysis in ES cells by regulated disruption of Bloom’s syndrome gene. Nature 429, 896–899. Yusa, K., Rad, R., Takeda, J., and Bradley, A. (2009). Generation of transgene-free induced pluripotent mouse stem cells by the piggyBac transposon. Nat. Methods 6, 363–369. Zijlstra, M., Li, E., Sajjadi, F., Subramani, S., and Jaenisch, R. (1989). Germ-line transmission of a disrupted beta 2-microglobulin gene produced by homologous recombination in embryonic stem cells. Nature 342, 435–438.

C H A P T E R

T H I R T E E N

Gene Trap Mutagenesis in the Mouse Roland H. Friedel*,† and Philippe Soriano† Contents 244 244 246 250 251 253 253 253 254 255 255 255 256 258 258 258 259 261 264 264 264

1. Introduction 2. Gene Trapping Strategies 2.1. Promoter and polyA trapping strategies 2.2. Trap modifications with site-specific recombinases 2.3. Induction trapping and phenotypic screens in ES cells 3. Design of the Gene Trap Vector 3.1. Splicing elements 3.2. Reporter and selection genes 3.3. Plasmid, retroviral, and transposon vectors 4. Gene Trapping Protocol for Retroviral Vectors 4.1. Generation of retroviral vectors 4.2. Production of virus supernatant 4.3. Infection of ES cells and clone picking 5. Gene Trapping Protocol for Transposon Vectors 5.1. Generation of transposon plasmids 5.2. Electroporation of ES cells with PiggyBac transposon 6. Identification of Trap Insertion Sites by Splinkerette PCR 7. Ordering and Handling of Gene Trap Clones from Consortia 8. Outlook Acknowledgments References

Abstract Gene trapping in mouse embryonic stem (ES) cells is an efficient method for the mutagenesis of the mammalian genome. Insertion of a gene trap vector disrupts gene function, reports gene expression, and provides a convenient tag for the identification of the insertion site. The trap vector can be delivered to ES cells by electroporation of a plasmid, by retroviral infection, or by transposonmediated insertion. Recent developments in trapping technology involve the utilization of site-specific recombination sites, which allow the induced * Department of Neurosurgery, Mount Sinai School of Medicine, New York, USA Department of Developmental and Regenerative Biology, Mount Sinai School of Medicine, New York, USA

{

Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77013-0

#

2010 Elsevier Inc. All rights reserved.

243

244

Roland H. Friedel and Philippe Soriano

modification of trap alleles in vitro and in vivo. Gene trapping strategies have also been successfully developed to screen for genes that are acting in specific biological pathways. In this chapter, we review different applications of gene trapping, and we provide detailed experimental protocols for gene trapping in ES cells by retroviral and transposon gene trap vectors.

1. Introduction The generation of mutations in the mouse is a prerequisite for understanding processes as diverse as development, disease, and cancer, and it is therefore the subject of concerted efforts worldwide to disrupt the function of all mouse genes. Three main strategies have been used to generate mutations: first, gene targeting by homologous recombination is used to produce defined mutations in specific genes; second, point mutations or small deletions can be generated by chemical mutagenesis using agents such as ENU or EMS; and third, random gene trap mutagenesis can be applied genome-wide using plasmid, virus, or transposon DNA vectors. These three approaches create different types of alleles, and they complement each other in the creation of allelic series for all mouse genes. The recent advances in sequencing of mammalian genomes have facilitated the expansion of these approaches to large-scale mutagenesis programs, which aim at generating public resources of mouse mutant alleles (for review, see Gondo, 2008). It will be this collection of allelic variants, with mutations ranging from complete deletion to subtle modifications, which will eventually allow us to dissect gene function at a detailed level. In this chapter, we review different approaches of gene trapping in mouse embryonic stem (ES) cells, and we provide detailed experimental protocols for trapping by retroviral and transposon vectors. The last section contains a practical guide for the ordering and handling of gene trap clones from consortia. Readers are also referred to Chapter 14 that discusses further aspects of gene trap mutagenesis.

2. Gene Trapping Strategies The concept of using insertional mutagenesis to disrupt gene function takes its source from the classic work of Barbara McClintock with transposons in maize. In the mouse, endogenous or exogenous proviruses as well as microinjected DNA can disrupt gene function by insertional mutagenesis (Copeland et al., 1983; Schnieke et al., 1983; Wagner et al., 1983). It was shown that transgene expression can be subject to gene regulatory elements in the vicinity of the insertion site (Jaenisch et al., 1981), a finding

245

Gene Trapping in Mice

subsequently extended to transgenes encoding an easily followed reporter gene, such as lacZ (Korn et al., 1992; Kothary et al., 1988). Based on these observations, the first generation of trapping vectors contained a minimal internal promoter that drives gene expression if the trap insertion has occurred in the vicinity of an endogenous enhancer element, hence the term ‘‘enhancer trap’’ (Fig. 13.1A). The enhancer trapping strategy is mainly attractive for screens that focus on generating lines with specific expression patterns (Gossler et al., 1989; Korn et al., 1992).

A Enhancer trap

B

mP

bgal

pA

E

Promoter trap SA

bgeo

pA

TM

C Secretory trap

SS

SA

bgeo

pA

D Exon trap neo

pA

E PolyA trap SA bgal pA P neo SD

pA

F Conditional trap “Flex” SA

G

bgeo

pA

Conditional trap “double switch” P puro SD

pA

SA IRES eGFP pA

Figure 13.1 Schematic diagrams of different types of gene trap alleles. (A) Insertion of an enhancer trap vector with minimal promoter (mP) nearby a transcriptionally active gene activates expression of the b-galactosidase reporter (bgal). (B) A promoter trap insertion, in which a splice acceptor (SA) fuses an endogenous transcript to a b-galactosidase/neomycin phophotransferase (bgeo) reporter/selection cassette. The fusion transcript of the promoter trap confers drug resistance, reports the expression pattern, and mutates the function of the endogenous trapped gene. Note that promoter trapping can also be performed with vectors that express drug resistance by a strong

246

Roland H. Friedel and Philippe Soriano

However, enhancer trap insertions do not always create functional mutations, and this strategy has not been exploited on a wider scale. With the advent of methods for culturing germ line competent ES cells, the concept of gene trap mutagenesis was expanded to a variety of approaches (Friedrich and Soriano, 1991; Gossler et al., 1989; von Melchner et al., 1992). The insertion of a gene trap vector in ES cells simultaneously disrupts gene function, reports expression pattern of the trapped gene, and provides a convenient tag for the molecular identification of the insertion site. Gene trapping has since been proven to be a highly efficient method to generate large number of mutations, and it has been successfully applied by several large-scale projects to generate libraries of trapped ES cell clones (Table 13.1). In the first years of gene trapping, one of the salient features was its effectiveness for the discovery of novel genes. This aspect of trapping seems less important now that the mouse genome has been sequenced. However, it has been recognized that the mammalian genome may also contain large numbers of yet unidentified nonprotein coding genes, and thus gene trapping may be a very valuable approach for the discovery of these novel genes (Roma et al., 2007).

2.1. Promoter and polyA trapping strategies Two main strategies were developed for the mutagenic trapping of mouse genes: ‘‘promoter trapping’’ and ‘‘polyA trapping.’’ The currently more widely applied strategy is promoter trapping, in which the trap vector heterologous promoter (see text for details). (C) The secretory trap vector is designed to trap genes that encode secretory signal peptide (SS). Secretion of the fusion protein from the endogenous gene and the trap vector is prevented by a transmembrane domain (TM) encoded in the trap vector. The TM anchors the fusion protein at the intracellular side of the cell membrane, thereby conferring drug resistance to trapped cells. (D) The insertion of an exon trap vector into an exon of a transcriptionally active gene leads to a direct fusion of a selection gene (neomycin phosphotransferase, neo) to the endogenous gene. (E) A polyA trap vector has two components: an upstream element with 50 splice acceptor (SA-bgal) to report gene expression, and a downstream element with promoter, selection gene, and splice donor (P-neo-SD), to select for drug-resistant intergenic insertions. (F) The conditional ‘‘Flex’’ design contains a promoter trap cassette that is flanked by upstream and downstream arrays of heterologous recombination sites for the Flp (semicircles) and Cre recombinases (triangles); for details see Schn€ utgen et al. (2005). Successive application of these recombinases leads to inversion (conditional allele) and reinversion (reinstated mutagenic allele) of the promoter trap cassette. (G) The conditional double switch vector contains an upstream 50 splice acceptor-GFP reporter component (SA-IRESeGFP-pA), which is in inverse orientation to gene transcription and flanked by half-mutant lox66 and lox71 sites (triangles). A downstream polyA trap component (P-puro-SD) is flanked by FRT recombination sites (semicircles). Excision of the polyA trap component by Flp recombinase creates the conditional allele, and subsequent inversion of the reporter element generates the mutated allele (see Xin et al., 2005 for details).

Table 13.1 Gene trap resources Program

Web site

Reference

International Gene Trap Consortium (IGTC)

www.genetrap.org

German Gene Trap Consortium (GGTC) Sanger Institute Gene Trap Resource (SIGTR) Centre for Modeling Human Disease (CMHD) TIGEM-IRBM Gene TRAP Project Texas A&M Institute of Genomic Medicine (TIGM) Database for exchangeable gene trap clones Soriano lab Centre for Mammalian Functional Genomics Baygenomics European Conditional Mouse Mutagenesis (EUCOMM) Unitrap gene trap browser Knockout Mouse Project (KOMP) Mouse Genome Informatics Mutant Mouse Regional Resource Centers (MMRRC) European Mouse Mutant Cell Repository (EuMMCR)

www.genetrap.de www.sanger.ac.uk/genetrap/ www.cmhd.ca/genetrap/ genetrap.tigem.it/ www.tigm.org www.egtc.jp www.drbsinai.org/pc/soriano.html www.escells.ca baygenomics.ucsf.edu www.eucomm.org unitrap.cbm.fvg.it www.knockoutmouse.org www.informatics.jax.org www.mmrrc.org www.eummcr.org

Nord et al. (2006), Skarnes et al. (2004) Hansen et al. (2003) To et al. (2004) Hansen et al. (2008) Taniwaki et al. (2005) Chen et al. (2004a) Hicks et al. (1997) Stryke et al. (2003) Friedel et al. (2007) Roma et al. (2008) Collins et al. (2007) Bult et al. (2009)

248

Roland H. Friedel and Philippe Soriano

contains a 50 splice acceptor that splices to the upstream exon of the trapped gene (Friedrich and Soriano, 1991; Gossler et al., 1989; Skarnes et al., 1992; Tate et al., 1998). In this way, the endogenous promoter of the trapped gene is exploited to drive the expression of the reporter gene (Fig. 13.1B). A promoter trap design that has been proven efficient to enrich for insertions into genes is the translational fusion of the reporter b-galactosidase to the selection gene neomycin phosphotransferase (neo; the fusion is termed bgeo; Friedrich and Soriano, 1991). Since this design involves expression of the selection gene from an endogenous promoter, insertions into transcriptionally silent regions are eliminated during selection. However, it should be noted that promoter trapping can also be performed with a vector containing a separate selection marker that is expressed by an exogenous promoter (e.g., bactin-neo-polyA or PGK-neo-polyA; Friedrich and Soriano, 1991; Gossler et al., 1989). This alternative strategy is useful for screens designed to achieve a more unbiased distribution of insertions throughout the genome. About 90–95% of insertions of a promoter trap with a PGK-neo selection gene are estimated to be in loci that are not expressed in ES cells (Friedrich and Soriano, 1991). A recent large-scale gene trap effort with a PGK-neo-polyA selection gene has demonstrated the feasibility of this approach by generating a trap library that has a mutational coverage of more than 90% of all mouse genes (Gragerov et al., 2007). As this strategy does not enrich for gene insertions, traps were not characterized by insertion site cloning of isolated clones, instead, traps in genes of interest can be identified by PCR screening of ES cell pools, and individual ES cell clones are subsequently obtained by successive dilutions of the pools (Gragerov et al., 2007). The reporter of a promoter trap can be translated independently from the trapped gene if it includes its own ATG translation start codon. This is a useful property for the detection of cells that express a trapped gene, since the reporter protein can diffuse throughout the cell body. An alternative strategy involves a reporter without ATG, which results in chimeric fusion proteins containing N-terminal parts of the endogenous protein and the reporter protein. The fusion proteins can be directed to a variety of different subcellular compartments, and this feature can be exploited to screen for genes encoding compartment-specific proteins, as has been successfully shown in a promoter trap screen for nuclear proteins (Tate et al., 1998). A promoter trap approach that is specifically designed to enrich for genes encoding secreted and transmembrane proteins is the ‘‘secretory trap’’ (Fig. 13.1C; De-Zolt et al., 2006; Leighton et al., 2001; Mitchell et al., 2001; Skarnes et al., 1995). This class of genes is underrepresented in standard promoter traps, because such trap fusions with secretory proteins contain a signal sequence and are secreted from cells. A secretory trap vector encodes a splice acceptor that is followed by a transmembrane domain and reporter/selection gene(s). This design enriches for secretory genes, since

Gene Trapping in Mice

249

only a trap vector insertion downstream of a signal sequence (or a type II transmembrane domain) creates a topology that allows for efficient drug selection: the vector-borne transmembrane domain anchors the fusion protein in the membrane and places the reporter/selection marker in the cytosol, where it has normal enzymatic activity. In contrast, trap insertions in nonsecretory genes are mostly nonproductive, since the absence of a signal sequence leads to an insertion of the transmembrane domain in type II orientation, which places the reporter/selection element in the extracellular space or the ER/lysosomal lumen. A further variant of promoter traps is the ‘‘exon trap,’’ which lacks a splice acceptor and needs to insert directly into an exon to form a functional trap (Fig. 13.1D; Hicks et al., 1997; von Melchner et al., 1992). The mutational target of exon trap vectors is much less than that of promoter traps, since exons comprise a much smaller fraction of the genome than introns. For this reason, the exon trap vector is less efficient in highthroughput screens, and it has not been used so widely. However, only this type of vector can trap genes that do not possess introns. As promoter traps that utilize the splice acceptor-bgeo (SA-bgeo) design depend on expression by the endogenous promoter to build up drug resistance, some genes are not accessible to this strategy simply because they are not expressed at high enough levels in mouse ES cells. Based on the results of large-scale promoter trap screens with SA-bgeo vectors, it is currently estimated that about 60% of mouse genes can be mutated with this vector type in ES cells (Skarnes et al., 2004). A design that aims at increasing the pool of genes accessible to promoter traps is to incorporate binding sites for transcription factors that are specifically active in mouse ES cells. Promising results have been obtained with a promoter trap vector that contains binding sites for the transcription factor Oct4, which is thought to enhance expression of trapped genes by recruiting Oct4 to the trapped locus (Schnu¨tgen et al., 2008). The binding sites can be removed at a later stage from the trapped allele by excision through the Flp recombinase (Schnu¨tgen et al., 2008). The ‘‘polyA trap’’ strategy was devised for the mutation of genes independent of their expression level in mouse ES cells. The method relies on using a vector with a strong internal promoter that drives expression of a resistance marker lacking its own polyA signal, followed by a splice donor (Fig. 13.1E; Niwa et al., 1993; Salminen et al., 1998; Yoshida et al., 1995; Zambrowicz et al., 1998). The splice donor design serves to enrich for insertions into genes, since polyA trap insertions outside of genes cannot form polyadenylated transcripts, leading to transcript degradation and elimination of ES cells during drug selection. Gene insertions of the polyA trap vector form a stable transcript by splicing the resistance gene to an endogenous polyA signal. A 50 splice acceptor with reporter gene is usually included in polyA trap vectors to monitor the expression of the trapped gene.

250

Roland H. Friedel and Philippe Soriano

Although the polyA trap design should allow in theory to trap all genes regardless of expression levels, the results of polyA trapping screens have often been below expectations, since insertions have been mainly found in 30 introns (Osipovich et al., 2005; Zambrowicz et al., 2003). One reason for this shortcoming could be the mechanism of nonsense-mediated decay (NMD), which degrades transcripts that contain a premature stop-codon upstream of a splice junction (Shigeoka et al., 2005). However, the NMD mechanism may not affect every type of polyA trap vector. In a gene trap screen for growth factorinduced genes, a polyA trap component of the vector was used to generate 30 RACE products for more than 2000 clones, indicating stable transcript expression from this vector (Chen et al., 2004a). A recent study also suggests that an optimized combination of promoter and splice donor elements may overcome the problems of polyA trapping (Tsakiridis et al., 2009). The development of new polyA trap vectors is an active field of research, and these attempts include the insertion of mRNA destabilizing elements after the splice donor to more efficiently eliminate unspliced transcripts (Ishida and Leder, 1999; Tsakiridis et al., 2009), or of an internal ribosomal entry site (IRES) sequence to protect transcripts against NMD (Shigeoka et al., 2005). However, these strategies have yet to be validated in a large-scale gene trap setting.

2.2. Trap modifications with site-specific recombinases The Cre, Flp, and FC31 recombinases mediate site-specific recombination at lox, FRT, and attB/P recognition sites, respectively, and allow the directed modification of the initial trap allele (see Schnu¨tgen et al., 2006 for review). Codon optimized versions of these enzymes have been engineered for efficient translation in mammalian cells (Buchholz et al., 1998; Raymond and Soriano, 2007). Recently, the Dre recombinase, which recognizes rox sites, has also become available for mammalian genome engineering (Anastassiadis et al., 2009). A trap vector can be flanked with loxP recombination sites in tandem orientation, which enables Cre-mediated excision of the trap vector in rescue experiments (Ishida and Leder, 1999). A reversion of a phenotype after trap vector excision indicates that the mutation was specifically caused by the insertion of the trap vector. An important application of recombinases in gene trapping is the conditional trap approach, which allows the time and tissuespecific ablation of gene function (Schnu¨tgen et al., 2005; Xin et al., 2005). The conditional designs are based on trap cassettes that are flanked by recombination sites in opposing orientations. This type of site arrangement causes inversion of the cassette upon recombination, rather than excision. The ‘‘Flex’’ system utilizes two arrays of loxP/lox5171 and FRT/F3 sites that flank the trap cassette in opposing orientation (Fig. 13.1F; Schnu¨tgen et al., 2005). The initial trap allele of a Flex vector is a mutagenic promoter trap insertion. By applying Flp recombinase, for example, by

Gene Trapping in Mice

251

breeding to an Flp germ line deleter mouse, the trap cassette can be inverted—thereby creating a neutral conditional allele, since the splice acceptor has been removed from the direction of transcription. The conditional trap can be selectively reactivated in a time- and tissue-specific manner by applying Cre recombinase, which reinverts the trap cassette back to its original mutagenic orientation. The ‘‘double switch’’ trap vector consists of an upstream 50 splice acceptor-GFP (green fluorescent protein) reporter component, which is in inverse orientation to gene transcription and flanked by half-mutant lox66 and lox71 sites in opposing orientation, and a downstream 30 polyA trap component flanked by FRT sites (Fig. 13.1G; Xin et al., 2005). The polyA trap component can be excised by Flp recombinase, which creates a conditional allele. This allele can be rendered mutagenic in a second step by Cre-mediated inversion of the splice acceptor-GFP component. The conditional version of a trap insertion may in some instances not be completely innocuous. For instance, it is possible that the insertion interferes with a cis-acting regulatory element, or the inverted cassette may disrupt another gene in antisense orientation to the initially trapped gene. These scenarios should be ascertained by phenotypic comparison of conditional trap mice with wild-type and null mutant mice. Another useful application of site-specific recombination sites is the exchange of cassettes in existing trap lines (‘‘recombinase-mediated cassette exchange,’’ RMCE). In this way, the initial trap allele can be modified in any conceivable way, for example, by inserting cDNAs encoding a fluorescent protein reporter, a Cre recombinase, a mutant version of the trapped gene, or an shRNA construct. The RMCE design is based on a cassette flanked by tandem recombination sites, which mediate the exchange of the residing cassette with an incoming vector. Vectors suitable for RMCE have been reported with flanking loxP sites (Hardouin and Nagy, 2000), halfmutant lox66/71 sites (Araki et al., 1999; Singla et al., 2010; Taniwaki et al., 2005), or heterotypic lox sites (Osipovich et al., 2005). Similar RMCE designs are also feasible with vectors that harbor heterotypic FRT sites or attB/P recombination sites (Baer and Bode, 2001).

2.3. Induction trapping and phenotypic screens in ES cells A mutagenesis project typically aims at identifying genes that are involved in a specific biological process. Induction trapping is a strategy for the enrichment of trap events for genes that are involved in such processes. For example, trap clones can be subjected to treatment with the differentiation factor retinoic acid (RA) in vitro, and the readout of the lacZ reporter can be used to select for genes that are sensitive to regulation by RA (Forrester et al., 1996). Conceptually, similar approaches have involved induction of clones with growth factors (Bonaldo et al., 1998), a secreted

252

Roland H. Friedel and Philippe Soriano

Engrailed-2 homeodomain (Mainguy et al., 2000), or the differentiation of trapped clones to endothelial cells (Hirashima et al., 2004). ES cells can also be differentiated via embryoid bodies to different cell lineages of all three germ layers (Stuhlmann, 2003). Bradley and colleagues devised an induction trapping strategy that allows trapping of genes only transiently expressed upon induction (Chen et al., 2004b). This strategy utilizes a Cre recombinase as intermediate readout to activate a permanent resistance marker from a responder allele. The use of a reporter allowing both positive and negative selection coupled with flow sorting has also been used successfully to identify hepatocyte growth factor (HGF)-responsive genes (Medico et al., 2001). A potential limitation of applying induction trapping strategies in ES cells may be that these cells are not appropriate responders for most exogenous stimuli. It may be biologically more appropriate to induce target genes in specialized cell lines or primary tissue, and to identify genes regulated by a specific stimulus by DNA microarray hybridization or massive parallel sequencing. Mouse mutants can then be conveniently generated by resorting to available gene trap resources. In two studies that performed gene trap screens in combination with DNA arrays, trap vectors contained promoter and splice donor to generate a trap transcript that can be used for the amplification of 30 RACE fragments. DNA arrays were generated from the 30 RACE fragments, thus representing the collection of gene trap clones. Arrays were then hybridized with cDNA probes that were specific for certain organs (Matsuda et al., 2004), or with probes derived from cells that were stimulated with PDGF (Chen et al., 2004a). Another approach to direct gene trapping toward genes in a specific pathway is to perform a phenotypic screen in ES cells. However, most gene trap insertions will cause heterozygous mutations (except for insertions in sex chromosomes), which will rarely generate detectable phenotypes. A strategy to obtain homozygous mutations is the application of ES cells that show high levels of somatic recombination due to a deficiency in the DNA helicase gene Bloom, which is mutated in human Bloom syndrome (Guo et al., 2004). Somatic recombination is a process of crossing-over of sister chromosomes during mitosis, which leads to daughter cells that are homozygote for genes distal to the cross-over site. Bloom/ ES cells with a heterozygous trap insertion will contain cells that are homozygous for the trapped allele after about 13 rounds of cell division. Phenotypic screening in Bloom/ ES cells has been applied to identify genes in the DNA mismatch repair machinery (Guo et al., 2004; Wang et al., 2009), host factors for retroviral infection (Wang and Bradley, 2007), and genes of the RNAi machinery (Trombly et al., 2009). Another strategy for homozygosing trap mutations in ES cells is the use of a gene trap vector with a hypomorphic version of the neo resistance gene (Lin et al., 2006). By simply increasing the drug concentration during neo selection, cells are selected that have doubled the gene dosage of the resistance marker by spontaneous loss-of-heterozygosity at the trapped locus.

Gene Trapping in Mice

253

3. Design of the Gene Trap Vector 3.1. Splicing elements A critical point in design of the gene trap cassette is the choice of splice acceptors and donors. For 50 splicing in promoter trapping, we recommend the adenoviral major late gene exon 2 splice acceptor (Friedrich and Soriano, 1991). We have not observed by-passing of this splice acceptor, which could potentially result in nonmutagenic or hypomorphic trap insertions, in any of 24 mutant mouse lines that we have analyzed at a molecular level. The adenoviral major late gene exon 2 splice donor is suitable as 30 splice donor for polyA trapping approaches, as this splice donor was successfully used for the cloning of 30 RACE transcripts from the ROSAFARY vector (Chen et al., 2004a). The splicing elements discussed above can be obtained from parts of pROSAFARY plasmid from the Addgene plasmid library (www.addgene.org).

3.2. Reporter and selection genes The most commonly used reporter is the b-galactosidase enzyme (bgal), which can be used to visualize gene expression by a color reaction of fixed embryos or tissue sections. For promoter trapping, the bgal enzyme is often used as a direct fusion to the neomycin phosphotransferase gene (bgeo; Friedrich and Soriano, 1991). A marker that allows labeling of cell membranes (e.g., axon tracts) is the placental alkaline phosphatase (PLAP; Leighton et al., 2001). For live imaging of gene expression, fluorescent proteins previously used in gene trap vectors include GFP (Ishida and Leder, 1999), Venus (Tanaka et al., 2008), and fusions of GFP to neo (Chen and Chen, 2004) or bgeo (Tsakiridis et al., 2007). An essential component of a gene trap vector is the resistance marker, typically a neomycin phosphotransferase (neo) gene, which allows for selection of clones that harbor the gene trap insertion. It should be noted that some vectors contain a hypomorphic variant of the neo that is less sensitive to drug selection (Skarnes et al., 1995). Alternative resistance markers for traps in ES cells are puromycin N-acetyl transferase and hygromycin phosphotransferase. A strategy that allows coexpression of multiple reporters (or resistance genes) under the control of the same promoter is to place an IRES between the open reading frames. Drawbacks of this strategy are that the gene following the IRES sequence is often translated at reduced levels and that the efficiency of translation may vary from cell type to cell type (Merrick, 2004). Another approach to ensure equal amounts of translation products is the use of viral 2A self-cleaving peptides (Szymczak et al., 2004). Genes that

254

Roland H. Friedel and Philippe Soriano

have been connected by 2A linker sequences are transcribed as one continuous open reading frame, and during translation one peptide bond in the 2A peptide is omitted by a ribosomal skip mechanism. The 2A peptides are 18 aa long, and leave a 17 aa sequence at the C-terminus of the upstream protein, and one proline residue at the N-terminus of the downstream protein. This strategy is therefore only effective as long as the final protein products remain functional with these additional amino acids.

3.3. Plasmid, retroviral, and transposon vectors Gene trap vectors can be delivered into the ES cell genome as plasmids, retroviruses, or transposons. Each of these methods has unique advantages and limitations. Plasmids are easy to prepare, have no significant size limitations, and can be introduced into ES cells simply by electroporation. The main drawback of plasmid traps is that undefined events can occur at the insertion site, such as deletions, and concatemeric insertions. Such complex insertions complicate the analysis of the mutation, and they hamper the efficient identification of genomic integration sites by linkermediated PCR. Retroviruses require little experimental work, since they are harvested as supernatants from a producer cell line, before application to ES cells. Moreover, they offer the advantage of clean, single insertions, which facilitate the high-throughput cloning of the genomic junctions. Retroviruses preferentially insert into the 50 region of genes, which is also advantageous for disrupting gene function with promoter traps. Gene trapping with transposons can be performed with class II transposons, which move in the genome by a cut-and-paste mechanism. Transposon traps can be easily introduced into ES cells by coelectroporation of a transposon trap plasmid together with a plasmid encoding the appropriate transposase. The main advantage of transposon traps, however, resides in their ability to transpose in somatic tissues at later stages, and they have thus been used very successfully for the identification of genes implicated in cancer (Collier et al., 2005; Dupuy et al., 2005; see also Chapters 4–6 in this volume). Two transposon systems are currently used for mutagenesis in the mouse: Sleeping Beauty, which has been synthetically reconstructed from salmonid fish (Ivics et al., 1997), and PiggyBac, which has been originally identified as transposon in the insect cabbage looper (Cary et al., 1989). Transposon insertions occur at specific target sites, which are characterized by the nucleotide sequences ‘‘TA’’ for Sleeping Beauty and ‘‘TTAA’’ for PiggyBac. The PiggyBac transposon system appears to transpose with higher activity, and it is therefore currently the preferred system for transposon mutagenesis in ES cells (Wang et al., 2008, 2009). Moreover, the lower transposition rate of the Sleeping Beauty transposon leads frequently to insertions in the vicinity of the donor site, a characteristic known as

Gene Trapping in Mice

255

‘‘local hopping.’’ Further advantages of the PiggyBac system are large cargo sizes (Ding et al., 2005), and a clean excision (Sleeping Beauty leaves a ‘‘footprint’’ mutation at the excised locus) (Luo et al., 1998; Wang et al., 2008). In contrast to retroviruses, however, transposons seem to have less preference for the insertion into the 50 end of genes. However, a bias for insertions of transposons into transcribed regions of the genome has been observed (Liang et al., 2009).

4. Gene Trapping Protocol for Retroviral Vectors 4.1. Generation of retroviral vectors A retroviral gene trap vector is generated by cloning of a plasmid that contains a gene trap cassette into two flanking 50 and 30 long terminal repeat (LTR) elements of the Moloney Murine Leukemia Virus (MMLV). The total size of the retroviral vector, including LTRs and trap cassette, should not exceed 11 kb. The direction of the trap cassette should be in reverse to the 50 –30 orientation of the viral LTRs (Friedrich and Soriano, 1991), and the LTRs should be engineered to lack viral enhancers (Soriano et al., 1991) or enhancers and promoters (Chen et al., 2004a). The plasmid for the production of the ROSAFARY gene trap virus (Chen et al., 2004a), which incorporates the design criteria described above, can be ordered from the Addgene plasmid library (www.addgene.org). The plasmid for the conditional rsFLP-ROSA-bgeo gene trap vector (Schnu¨tgen et al., 2005) is available through the EUCOMM program (www.eucomm.org). The latter plasmid contains a ‘‘PINCO’’ backbone with the EBNA1 gene of the Epstein-Barr virus (Grignani et al., 1998), which leads to increased virus titers upon transfection of packaging cell lines.

4.2. Production of virus supernatant Retrovirus particles for gene trapping can be easily obtained by transfecting a retroviral gene trap plasmid into a packaging cell line that carries stable integrations of the gag, pol, and env genes of the MMLV. We have been routinely using the ‘‘GPþE86’’ (Markowitz et al., 1988; Soriano et al., 1991) and the ‘‘Phoenix-Eco’’ cells line (Swift et al., 2001; http://www. stanford.edu/group/nolan) with stable and transient transfection protocols, respectively. For the generation of stably transfected GPþE86 cells, 10 mg linearized vector plasmid are electroporated in 1  106 cells in 0.8 ml phosphatebuffered saline (PBS). After selecting in the presence of the appropriate

256

Roland H. Friedel and Philippe Soriano

antibiotic for 10–12 days, resistant clones are pooled, expanded, and frozen in batch. To collect infectious viral particles, virus-producing cells are grown to near confluence and overlaid with fresh medium. After 16 h, virus-containing medium supernatant is filtered through a 0.45-m filter to remove cells and debris before being used for infection. The supernatant is either applied directly on ES cells (the half-life of retroviruses at 37  C is approximately 6 h), or stored at 80  C. However, freezing and thawing of virus supernatants reduces titers by about half. To determine the effective colony generating viral titer, serial 10-fold dilutions of the viral supernatant are prepared to infect log phase 3T3 cells or ES cells in 6-well dishes. Media should contain 4 g/ml polybrene (hexadimethrin bromide, Sigma) to increase infection efficiency by enhancing virus adsorption on target cell membranes. Cells are selected in antibiotic-containing medium until colonies can be counted by staining with 0.5% methylene blue in 50% methanol. For virus production by transient transfection, Phoenix cells are expanded on 10 cm tissue culture dishes until they reach subconfluency. Ten micrograms of circular plasmid DNA are transfected per dish by the calcium-phosphate method or with a commercial lipid-based transfection reagent. On the day following transfection, overlay the transfected cells with 7 ml of ES cell media, incubate cells overnight and collect the supernatant on the following day. The transfected cells can be conditioned again overnight for a second harvest of virus supernatant.

4.3. Infection of ES cells and clone picking Culture ES cells according to the protocols for the particular cell line. Each ES cell line has its own specific requirements for feeders, media, trypsinization, and split ratios. Grow ES cells to subconfluency in a 10-cm tissue culture dish and infect with virus supernatant in a total volume of 4 ml media in the presence of 4 g/ml polybrene (see Fig. 13.2A for an overview of the workflow). After 6 h, change to normal media without virus. On the following day, start selection with media containing 150 g/ml active concentration of G418 (Invitrogen 11811-031). Change selection media every day for 8–10 days. ES clones should be visible after 7 days, and colonies can be usually picked after 10 days of selection. For picking of clones, wash cells on selection plate once with PBS and cover them again with PBS. Pick clones with a glass capillary or a micropipette in a volume of 10 l by gently scratching them from the plate. Transfer picked clones into a 96-well U-bottom plate containing 30 l of trypsin in each well and incubate for 10 min at 37  C. Add 150 l of medium (þG418) to each well and triturate cells thoroughly to generate a suspension of single cells. Transfer cell suspension to a 96-well flat-bottom plate. Change medium (þG418) on 96-well plates every day. After 2–4 days, when the

257

Gene Trapping in Mice

A

B Transfect cell line and collect virus

Grow ES cells to subconfluency

Grow ES cells to subconfluency

Electroporate transposon and transposase plasmid

Infect with virus for 6 h +

−

7−10 days drug selection

Pick ES cell clones in 96 well plate

Split in replicates for archiving and DNA analysis

Figure 13.2 Workflow of gene trapping by retroviral infection (A) or by electroporation with transposon vectors (B).

majority of wells have reached subconfluence, split cells to four 96-well plates (copies A and B for freezing cells, copies C and D for DNA analysis). Copies A and B are frozen for archiving when the majority of wells have reached subconfluence. For freezing of cells, wash each well with 100 l PBS, trypsinize with 25 l trypsin for 10 min at 37  C, add 75 l media and triturate cells, add 100 l 2 freezing media (20% FCS, 10% DMSO in DMEM) and mix, and cover with 50 l mineral oil. For long-term storage, it is advisable to freeze cells in racks that hold 96 microtubes (e.g., Neptune Minitube System or Thermo Scientific Matrix 2D Barcoded Storage Tubes). Cultivate copies C and D for 2–3 extra days until they appear overgrown in order to get sufficient amounts of cells for DNA analysis.

258

Roland H. Friedel and Philippe Soriano

5. Gene Trapping Protocol for Transposon Vectors 5.1. Generation of transposon plasmids This protocol is described for PiggyBac vectors, but can be easily adapted for the application of the Sleeping Beauty transposon. For gene trapping with a PiggyBac transposon, generate a plasmid that contains a gene trap cassette that is flanked by the 50 and 30 PiggyBac terminal repeats (TR). For efficient transposition of PiggyBac, the minimal sizes for the TRs have been determined to be 310 bp for the 50 TR and 235 bp for the 30 TR (these TRs also contain nonrepeating internal sequences that enhance transposition) (Li et al., 2005). A plasmid containing the minimal PiggyBac repeats can be obtained from Dr. Fraser at the University of Notre Dame (http://piggybac. bio.nd.edu). A PiggyBac transposon can carry up to 9.1 kb of foreign DNA without a significant drop in transposition efficiency (Ding et al., 2005). As transposase, either the original insect transposase (iPBase), or a PiggyBac transposase that has been optimized for mammalian codon usage (mPBase) can be used (Cadinanos and Bradley, 2007). In our case, the activity of original iPBase expressed under the synthetic CAG promoter (Wang et al., 2008) was adequate for gene trapping experiments in ES cells.

5.2. Electroporation of ES cells with PiggyBac transposon The PiggyBac transposon system is introduced into ES cells by coelectroporation of two circular plasmids: a plasmid encoding the transposase and a plasmid containing the PiggyBac transposon gene trap vector. The transposase enzyme will cut out the transposon and paste it randomly at a TTAA target site in the ES cell genome. Depending on the amounts of DNA that are electroporated, multiple independent insertions of PiggyBac transposons can occur in a single clone. To reduce the frequency of clones with multiple insertions, it is advisable to use low amounts of plasmid DNA for electroporation. We typically electroporate 0.4 g PiggyBac transposon plasmid and 2 g iPBase plasmid for 107 cells in a total volume of 800 l PBS. Parameter settings are 250 V, 500 F, and ‘‘exponential decay’’ mode for a Biorad Gene Pulser with a 4-mm cuvette, which should result in a t discharge time of about 0.7 ms. Applying these parameters for a PiggyBac promoter-trap vector with drug-selectable marker, we obtain more than 100 clones per electroporation, of which about 95% harbor single insertions (see also Wang et al., 2008, 2009). For one round of electroporation, expand ES cells in one 10 cm culture dish until they reach subconfluency (see Fig. 13.2B for an overview of the workflow). On the day of electroporation, change media on cells 3 h before

Gene Trapping in Mice

259

electroporation. Trypsinize and count cells in a Neubauer chamber. Spin down 107 cells each and resuspend with PBS to a total volume of 800 l. Add desired amount of plasmid DNA (dissolved in Tris–EDTA, in a volume of no more than 20 l), transfer cell/DNA mix into an electroporation cuvette of 4 mm gap width, and electroporate as described above. The electroporated cells are plated on three 10-cm dishes. Antibiotic selection is applied about 20 h after electroporation by changing media to selective media with 150 g/ml active G418 (Invitrogen 11811-031). ES clones are cultivated in selective media for the following 7–10 days. Perform picking, expansion, and freezing of clones as described in Section 4.3.

6. Identification of Trap Insertion Sites by Splinkerette PCR The insertion sites of gene traps have been previously mainly determined by methods that generate cDNA fragments containing parts of the vector cassette and exons flanking the trap insertion. For promoter traps, intronic locations can be determined by sequencing exons upstream of the trap insertion with the 50 RACE method, and for polyA traps, insertions can be identified by 30 RACE to sequence exons downstream of the insertion. Detailed protocols for 50 and 30 RACE analysis of gene trap clones can be found in Chen and Soriano (2003). However, these methods rely on the analysis of mRNA, which can be a limiting factor for genes that are expressed at low levels in ES cells. In addition, the results give only information about the intron in which the trap vector inserted, but not about the precise genomic location of the insertion, which is a prerequisite for PCR genotyping of gene trap mouse lines from genomic tail DNA. With the advent of the complete sequence of the mouse genome, it became feasible to identify gene trap insertion loci by analysis of the flanking genomic sequences at the insertion site. Different methods for the cloning of junction fragments that contain vector and flanking genomic sequences are available, such as plasmid rescue, inverse PCR, and linker-mediated PCR. A method that has proven to be particularly efficient for this purpose is the ‘‘splinkerette,’’ an optimized protocol of linker-mediated PCR (Devon et al., 1995). It has been successfully applied for the highthroughput identification of gene trap insertions (Horn et al., 2007), and it can be applied for both retroviral or transposon insertions. The main improvement of splinkerette in respect to other linker-mediated strategies is the introduction of a hairpin loop in the linker oligos, which suppresses amplification of unspecific side products (Fig. 13.3). Splinkerette is performed in two separate reactions for the 50 and 30 junctions of the vector. We provide here a protocol for the use of the restriction enzyme BstYI, which recognizes an R0 GATCY site, and was

260

Roland H. Friedel and Philippe Soriano

BstYI

BstYI A

BstYI

Trap vector

B

SPLK

Tra

L1

SPLK

F SPLK

C

Tra V1

SPLK

V1 p vector

SPLK

SPLK SPLK

SPLK

L1 D V1 L2 E V2

Figure 13.3 Identification of gene trap insertion sites by splinkerette PCR. (A) A DNA preparation of an ES cell clone is digested with BstYI restriction endonuclease. (B) Ligation to the splinkerette linker. (C) In the first round of PCR, only the vectorspecific primer V1 can anneal and initiate synthesis of a strand that complements the splinkerette linker. (D) The linker-specific primer L1 and the vector-specific primer V1 PCR amplify a junction fragment of the vector and the adjacent genomic DNA. (E) A PCR with nested primers L2 and V2 serves to improve specificity and yield of the reaction. (F) Generation of unspecific products is minimized, since the primer L1 cannot anneal to linkers that contains the self-complementary ‘‘hook’’ region.

successfully used for a large-scale annotation of a gene trap library (Horn et al., 2007). The reactions can also be repeated with other restriction enzymes, such as Sau3AI, which may increase the recovery rate of junction fragments. The following protocol serves for the cloning of transposon insertion sites. The protocol can also be applied for the cloning of retroviral insertion sites, only the vector-specific primers need to be adjusted (see Horn et al., 2007; Uren et al., 2009). For DNA preparation, ES cells are grown in 96-well plates until they reach confluency. Plates are washed once with PBS and can then be stored dry at  20  C until further processing. Incubate each well with 50 l lysis buffer (10 mM Tris–HCl, pH 7.5, 10 mM EDTA, 10 mM NaCl, 0.5% (w/v) sodium dodecyl sulfate (SDS); add proteinase K to 1 mg/ml fresh before use) for 2 h at 60  C. Add 150 l ice cold ethanol–salt solution (150 l 5 M NaCl to 10 ml of ethanol) per well, let sit at room temperature for 1 h. The precipitated DNA threads stick to the bottom of the well. Wash wells three times with 150 l 70% EtOH. Drain plates by slowly inverting on paper towels. Air dry precipitated DNA and redissolve with 40 l of 5 mM Tris, pH 8.0, overnight at room temperature.

Gene Trapping in Mice

261

For digest, mix 9 l of the genomic DNA with 5 units of BstYI in a total volume of 20 l with buffers supplied by the manufacturer. Incubate reactions in a 96-well plate in a PCR thermal cycler at 60  C for 1 h. Stop digest by heat inactivation at 80  C for 20 min. For alternative restriction enzymes, adjust conditions accordingly. Prepare the double-stranded splinkerette adapter by combining 400 pmol each of the splinkerette primers HMSpAA and HMSpBB (see Table 13.2) in a total volume of 40 l of 5 mM Tris–HCl, pH 7.5, 25 mM NaCl, 0.5 mM DTT. Denature for 3 min at 97  C and allow strands to anneal by slowly cooling to room temperature. Set up ligation in a final volume of 30 l using 20 l of heat inactivated restriction digest, 1 l of the annealed splinkerette adapter mix and 400 units T4-DNA ligase (New England Biolabs) in the provided ligation buffer. Incubate ligation overnight at 16  C. Purify ligation products using a PCR purification kit (QIAGEN) and elute the DNA into 45 l of 5 mM Tris–HCl, pH 8.0. Processing of 96-well plates can be performed with multiwell purification plates. Set up PCR reactions in 0.2-ml thin-walled PCR tubes in a volume of 15 l containing 5 l of the adapter ligated DNA, 1 ThermoPol PCR buffer (NEB), 0.2 mM dNTPs, 0.2 mM of each primer (HMSp1 and PB50 -1 for 50 junction, and HMSp1 and PB30 -1 for 30 junction; Table 13.2), and 0.4 unit Taq Polymerase (NEB). The primary PCR is performed with the following cycling conditions: 2 min 30 s at 95  C for initial denaturation, then 30 cycles of 15 s at 95  C, 30 s at 65  C, and 2 min at 72  C, and a final elongation step at 72  C for 7 min. Use 1 l of the primary amplification reaction without further purification in a nested PCR with the primers HMSp2 and PB50 -2, and HMSp2 and PB30 -2, respectively (Table 13.2), using the same PCR conditions as for the primary PCR. Use 1 l of the final amplification product for direct sequencing with primers PB50 -seq or PB30 -seq (Table 13.2).

7. Ordering and Handling of Gene Trap Clones from Consortia Collections of gene trap clones have been generated by several laboratories worldwide (see Table 13.1 for web links to gene trap consortia and databases). The annotation data of these clones has been combined in a central database, accessible at the web site of the International Gene Trap Consortium (IGTC; www.igtc.org). Other useful databases for accessing gene trap clones can be found at the UniTrap web site (http://unitrap.cbm. fvg.it), and at the Mouse Genome Informatics web site (www.informatics. jax.org), which lists available gene trap clones on its gene information pages.

Table 13.2

Splinkerette primers

Adapter HMSpAA CGAAGAGTAACCGTTGCTAGGAGAGACCGTGGCTGAATGAGACTGGTGTCGACACTAGTGG HMSpBB GATCCCACTAGTGTCGACACCAGTCTCTAATTTTTTTTTTCAAAAAAA First round PCR HMSp1 CGAAGAGTAACCGTTGCTAGGAGAGACC PB50 -1 CAAAATCAGTGACACTTACCGCATTGACAA PB30 -1 TAAATAAACCTCGATATACAGACCGATAAA Second round PCR HMSp2 GTGGCTGAATGAGACTGGTGTCGAC PB50 -2 CTTACCGCATTGACAAGCACGCCTCACGGG PB30 -2 ATATACAGACCGATAAAACACATGCGTCAA Direct sequencing PB50 -seq TTAGAAAGAGAGAGCAATATTTCAAGAATG PB30 -seq TTTTACGCATGATTATCTTTAACGTACGTC HM primers from Mikkers et al. (2002). PB primers from Wang et al. (2009).

Gene Trapping in Mice

263

To find gene trap clones for your gene of interest, enter the gene name in the search query, or blast your sequence against the gene trap sequence tag database. The search result will provide an annotation for each clone, including ordering information, the specific ES cell line that was used, and the insertion site of the clone. The insertion site will be mapped to an intronic location, if 50 /30 RACE was used to identify the clone, or to the exact genomic coordinates, if the clone was identified by genomic PCR. The databases typically list for each gene multiple ES cell clones representing traps in different introns of the gene. For most phenotypic studies, trap clones with insertions in the 50 region of the gene should be chosen, since these are more likely to be functionally mutagenic. The germ line transmission capacity of gene trap ES cell clones is in practice about 50–70%. That is, not every gene trap clone will result in germ line transmission, and it may be necessary in some cases to inject an alternative gene trap clone. Gene trap ES cell clones are shipped in cryo-vials on dry ice, and should be placed immediately in liquid nitrogen storage upon receipt. The sending institution will provide protocols for the handling of the ES cell line and its specific media requirements. Before injection of a gene trap clone into blastocysts for the generation of chimeric mice, the identity of the obtained cell clone should be reconfirmed by RT-PCR. For this purpose, split a fraction of the gene trap cell line during expansion onto one 35 mm dish, let cells grow to confluency, and use cells to prepare RNA and subsequently oligo-dT primed cDNA. Perform the diagnostic RT-PCR reaction with a primer that is located in an exon upstream of the annotated trap insertion (or downstream, in the case of polyA traps), and a matching reverse primer that is specific for the gene trap vector cassette. In few sporadic cases, partial by-passing of the gene trap splice acceptor has been observed, resulting in hypomorphic or nonmutagenic alleles. The full mutagenicity of a gene trap allele should therefore be verified in homozygous mutant animals. The amount of wild-type transcript or protein should be assessed in homozygous mutant and wild-type littermates by Northern or Western blot analysis, respectively. Genotyping of gene trap mouse lines can be easily accomplished by using genomic PCR with a crude DNA preparation of the tail tip. We recommend a three-primer PCR strategy, with one primer specific for the vector, and two primers flanking the insertion site (see Chen and Soriano, 2003 for details). The b-galactosidase (bgal) enzyme of Escherichia coli is frequently included as a reporter in gene trapping vectors to monitor the expression of the trapped gene. Detailed protocols for the application of this reporter in whole-mount embryo and tissue section staining can be found in Chen and Soriano (2003). For gene trap lines with an appropriate expression pattern of the trapped gene, it is also possible to take advantage of the bgal reporter for genotyping of mice by X-gal staining of the tail tip.

264

Roland H. Friedel and Philippe Soriano

8. Outlook Gene trapping has been one of the most successful strategies for the mutagenesis of the mammalian genome, and it is a cornerstone in the worldwide efforts to generate a public library of ES cells with mutations in all mouse genes. Although several consortia have shifted their efforts now on gene targeting, which allows the tailored modification of all genes, gene trapping remains by far the most efficient mutagenesis strategy for the generation of large collections of mutant alleles. Powerful new applications of gene trapping are also emerging in methodologies that aim at modifying gene trap alleles in vivo. The most prominent examples are conditional gene traps that can be activated by recombinases in a time and tissue-specific manner, and the utilization of gene trapping together with somatic transposon mutagenesis in vivo, which has been shown to be a very effective tool for identifying cancer-causing genes. These methods are likely to shed insight into physiological processes well beyond embryonic development, where gene trap mutagenesis has been mainly applied so far. Much progress has been made in the development of gene trap technology over the past decade, and it will be exciting to see its future applications.

ACKNOWLEDGMENTS We thank our colleagues for critical reading of the manuscript. Work in the authors’ laboratory is supported by grants from the National Institute of Child Health and Human Development to P. S.

REFERENCES Anastassiadis, K., Fu, J., Patsch, C., Hu, S., Weidlich, S., Duerschke, K., Buchholz, F., Edenhofer, F., and Stewart, A. F. (2009). Dre recombinase, like Cre, is a highly efficient site-specific recombinase in E. coli, mammalian cells and mice. Dis. Model Mech. 2, 508–515. Araki, K., Imaizumi, T., Sekimoto, T., Yoshinobu, K., Yoshimuta, J., Akizuki, M., Miura, K., Araki, M., and Yamamura, K. (1999). Exchangeable gene trap using the Cre/mutated lox system. Cell Mol. Biol. (Noisy-le-grand) 45, 737–750. Baer, A., and Bode, J. (2001). Coping with kinetic and thermodynamic barriers: RMCE, an efficient strategy for the targeted integration of transgenes. Curr. Opin. Biotechnol. 12, 473–480. Bonaldo, P., Chowdhury, K., Stoykova, A., Torres, M., and Gruss, P. (1998). Efficient gene trap screening for novel developmental genes using IRES beta geo vector and in vitro preselection. Exp. Cell Res. 244, 125–136. Buchholz, F., Angrand, P. O., and Stewart, A. F. (1998). Improved properties of FLP recombinase evolved by cycling mutagenesis. Nat. Biotechnol. 16, 657–662.

Gene Trapping in Mice

265

Bult, C. J., Kadin, J. A., Richardson, J. E., Blake, J. A., and Eppig, J. T. (2009). The mouse genome database: Enhancements and updates. Nucleic Acids Res. 34, D562–D567. Cadinanos, J., and Bradley, A. (2007). Generation of an inducible and optimized piggyBac transposon system. Nucleic Acids Res. 35, e87. Cary, L. C., Goebel, M., Corsaro, B. G., Wang, H. G., Rosen, E., and Fraser, M. J. (1989). Transposon mutagenesis of baculoviruses: Analysis of Trichoplusia ni transposon IFP2 insertions within the FP-locus of nuclear polyhedrosis viruses. Virology 172, 156–169. Chen, W. V., and Chen, Z. (2004). Differentiation trapping screen in live culture for genes expressed in cardiovascular lineages. Dev. Dyn. 229, 319–327. Chen, W. V., and Soriano, P. (2003). Gene trap mutagenesis in embryonic stem cells. Methods Enzymol. 365, 367–386. Chen, W. V., Delrow, J., Corrin, P. D., Frazier, J. P., and Soriano, P. (2004a). Identification and validation of PDGF transcriptional targets by microarray-coupled gene-trap mutagenesis. Nat. Genet. 36, 304–312. Chen, Y. T., Liu, P., and Bradley, A. (2004b). Inducible gene trapping with drug-selectable markers and Cre/loxP to identify developmentally regulated genes. Mol. Cell Biol. 24, 9930–9941. Collier, L. S., Carlson, C. M., Ravimohan, S., Dupuy, A. J., and Largaespada, D. A. (2005). Cancer gene discovery in solid tumours using transposon-based somatic mutagenesis in the mouse. Nature 436, 272–276. Collins, F. S., Rossant, J., and Wurst, W. (2007). A mouse for all reasons. Cell 128, 9–13. Copeland, N. G., Hutchison, K. W., and Jenkins, N. A. (1983). Excision of the DBA ecotropic provirus in dilute coat-color revertants of mice occurs by homologous recombination involving the viral LTRs. Cell 33, 379–387. Devon, R. S., Porteous, D. J., and Brookes, A. J. (1995). Splinkerettes—Improved vectorettes for greater efficiency in PCR walking. Nucleic Acids Res. 23, 1644–1645. De-Zolt, S., Schnutgen, F., Seisenberger, C., Hansen, J., Hollatz, M., Floss, T., Ruiz, P., Wurst, W., and von Melchner, H. (2006). High-throughput trapping of secretory pathway genes in mouse embryonic stem cells. Nucleic Acids Res. 34, e25. Ding, S., Wu, X., Li, G., Han, M., Zhuang, Y., and Xu, T. (2005). Efficient transposition of the piggyBac (PB) transposon in mammalian cells and mice. Cell 122, 473–483. Dupuy, A. J., Akagi, K., Largaespada, D. A., Copeland, N. G., and Jenkins, N. A. (2005). Mammalian mutagenesis using a highly mobile somatic Sleeping Beauty transposon system. Nature 436, 221–226. Forrester, L. M., Nagy, A., Sam, M., Watt, A., Stevenson, L., Bernstein, A., Joyner, A. L., and Wurst, W. (1996). An induction gene trap screen in embryonic stem cells: Identification of genes that respond to retinoic acid in vitro. Proc. Natl. Acad. Sci. USA 93, 1677–1682. Friedel, R. H., Seisenberger, C., Kaloff, C., and Wurst, W. (2007). EUCOMM—The European conditional mouse mutagenesis program. Brief. Funct. Genomic Proteomic 6, 180–185. Friedrich, G., and Soriano, P. (1991). Promoter traps in embryonic stem cells: A genetic screen to identify and mutate developmental genes in mice. Genes Dev. 5, 1513–1523. Gondo, Y. (2008). Trends in large-scale mouse mutagenesis: From genetics to functional genomics. Nat. Rev. Genet. 9, 803–810. Gossler, A., Joyner, A. L., Rossant, J., and Skarnes, W. C. (1989). Mouse embryonic stem cells and reporter constructs to detect developmentally regulated genes. Science 244, 463–465. Gragerov, A., Horie, K., Pavlova, M., Madisen, L., Zeng, H., Gragerova, G., Rhode, A., Dolka, I., Roth, P., Ebbert, A., Moe, S., Navas, C., et al. (2007). Large-scale, saturating insertional mutagenesis of the mouse genome. Proc. Natl. Acad. Sci. USA 104, 14406–14411.

266

Roland H. Friedel and Philippe Soriano

Grignani, F., Kinsella, T., Mencarelli, A., Valtieri, M., Riganelli, D., Grignani, F., Lanfrancone, L., Peschle, C., Nolan, G. P., and Pelicci, P. G. (1998). High-efficiency gene transfer and selection of human hematopoietic progenitor cells with a hybrid EBV/ retroviral vector expressing the green fluorescence protein. Cancer Res. 58, 14–19. Guo, G., Wang, W., and Bradley, A. (2004). Mismatch repair genes identified using genetic screens in Blm-deficient embryonic stem cells. Nature 429, 891–895. Hansen, J., Floss, T., Van Sloun, P., Fuchtbauer, E. M., Vauti, F., Arnold, H. H., Schnutgen, F., Wurst, W., von Melchner, H., and Ruiz, P. (2003). A large-scale, gene-driven mutagenesis approach for the functional analysis of the mouse genome. Proc. Natl. Acad. Sci. USA 100, 9918–9922. Hansen, G. M., Markesich, D. C., Burnett, M. B., Zhu, Q., Dionne, K. M., Richter, L. J., Finnell, R. H., Sands, A. T., Zambrowicz, B. P., and Abuin, A. (2008). Large-scale gene trapping in C57BL/6N mouse embryonic stem cells. Genome Res. 18, 1670–1679. Hardouin, N., and Nagy, A. (2000). Gene-trap-based target site for cre-mediated transgenic insertion. Genesis 26, 245–252. Hicks, G. G., Shi, E. G., Li, X. M., Li, C. H., Pawlak, M., and Ruley, H. E. (1997). Functional genomics in mice by tagged sequence mutagenesis. Nat. Genet. 16, 338–344. Hirashima, M., Bernstein, A., Stanford, W. L., and Rossant, J. (2004). Gene-trap expression screening to identify endothelial-specific genes. Blood 104, 711–718. Horn, C., Hansen, J., Schnutgen, F., Seisenberger, C., Floss, T., Irgang, M., De-Zolt, S., Wurst, W., von Melchner, H., and Noppinger, P. R. (2007). Splinkerette PCR for more efficient characterization of gene trap events. Nat. Genet. 39, 933–934. Ishida, Y., and Leder, P. (1999). RET: A poly A-trap retrovirus vector for reversible disruption and expression monitoring of genes in living cells. Nucleic Acids Res. 27, e35. Ivics, Z., Hackett, P. B., Plasterk, R. H., and Izsvak, Z. (1997). Molecular reconstruction of Sleeping Beauty, a Tc1-like transposon from fish, and its transposition in human cells. Cell 91, 501–510. Jaenisch, R., Jahner, D., Nobis, P., Simon, I., Lohler, J., Harbers, K., and Grotkopp, D. (1981). Chromosomal position and activation of retroviral genomes inserted into the germ line of mice. Cell 24, 519–529. Korn, R., Schoor, M., Neuhaus, H., Henseling, U., Soininen, R., Zachgo, J., and Gossler, A. (1992). Enhancer trap integrations in mouse embryonic stem cells give rise to staining patterns in chimaeric embryos with a high frequency and detect endogenous genes. Mech. Dev. 39, 95–109. Kothary, R., Clapoff, S., Brown, A., Campbell, R., Peterson, A., and Rossant, J. (1988). A transgene containing lacZ inserted into the dystonia locus is expressed in neural tube. Nature 335, 435–437. Leighton, P. A., Mitchell, K. J., Goodrich, L. V., Lu, X., Pinson, K., Scherz, P., Skarnes, W. C., and Tessier-Lavigne, M. (2001). Defining brain wiring patterns and mechanisms through gene trapping in mice. Nature 410, 174–179. Li, X., Harrell, R. A., Handler, A. M., Beam, T., Hennessy, K., and Fraser, M. J., Jr. (2005). piggyBac internal sequences are necessary for efficient transformation of target genomes. Insect Mol. Biol. 14, 17–30. Liang, Q., Kong, J., Stalker, J., and Bradley, A. (2009). Chromosomal mobilization and reintegration of Sleeping Beauty and PiggyBac transposons. Genesis 47, 404–408. Lin, Q., Donahue, S. L., Moore-Jarrett, T., Cao, S., Osipovich, A. B., and Ruley, H. E. (2006). Mutagenesis of diploid mammalian genes by gene entrapment. Nucleic Acids Res. 34, e139. Luo, G., Ivics, Z., Izsvak, Z., and Bradley, A. (1998). Chromosomal transposition of a Tc1/ mariner-like element in mouse embryonic stem cells. Proc. Natl. Acad. Sci. USA 95, 10769–10773.

Gene Trapping in Mice

267

Mainguy, G., Montesinos, M. L., Lesaffre, B., Zevnik, B., Karasawa, M., Kothary, R., Wurst, W., Prochiantz, A., and Volovitch, M. (2000). An induction gene trap for identifying a homeoprotein-regulated locus. Nat. Biotechnol. 18, 746–749. Markowitz, D., Goff, S., and Bank, A. (1988). A safe packaging line for gene transfer: Separating viral genes on two different plasmids. J. Virol. 62, 1120–1124. Matsuda, E., Shigeoka, T., Iida, R., Yamanaka, S., Kawaichi, M., and Ishida, Y. (2004). Expression profiling with arrays of randomly disrupted genes in mouse embryonic stem cells leads to in vivo functional analysis. Proc. Natl. Acad. Sci. USA 101, 4170–4174. Medico, E., Gambarotta, G., Gentile, A., Comoglio, P. M., and Soriano, P. (2001). A gene trap vector system for identifying transcriptionally responsive genes. Nat. Biotechnol. 19, 579–582. Merrick, W. C. (2004). Cap-dependent and cap-independent translation in eukaryotic systems. Gene 332, 1–11. Mikkers, H., Allen, J., Knipscheer, P., Romeijn, L., Hart, A., Vink, E., and Berns, A. (2002). High-throughput retroviral tagging to identify components of specific signaling pathways in cancer. Nat. Genet. 32, 153–159. Mitchell, K. J., Pinson, K. I., Kelly, O. G., Brennan, J., Zupicich, J., Scherz, P., Leighton, P. A., Goodrich, L. V., Lu, X., Avery, B. J., Tate, P., Dill, K., et al. (2001). Functional analysis of secreted and transmembrane proteins critical to mouse development. Nat. Genet. 28, 241–249. Niwa, H., Araki, K., Kimura, S., Taniguchi, S., Wakasugi, S., and Yamamura, K. (1993). An efficient gene-trap method using poly A trap vectors and characterization of gene-trap events. J. Biochem. 113, 343–349. Nord, A. S., Chang, P. J., Conklin, B. R., Cox, A. V., Harper, C. A., Hicks, G. G., Huang, C. C., Johns, S. J., Kawamoto, M., Liu, S., Meng, E. C., Morris, J. H., et al. (2006). The International Gene Trap Consortium Website: A portal to all publicly available gene trap cell lines in mouse. Nucleic Acids Res. 34, D642–D648. Osipovich, A. B., Singh, A., and Ruley, H. E. (2005). Post-entrapment genome engineering: First exon size does not affect the expression of fusion transcripts generated by gene entrapment. Genome Res. 15, 428–435. Raymond, C. S., and Soriano, P. (2007). High-efficiency FLP and PhiC31 site-specific recombination in mammalian cells. PLoS One 2, e162. Roma, G., Cobellis, G., Claudiani, P., Maione, F., Cruz, P., Tripoli, G., Sardiello, M., Peluso, I., and Stupka, E. (2007). A novel view of the transcriptome revealed from gene trapping in mouse embryonic stem cells. Genome Res. 17, 1051–1060. Roma, G., Sardiello, M., Cobellis, G., Cruz, P., Lago, G., Sanges, R., and Stupka, E. (2008). The UniTrap resource: Tools for the biologist enabling optimized use of gene trap clones. Nucleic Acids Res. 36, D741–D746. Salminen, M., Meyer, B. I., and Gruss, P. (1998). Efficient poly A trap approach allows the capture of genes specifically active in differentiated embryonic stem cells and in mouse embryos. Dev. Dyn. 212, 326–333. Schnieke, A., Harbers, K., and Jaenisch, R. (1983). Embryonic lethal mutation in mice induced by retrovirus insertion into the alpha 1(I) collagen gene. Nature 304, 315–320. Schnu¨tgen, F., De-Zolt, S., Van Sloun, P., Hollatz, M., Floss, T., Hansen, J., Altschmied, J., Seisenberger, C., Ghyselinck, N. B., Ruiz, P., Chambon, P., Wurst, W., et al. (2005). Genomewide production of multipurpose alleles for the functional analysis of the mouse genome. Proc. Natl. Acad. Sci. USA 102, 7221–7226. Schnu¨tgen, F., Stewart, A. F., von Melchner, H., and Anastassiadis, K. (2006). Engineering embryonic stem cells with recombinase systems. Methods Enzymol. 420, 100–136. Schnu¨tgen, F., Hansen, J., De-Zolt, S., Horn, C., Lutz, M., Floss, T., Wurst, W., Noppinger, P. R., and von Melchner, H. (2008). Enhanced gene trapping in mouse embryonic stem cells. Nucleic Acids Res. 36, e133.

268

Roland H. Friedel and Philippe Soriano

Shigeoka, T., Kawaichi, M., and Ishida, Y. (2005). Suppression of nonsense-mediated mRNA decay permits unbiased gene trapping in mouse embryonic stem cells. Nucleic Acids Res. 33, e20. Singla, V., Hunkapiller, J., Santos, N., Seol, A. D., Norman, A. R., Wakenight, P., Skarnes, W. C., and Reiter, J. F. (2010). Floxin, a resource for genetically engineering mouse ESCs. Nat. Methods 7, 50–52. Skarnes, W. C., Auerbach, B. A., and Joyner, A. L. (1992). A gene trap approach in mouse embryonic stem cells: The lacZ reported is activated by splicing, reflects endogenous gene expression, and is mutagenic in mice. Genes Dev. 6, 903–918. Skarnes, W. C., Moss, J. E., Hurtley, S. M., and Beddington, R. S. (1995). Capturing genes encoding membrane and secreted proteins important for mouse development. Proc. Natl. Acad. Sci. USA 92, 6592–6596. Skarnes, W. C., von Melchner, H., Wurst, W., Hicks, G., Nord, A. S., Cox, T., Young, S. G., Ruiz, P., Soriano, P., Tessier-Lavigne, M., Conklin, B. R., Stanford, W. L., et al. (2004). A public gene trap resource for mouse functional genomics. Nat. Genet. 36, 543–544. Soriano, P., Friedrich, G., and Lawinger, P. (1991). Promoter interactions in retrovirus vectors introduced into fibroblasts and embryonic stem cells. J. Virol. 65, 2314–2319. Stryke, D., Kawamoto, M., Huang, C. C., Johns, S. J., King, L. A., Harper, C. A., Meng, E. C., Lee, R. E., Yee, A., L’Italien, L., Chuang, P. T., Young, S. G., et al. (2003). BayGenomics: A resource of insertional mutations in mouse embryonic stem cells. Nucleic Acids Res. 31, 278–281. Stuhlmann, H. (2003). Gene trap vector screen for developmental genes in differentiating ES cells. Methods Enzymol. 365, 386–406. Swift, S., Lorens, J., Achacoso, P., and Nolan, G. P. (2001). Rapid production of retroviruses for efficient gene delivery to mammalian cells using 293T cell-based systems. Curr. Protoc. Immunol. Chapter 10, Unit 10, 17C. Szymczak, A. L., Workman, C. J., Wang, Y., Vignali, K. M., Dilioglou, S., Vanin, E. F., and Vignali, D. A. (2004). Correction of multi-gene deficiency in vivo using a single ‘‘selfcleaving’’ 2A peptide-based retroviral vector. Nat. Biotechnol. 22, 589–594. Tanaka, T. S., Davey, R. E., Lan, Q., Zandstra, P. W., and Stanford, W. L. (2008). Development of a gene-trap vector with a highly sensitive fluorescent protein reporter system for expression profiling. Genesis 46, 347–356. Taniwaki, T., Haruna, K., Nakamura, H., Sekimoto, T., Oike, Y., Imaizumi, T., Saito, F., Muta, M., Soejima, Y., Utoh, A., Nakagata, N., Araki, M., et al. (2005). Characterization of an exchangeable gene trap using pU-17 carrying a stop codon-beta geo cassette. Dev. Growth Differ. 47, 163–172. Tate, P., Lee, M., Tweedie, S., Skarnes, W. C., and Bickmore, W. A. (1998). Capturing novel mouse genes encoding chromosomal and other nuclear proteins. J. Cell Sci. 111 (Pt. 17), 2575–2585. To, C., Epp, T., Reid, T., Lan, Q., Yu, M., Li, C. Y., Ohishi, M., Hant, P., Tsao, N., Casallo, G., Rossant, J., Osborne, L. R., et al. (2004). The Centre for Modeling Human Disease Gene Trap resource. Nucleic Acids Res. 32, D557–D559. Trombly, M. I., Su, H., and Wang, X. (2009). A genetic screen for components of the mammalian RNA interference pathway in Bloom-deficient mouse embryonic stem cells. Nucleic Acids Res. 37, e34. Tsakiridis, A., Tzouanacou, E., Larralde, O., Watts, T. M., Wilson, V., Forrester, L., and Brickman, J. M. (2007). A novel triple fusion reporter system for use in gene trap mutagenesis. Genesis 45, 353–360. Tsakiridis, A., Tzouanacou, E., Rahman, A., Colby, D., Axton, R., Chambers, I., Wilson, V., Forrester, L., and Brickman, J. M. (2009). Expression-independent gene trap vectors for random and targeted mutagenesis in embryonic stem cells. Nucleic Acids Res. 37, e129.

Gene Trapping in Mice

269

Uren, A. G., Mikkers, H., Kool, J., van der Weyden, L., Lund, A. H., Wilson, C. H., Rance, R., Jonkers, J., van Lohuizen, M., Berns, A., and Adams, D. J. (2009). A highthroughput splinkerette-PCR method for the isolation and sequencing of retroviral insertion sites. Nat. Protoc. 4, 789–798. von Melchner, H., DeGregori, J. V., Rayburn, H., Reddy, S., Friedel, C., and Ruley, H. E. (1992). Selective disruption of genes expressed in totipotent embryonal stem cells. Genes Dev. 6, 919–927. Wagner, E. F., Covarrubias, L., Stewart, T. A., and Mintz, B. (1983). Prenatal lethalities in mice homozygous for human growth hormone gene sequences integrated in the germ line. Cell 35, 647–655. Wang, W., and Bradley, A. (2007). A recessive genetic screen for host factors required for retroviral infection in a library of insertionally mutated Blm-deficient embryonic stem cells. Genome Biol. 8, R48. Wang, W., Lin, C., Lu, D., Ning, Z., Cox, T., Melvin, D., Wang, X., Bradley, A., and Liu, P. (2008). Chromosomal transposition of PiggyBac in mouse embryonic stem cells. Proc. Natl. Acad. Sci. USA 105, 9290–9295. Wang, W., Bradley, A., and Huang, Y. (2009). A piggyBac transposon-based genome-wide library of insertionally mutated Blm-deficient murine ES cells. Genome Res. 19, 667–673. Xin, H. B., Deng, K. Y., Shui, B., Qu, S., Sun, Q., Lee, J., Greene, K. S., Wilson, J., Yu, Y., Feldman, M., and Kotlikoff, M. I. (2005). Gene trap and gene inversion methods for conditional gene inactivation in the mouse. Nucleic Acids Res. 33, e14. Yoshida, M., Yagi, T., Furuta, Y., Takayanagi, K., Kominami, R., Takeda, N., Tokunaga, T., Chiba, J., Ikawa, Y., and Aizawa, S. (1995). A new strategy of gene trapping in ES cells using 30 RACE. Transgenic Res. 4, 277–287. Zambrowicz, B. P., Friedrich, G. A., Buxton, E. C., Lilleberg, S. L., Person, C., and Sands, A. T. (1998). Disruption and sequence identification of 2,000 genes in mouse embryonic stem cells. Nature 392, 608–611. Zambrowicz, B. P., Abuin, A., Ramirez-Solis, R., Richter, L. J., Piggott, J., BeltrandelRio, H., Buxton, E. C., Edwards, J., Finch, R. A., Friddle, C. J., Gupta, A., Hansen, G., et al. (2003). Wnk1 kinase deficiency lowers blood pressure in mice: A genetrap screen to identify potential targets for therapeutic intervention. Proc. Natl. Acad. Sci. USA 100, 14109–14114.

C H A P T E R

F O U R T E E N

A Wider Context for Gene Trap Mutagenesis Joshua M. Brickman,* Anestis Tsakiridis,* Christine To,† and William L. Stanford† Contents 1. 2. 3. 4. 5.

Introduction Gene Trap Vectors Random Integration as a Means to Generate New Mutations Targeted Trapping Public Resources of Mutagenized ES Cell Clones 5.1. Gene trap resources 5.2. High-throughput gene targeting resources 5.3. New tools 6. Gene Discovery and Annotation 7. Hypothesis-Driven Screens 8. Protocols 8.1. ES cell transfection, colony picking, and expansion 8.2. Gene trap vectors for trapping and targeting 8.3. Splinkerette Acknowledgments References

272 273 278 279 280 280 284 285 285 287 288 288 289 290 291 291

Abstract Gene trapping is a technology originally developed for the simultaneous identification and mutation of genes by random integration in embryonic stem (ES) cells. While gene trapping was developed before efficient and high-throughput gene targeting, a significant proportion of the publically available mutant ES cell lines and mice were generated through a number of large-scale gene trapping initiatives. Moreover, elements of gene trap vectors continue to be incorporated into gene targeting vectors as a means to increase the efficiency of homologous recombination. Here, we review the current state of gene trapping technology and the applications of specific types of gene trap vector. As a * MRC Centre for Regenerative Medicine, Institute for Stem Cell Research, School of Biological Sciences, University of Edinburgh, Edinburgh, United Kingdom Institute of Biomaterials and Biomedical Engineering, University of Toronto, Toronto, Ontario, Canada

{

Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77014-2

#

2010 Elsevier Inc. All rights reserved.

271

272

Joshua M. Brickman et al.

component of this analysis, we consider the behavior of specific vector types both from the perspective of their application and how they can inform our annotation of the mammalian transcriptome. We consider the utility of gene trap vectors as tools for cell-based expression analysis, targeted screening in embryonic differentiation, and for use in cell lines derived from different lineages.

1. Introduction Gene trapping is a method for the simultaneous identification and mutation of genes. Conventionally, gene trap vectors were promoterless, employing splicing to drive the expression of a reporter or antibiotic resistance marker. The use of an enzymatic or fluorescent reporter meant that in addition to generating a potential mutant, the gene trap could also report on the endogenous gene’s expression, allowing expression based screening of potential mutants prior to extensive phenotypic characterization. Gene trapping was initiated in Drosophila by enhancer trapping (O’Kane and Gehring, 1987). In the late 1980s and early 1990s, this rapidly evolved into gene trapping in cell culture. However, with the advent of embryonic stem (ES) cell culture, gene trapping became a means to introduce mutations into the germ line (Friedrich and Soriano, 1991; Gossler et al., 1989; Skarnes et al., 1992). Integration of gene trap vectors could be stably selected for through the use of antibiotic resistance markers, in most cases neomycin, and mutant cell lines could be introduced into the mouse germ line by blastocyst injection or morulae aggregation. This random mutagenesis strategy was pioneered at the same time that a handful of labs generated the first targeted mouse strains via homologous recombination. Gene trapping has also been used in a number of other nonmammalian species as a means to generate insertional mutations in the absence of a technology for homologous recombination (Bronchain et al., 1999; Gaiano et al., 1996; Kawakami et al., 2004; Pietsch et al., 2006). Despite the rapid evolution of gene targeting, gene trapping remained the most rapid means to develop mutagenic resources for some time. However, the high-throughput utility of gene trapping has its limits. Gene trap vectors exhibit bias (discussed in Section 2) based on the level of transcription and the size of genes (Nord et al., 2007). While some of this bias can be corrected with different vector types, high-throughput homologous recombination approaches are expected to be able to more efficiently mutate the remainder of the genome. So in the wake of high-throughput mutagenesis programs designed for genome saturation (Floss and Schnutgen, 2008; Friedel et al., 2007), does gene trapping have a future? In this review, we will focus on the utility of gene trapping in a postgenome world. We will explore how gene trap technology can be applied for both high-throughput homologous

A Wider Context for Gene Trap Mutagenesis

273

recombination and in specific hypothesis-driven screens. While highthroughput targeted trapping is predicated on the ability to define the genes that will be targeted, we will also consider the number of sequences that have been identified in specific gene trap screens and how they correspond to identified reading frames within the genome. As particular screens appear to have generated insertions in transcripts that were not at the time annotated reading frames, we consider whether gene trapping continues to be an effective tool for gene identification.

2. Gene Trap Vectors We have previously reviewed in detail the evolution of various gene trap vectors (Stanford et al., 2001). In addition, Friedel and Soriano (Chapter 13) have also reviewed a number of recent vector technologies and therefore we will not cover this material in extensive detail. In this section, we will overview widely used gene trap vectors, many of which have been used to generate the public gene trap resources, which the reader may also choose to incorporate into hypothesis-driven screens. The most widely used gene trap vectors are promoterless and contain a splice acceptor sequence (SA) upstream of a selectable marker or reporter gene (‘‘SA-type’’ or ‘‘promoter trap vectors,’’ see Fig. 14.1) (Friedrich and Soriano, 1991; Gossler et al., 1989; Skarnes et al., 1995). When this type of vector integrates into a gene transcribed in ES cells, the gene trap cassette’s selectable marker is expressed under the control of the endogenous gene’s promoter. The splicing of the vector onto the endogenous transcript would then also disrupt the endogenous coding sequence and result in the production of a truncated fusion protein that in the ideal scenario would produce a complete loss-of-function allele. However, in reality, it is important to check that the endogenous gene does not generate a wild-type transcript by splicing around the insertion. Splice acceptor vectors generally rely on a fusion of the antibiotic resistance gene to a selectable cassette, with the most popular being a fusion of b-galactosidase to the neomycin phosphotransferase gene (neo) known as bgeo. Recent improvements to this design have included the use of the T2A (Wang et al., 2009) peptide, fusion of GFP directly to antibiotic resistance (Cobellis et al., 2005), and our vectors that have employed a triple fusion that includes both GFP and either bgeo or a hygromycin phosphotransferase gene (Hygro) equivalent, bhygro (Tsakiridis et al., 2007). As this type of vector employs selection that relies on the endogenous genes transcription, the stringency of antibiotic selection influences the range of insertions generated. As neo is extremely sensitive, two versions of the neo allele are currently used. The wild-type allele requires extremely low levels of

274

Joshua M. Brickman et al.

Promoter trapping Active target locus

Promoter targeted trapping Active target locus

Enhancers P Exon 1

SA-based gene trap vector

SA

Exon 2

bgeo

Exon 3

Enhancers P Exon 1

Exon 2

Exon 3

SA-based gene targeting vector

pA

Homology arms Homologous recombination

Splicing

Fusion transcript

Splicing

Fusion protein 5⬘RACE PCR aids the cloning of the trapped gene Reporter expression reflects trapped locus activity

Fusion transcript

Germ line transmission

Fusion protein

Use in conjunction with in vitro ES cell differentiation

PolyA trapping Inactive target locus

PolyA targeted trapping

Enhancers

Inactive target locus

pA P Exon 1

PolyA trap vector

Exon 2

Exon 3

Enhancers

PolyA-trap gene targeting vector

SA Reporter pA P Selector SD

pA P Exon 1

Exon 2

Exon 3

SA Reporter pA P Selector SD

pA Splicing

Splicing

Fusion transcript

Homology arms

pA

Homologous recombination

Fusion protein Reporter expression upon gene activation

Drug resistance independent of trapped gene expression

3⬘RACE PCR enables identification of downstream trapped exons

pA Splicing

Fusion transcript Fusion protein

Figure 14.1 Expression-dependent and -independent approaches for transcript entrapment. Promoter or SA-trapping methods rely on the levels of expression of the target locus and can be either random (‘‘promoter trapping’’) or ‘‘targeted,’’ employing homology arms to drive homologous recombination and the introduction of the SA vector into a specific intronic position. PolyA trap strategies are in theory expression-independent as they employ a splice donor element to capture an endogenous polyadenylation signal. They can also be random (‘‘polyA trapping’’) or targeted (‘‘polyA targeted trapping’’). PolyA targeted trapping employs homology arms in the same way as promoter targeted trapping. P, promoter; SA, splice acceptor; pA, polyadenylation signal; SD, splice donor.

expression to generate antibiotic resistance and therefore allows for the selection of stable clones that do not express sufficient levels of the bgeo fusion protein to generate detectable b-galactosidase reporter expression. A mutant version of neo (Skarnes et al., 1995; Yenofsky et al., 1990) that requires higher levels of the enzyme to generate G418 resistance has also been used in a number of gene trap vectors and it appears to increase the frequency of antibiotic resistant colonies that also express the b-galactosidase reporter. We have found that the replacement of the neo coding sequence with Hygro, results in a similar increase in the stringency of selection and this has also been shown to increase the frequency of clones

A Wider Context for Gene Trap Mutagenesis

275

expressing sufficient levels of the reporter gene for either fluorescent or enzymatic detection (Tsakiridis et al., 2007). The use of a splice acceptor to generate a direct fusion of the bgeo or similar reporter to the endogenous genes coding sequence, generates fusion proteins that, in a large number of cases, replicate the subcellular localization of the disrupted gene. These observations have led to the application of bgeo in screens for protein components localized to specific subcellular locations (Sutherland et al., 2001; Tate et al., 1998). The subcellular localization of target proteins was also exploited in the development of a secretory trap, where the b-galactosidase will only be active when it is fused to transmembrane proteins (Skarnes et al., 1995). SA-type vectors have the caveat that they depend on the expression of the disrupted gene. Even with the extremely high sensitivity of the unmodified neo enzyme, selection requires that a gene be expressed in ES cells above a certain threshold level to ensure selection and approximately 60–70% of the genome is expressed at this level (Friedel et al., 2005). To circumvent this problem, vectors have been designed in which selection has been rendered expression independent. These vectors have employed heterologous promoters to drive the expression of selectable markers that lack a polyA sequence, but contain a splice donor (SD) (Fig. 14.1). Integration of this type of vector upstream of a functional polyA sequence then generates a stable transcript and drug resistance (Niwa et al., 1993; Salminen et al., 1998; Zambrowicz et al., 1998). The uncoupling of antibiotic resistance from the requirement for endogenous gene expression implies that polyA trap vectors can theoretically disrupt a wider range of genes including those that are not expressed in ES cells as well as nonprotein-coding transcripts. While the uncoupling of antibiotic resistance from the requirement for endogenous gene expression should enhance the accessibility of the genome to trapping, these vectors have not been without their problems. The combination of a strong internal promoter with an inefficient SD can produce antibiotic resistance even in the absence of splicing onto an endogenous transcript (Hirashima et al., 2004; Zambrowicz et al., 2003), although the inclusion of an RNA instability element downstream of the SD has been incorporated into vector design (Ishida and Leder, 1999; Tsakiridis et al., 2009) and we have recently shown that this can achieve as much as a 7.7-fold enhancement in the splicing efficiency with which the vector is correctly spliced to an endogenous exon (Tsakiridis et al., 2009). Modifications that improve vector efficiency include different promoter and SD combinations (Matsuda et al., 2004; Osipovich et al., 2005; Shigeoka et al., 2005; Zambrowicz et al., 2003) and the insertion of a synthetic intron within the selectable marker gene (Lin et al., 2006). However, these vectors are limited because of the action of an mRNA surveillance mechanism called nonsense-mediated mRNA decay (NMD) (Maquat, 2004; Shigeoka et al., 2005). NMD promotes the selection of

276

Joshua M. Brickman et al.

insertional events in the 30 -most intron of target sequences as it triggers the degradation of the selectable marker’s transcript based on the presence of a premature termination codon (Lin et al., 2006; Osipovich et al., 2005; Shigeoka et al., 2005). Although this bias is still compatible with the engineering of 30 insertions that generate hypomorphic alleles, it clearly jeopardizes the mutagenic potential of these vectors. However, SD vectors can be made position independent through the incorporation of an IRES downstream of the selectable marker (Shigeoka et al., 2005) or in our case, by modifying the promoter used to drive the neo transcript (Tsakiridis et al., 2009). We attempted to quantitate the effect of NMD on specific promoter-SD combinations, and found that the overall level of neo transcription from the particular SD-promoter combination rather than NMD susceptiblity was the relevant factor ensuring selection of individual clones (Tsakiridis et al., 2009). Thus, what determined NMD independence was the levels of neo transcript generated, despite the activity of NMD, which were sufficient to enable antibiotic resistance. In the case of SA and polyA trap combinations, the polyA trap component may still be useful for 30 RACE, despite its being subject to the NMD driven 30 bias (Chen et al., 2004). In addition to the use of either an IRES or alternative promoters to drive neo expression, empirical evidence suggests that the selectable marker can influence NMD susceptibility. Because the overall level of neo expression is crucial to overcome NMD sensitivity, it is not surprising that NMD affects polyA trap vectors encoding the mutant neo (bgeo) version. Furthermore, vectors with different selectable markers appear less affected. Specifically, when we compared the insertional behavior of five vectors based on the set of sequence tags generated in genes with known exon/intron structure, we found that two of our vectors (pGTIV3 and pMS) with wild-type neo and pGTlox4 (containing the blasticidin resistance gene) exhibited only a minimal 30 insertional bias (Fig. 14.2). Unfortunately, the limitations of our analysis are that Gep-SD5 and RET are retroviral vectors while pGTlox4 and pMS are plasmid vectors and multiple integrations could have occurred that increased drug resistance. However, while this caveat may affect some of the insertions considered as part of our dataset, we believe that these events will be rare and do not detract from the conclusions drawn from Fig. 14.2. At least some of the 50 integrations obtained with pGTIV3 were based on a retrovirus and pGTIV3 was used successfully in gene targeting experiments at the 50 side of two genes and the results confirmed by southern blot (Tsakiridis et al., 2009). In conclusion, it appears possible to modulate the affects of NMD by either changing the promoter-SD configuration and/or the selection employed. In a complementary approach, we have co-opted NMD to degrade the trapped transcript, thereby using NMD as the mutagenic agent (http:// www.cmhd.ca/genetrap/). To induce NMD of the fusion transcript between the trapped gene and a fluorescent reporter, we inserted three

277

A Wider Context for Gene Trap Mutagenesis

100

% frequency of clones

90 80 70 60 50 40 30 20 10 0

3⬘intron

5⬘intron pGTIV3

pMS

pGTIox4

3⬘-most intron Gep-SD5

RET

Figure 14.2 PolyA trap vectors utilizing selectable markers other than the mutant neo gene do not exhibit a severe 30 bias. Distribution of insertion sites within trapped genes for five different polyA trap vectors. Vectors pGTIV3 (Tsakiridis et al., 2009) and pMS (1, 2, 3) (Salminen et al., 1998; To et al., 2004 ) employ a splice donor cassette in which the expression of the nonmutant neo gene version is driven by the human b-actin promoter. Vector pGTlox4 (http://www.cmhd.ca/genetrap/vectors.html) (To et al., 2004) contains the blasticidin resistance gene as a selectable marker. Vectors Gep-SD5 (http://www. cmhd.ca/genetrap/vectors.html) (To et al., 2004) and RET (Ishida and Leder, 1999) both employ the mutant neo gene within their splice donor modules. The vector integration sites were determined by performing BLAST analysis of the RACE sequence tags against the Ensembl mouse database. A BioPerl script was used to retrieve the insertion sites via the Ensembl Perl API. Only integrations within genes with well-defined exon–intron structure were included. Insertions within two-exon genes were excluded from the analysis. The 50 - and 30 -introns were defined as being 50 and 30 to the middle exon or intron of a gene, respectively. Vector insertions defined as ‘‘30 -most intron’’ were independently counted and excluded from the 30 -intron group. N ¼ 49 for pGTIV3, N ¼ 114 (pMS), N ¼ 116 (pGTlox4), N ¼ 310 (Gep-SD5), N ¼ 100 (RET).

floxed internal exons containing premature stop codons downstream of the Venus fluorescent reporter. The premature stop codons target the fusion transcript for NMD. For this strategy, 30 terminal integrations are desired as Cre-mediated recombination will often rescue the allele, enabling a C-terminal fluorescently tagged allele to be expressed which can be used for real-time subcellular localization studies. A large class of polyA trap vectors do not just contain the SD selection cassette, but also an SA-type reporter that either contains a second selection and/ or reporter. These reporters can in principle allow for lineage selection in addition to generating integrations. Thus, the first attempts to generate neural differentiation from ES cells employed a reporter gene to select for the expression of an early neural marker, Sox2 (Li et al., 1998).

278

Joshua M. Brickman et al.

3. Random Integration as a Means to Generate New Mutations Based on the rate of accumulation of new mutations, approximately 60–70% of all mouse genes are predicted to be accessible to SA-type vectors (Schnutgen et al., 2005; Skarnes, 2005). The accessibility of a locus to trapping (‘‘trappability’’) depends on both gene size and expression levels (Nord et al., 2007). Furthermore, different gene trap vectors appear to each have their own insertional ‘‘hot spots’’ (Nord et al., 2007). Thus, while 60–70% of the genome may be accessible to random trapping, all the high probability loci have been trapped already and efficiency with which one generates new insertions is progressively decreasing. As a result, while the current level of genome saturation according to IGTC (International Gene Trap Consortium) web site and the latest Ensembl assembly is 37% (www. igtc.org), the next 30% of genes mutated by random trapping will be much less efficient and thus much more costly. However, the advent of polyA trap vectors also appears to broaden the range of the genome that is accessible to gene trap insertions. To determine the rate of trapping of new genes by specific vector types, we took the total sequence tag set available on IGTC with vector annotation and asked how different vector types were performing (Table 14.1). If we define the rate of trapping (Skarnes, 2005) as the percentage of unique targets/total number of sequence tags, SA plasmid vectors appear most effective at 11%, polyA trap (either vector or retroviral) is around 7–8%, Table 14.1 Comparison of gene trapping efficiency between different vectors

Vector type

New gene trapping rate (%)a

New gene trapping efficiency (%)b

PolyA (retroviral) SA (retroviral) PolyA (plasmid) SA (plasmid)

7 (1030/14731) 3 (316/9879) 8 (311/3914) 11 (2868/26381)

17 (1030/6245) 5 (316/6288) 40 (311/783) 15 (2868/19182)

Over 640,000 gene trap sequence entries available from NCBI dbGSS database were downloaded in December 2009. Tags were classified into four vector types based on the available vector information. Retroviral polyA and plasmid polyA vectors were from CMHD. Plasmid SA vectors were primarily from BayGenomics, SIGTR, and GGTC. The retroviral SA was from GGTC and Soriano Gene Trap resources. Sequence tags were mapped to the Ensembl transcripts. Only gene trap insertions that were in the sense orientation to the gene were considered. a No. of genes trapped exclusively by the indicated class of vectors/total no. of cDNA sequence tags generated through the use of indicated class of vectors. b No. of genes trapped exclusively by the indicated class of vectors/total no. of cDNA sequence tags generated through the use of indicated class of vectors and corresponding to exonic sequences of gene transcripts.

A Wider Context for Gene Trap Mutagenesis

279

and retroviral SA vectors at 3%. The low rate currently being achieved by retroviral SA vectors suggests they are particularly biased toward hot spots that are already saturated. However, these values depend on the way in which a gene is defined and polyA traps appear to have an enhanced capacity to trap unknown, but transcribed sequences (Roma et al., 2007; Tsakiridis et al., 2009). As a result we have assessed the performance of each vector by a new criteria, new gene trapping efficiency (Table 14.1), the number of unique genes identified/number of sequence tags corresponding to annotated genes. Based on this indicator, the polyA trap plasmid vectors have a ‘‘new gene trapping efficiency’’ of 40%, while the plasmid-based splice acceptor vectors are trapping new genes at a rate of 15%. As with the above analysis there appears a reduction in the rates of trapping when retroviral vectors are compared to both their plasmid counterparts. Taken together, these data suggest that random integration is still an effective means to generate new mutations using specific vectors.

4. Targeted Trapping Historically, SA-type vectors have been used for gene targeting in the case of genes expressed at high levels in ES cells (e.g., Oct4; Mountford et al., 1994) and that this has traditionally been a means to increase the efficiency of homologous recombination. Conventional gene targeting relies on a promoter driven selection cassette that generates antibiotic resistance whether it is introduced into the locus of the target gene or not. A SA-type gene trap vector has no promoter and therefore it is much less likely to generate antibiotic resistance through random integration and thus will generate fewer colonies, but with the majority being a result of homologous recombination. Therefore, Friedel et al. (2005) attempted to target 24 genes using a SA-type vector via homologous recombination. Remarkably, 18 genes were targeted at frequencies of greater than 50%. These finding suggested that promoterless targeting, or so called ‘‘targeted trapping,’’ had the potential to dramatically improve targeting efficiencies for a wide range of genes. As a result, this technology has been extensively utilized to target ES cell expressed genes in public high-throughput initiatives (KOMP, EuCOMM, NorCOMM) focusing mainly on the conditional disruption of currently untrapped loci by gene targeting (Floss and Schnutgen, 2008; Friedel et al., 2007). This has been largely effective and the majority of genes left to target through these initiatives are probably not expressed in ES cells at sufficient levels for promoterless targeting. Based on the effective range of SD-type vectors, it is possible that they can also improve the efficiency of gene targeting. While they contain an internal promoter, effective SD-type vectors require integration in a gene

280

Joshua M. Brickman et al.

sequence and proper splicing for selection and therefore could reduce the background of random integration that normally occurs in gene targeting experiments. We have recently demonstrated that these vectors can function in homologous recombination and shown that for at least one untrappable gene, SD-type targeting is effective (Tsakiridis et al., 2009). Only SDtype vectors that are not subject to NMD are applicable for homologous recombination, but even with this limited set of vectors the potential to improve global targeting efficiencies is significant although this needs to be rigorously tested.

5. Public Resources of Mutagenized ES Cell Clones We have previously presented a users’ guide to publically available gene trap resources, which included an overview of online resources including basic protocols for bioinformatic and gene trap gene annotation, and advice on clone ordering and confirmation of trapped gene identity (Stanford et al., 2006). Here, we will confine this section to updating the gene trapping resources and reviewing the new high-throughput conditional gene targeting resources, which in large part developed out of the targeted trapping approach. Finally, we will survey novel tools available to enrich the utility of the gene trap and targeted resources.

5.1. Gene trap resources For some years, the development of a mutational resource through gene trapping has been a major goal of large-scale international collaborations organized through the IGTC (Nord et al., 2006; Skarnes et al., 2004). Table 14.2 depicts the publically available gene trap resources according to the generators and distributors of the resources and the web sites to request ES cell clones. The autonomy of each group within the IGTC has led to a variety of different types of gene trap vectors and parental ES cell lines used, enriching and expanding the resource. As of February 2010, the IGTC resource contains 435,842 ES cell clones, which represents 11,941 genes by the Ensembl genome database (or 37.9% of the genome) or 15,084 genes by Entrez/NCBI curation (or 24.72% of the genome). The entire IGTC data set can be displayed on the Ensembl genome browser (http:// www.ensembl.org) by toggling the gene trap ‘‘DAS’’ (distributed annotation system) check-box. The ‘‘Detailed View’’ window of the ‘‘ContigView’’ page contains a dropdown menu labeled ‘‘DAS’’ which can be checked to display the mapped gene trap sequence tags.

Table 14.2 Publically available gene trap and gene targeting resources Resource generator

Gene trap resources International Gene Trap Consortium (IGTC) International Mouse Strain Resource (IMSR) BayGenomics Genetrap Project Sanger Institute Gene Trap Resource (SIGTR) Soriano Gene Trap Resource Centre for Modeling Human Disease Genetrap Resource (CMHD) European Conditional Mouse Mutagenesis Program (EuCOMM) German Gene Trap Consortium (GGTC) Mammalian Functional Genomics Centre (MFGC) Telethon Institute of Genetics and Medicine (TIGEM)–IRBM Gene Trap Resource Texas A&M Institute for Genomic Medicine (TIGM)

Resource distributor

Resource distributor URL

Online catalog only

http://www.genetrap.org/

Online catalog only

http://www.findmice.org//index.jsp

Mouse Mutant Regional Resource Centers (MMRRC) Mouse Mutant Regional Resource Centers (MMRRC) Mouse Mutant Regional Resource Centers (MMRRC) Canadian Mouse Mutant Repository (CMMR) European Mouse Mutant Repository (EuMMCR) European Mouse Mutant Repository (EuMMCR) Canadian Mouse Mutant Repository (CMMR) TIGEM–IRBM

http://www.mmrrc.org/distribution/ overview_BG.html http://www.mmrrc.org/distribution/ overview_SIGTR.html http://www.mmrrc.org/distribution/ overview_SorianoGeneTrap.html http://www.phenogenomics.ca/services/ cmmr/escell_services.html http://www.eummcr.org/products/es-cells.php

TIGM

http://www.tigm.org/database/

http://www.eummcr.org/products/es-cells.php http://www.phenogenomics.ca/services/ cmmr/escell_services.html http://genetrap.tigem.it/public/browse.php

(continued)

Table 14.2 (continued) Resource generator

Resource distributor

Resource distributor URL

Exchangeable Gene Trap Clones Project (EGTC) Gene targeted resources International Knockout Mouse Consortium (IKMC) European Conditional Mouse Mutagenesis Program (EuCOMM) Knockout Mouse Project (KOMP) North American Conditional Mouse Mutagenesis Project (NorCOMM)

Kumamoto University

http://egtc.jp/action/access/index

Online catalog only for all targeted IKMC resources European Mouse Mutant Repository (EuMMCR) KOMP Repository Canadian Mouse Mutant Repository (CMMR)

http://www.knockoutmouse.org/ http://www.eummcr.org/products/es-cells.php https://www.komp.org/ http://www.phenogenomics.ca/services/cmmr/ escell_services.html

A Wider Context for Gene Trap Mutagenesis

283

The vast majority of the IGTC clones are mutagenized by variations on the SA-based vector and delivered to ES cells by retrovirus. However, the resources generated by BayGenomics and the Sanger Institute Gene Trap Resource (SIGTR) (Stryke et al., 2003) as well as the Telethon Institute of Genetics and Medicine-IRBM Gene Trap Resource (TIGEM) and Exchangeable Gene Trap Clones Project (EGTC) have been exclusively plasmid based. Plasmid-based vectors have the disadvantage of inserting as multiple copies per cell—either as concatemers or as singletons at different sites; furthermore, electroporation is more difficult to scale up for highthroughput mutagenesis. However, we are not aware of these issues being problematic for these centers. Furthermore, our preliminary analysis suggests plasmid-based vectors have a different integration bias than retroviral vectors (Table 14.1). Within the CMHD, we tested new vectors as plasmids and put the best gene trap vectors into the pGEN-virus (Soriano et al., 1991) for high-throughput production using a low multiplicity of infection (MOI). BayGenomics and the German Gene Trap Consortium (GGTC) implemented secretory trap screens using complementary vectors to specifically isolate transmembrane and secreted proteins (Hansen et al., 2003; Mitchell et al., 2001). Yet, other vectors demonstrated less subcellular localization bias. One such class of vectors are the U3neo gene entrapment vectors used by Mammalian Functional Genomics Centre (MFGC) in Manitoba (Hicks et al., 1997). These vectors do not contain a SA and are therefore designed to trap insertions within exons (although cryptic splice SAs in these vectors have generated splicing based insertions for about half of this library; Osipovich et al., 2004). Vectors that introduce an IRES upstream of a reporter (and thus do not create a 50 fusion transcript) also avoid a subcellular bias and have been used by a number of the gene trap centers including the CMHD (To et al., 2004), EGTC (Taniwaki et al., 2005), and TIGEM (Romito et al., 2010). In addition, the GGTC generated conditional gene trap vectors and generated a cohort of clones encoding insertions with the conditional vectors (Schnutgen et al., 2005). Validated conditional vectors from the GGTC were put into production by the European Conditional Mouse Mutagenesis Program (EuCOMM) using both 129 (E14 TG2a) and C57Bl/ 6 ( JM8F6) ES cell lines. The two currently active IGTC programs, EGTC and TIGEM, are also performing trapping with vectors designed for recombinase-mediated cassette exchange (RMCE) (Romito et al., 2010; Taniwaki et al., 2005) (discussed in Section 5.3 in greater detail) The Soriano lab gene trap resource is populated with insertions generated by two approaches; conventional gene traps employing the ROSA-bgeo vector (the first SA-bgeo gene trap vector; Friedrich and Soriano, 1991) and a polyA trap microarray combined platform known as ROSAFARY (Chen et al., 2004). Former Postdoctoral fellows from the Soriano lab founded

284

Joshua M. Brickman et al.

Lexicon Genetics and used polyA trap vectors to create OMNIBANK I (Zambrowicz et al., 1998), which is now distributed by the Texas A&M Institute for Genomic Medicine (TIGM). OMNIBANK II was created in a C57Bl/6 library by Lexicon Genetics using an SA-bgeo vector through a state of Texas grant for distribution by TIGM. Finally, as discussed above, the CMHD gene trap resource generated a polyA library using vectors that we developed to use NMD as a mutagenesis tool as well as the uPA vector developed by Yasumasa Ishida to suppress NMD (Shigeoka et al., 2005). While much of the early resource was generated in well-validate 129 ES cell lines, as the International Mouse Knockout Consortium (IKMC) that is designed to knock out every gene in the C57Bl/6 background prompted several facilities to put validated C57Bl/6 ES cell lines into the gene trap pipelines. Upon receiving trapped clones, the requesting lab should confirm the identity of the trapped gene and will also need to develop a genotyping strategy. We have found splinkerette PCR (Section 8.3) ideal for genomic integration site identification, which enables the development of a genotyping assay.

5.2. High-throughput gene targeting resources The IKMC (Table 14.2) is comprised of three international consortia to systematically mutagenize all protein-coding genes in the mouse. The prioritization of the genes for knockout is based upon their underrepresentation in the IGTC library and absence of publically available knockouts. In addition, genes could be nominated for knockout on the IKMC web site. All the targeting is being performed on C57Bl/6 ES cell lines demonstrating high levels of germline transmission. The consortium agreed upon generating, when possible, ‘‘null-first, condition-ready’’ alleles, such that the targeted ES cells are essentially a null mutation but can be converted to a conditional allele using recombinase/integrase expression (Collins et al., 2007). However, not all genes have structures amenable to this strategy. Therefore, Regeneron Pharmaceuticals Inc. has been contracted to target 3500 genes using their Bac targeting vectors (VelociGene) technology (Valenzuela et al., 2003). The remainder of the targeting is being performed by consortia including gene trap facilities that switched to high-throughput conditional gene targeting when it became more effective to mutate new genes by homologous recombination as opposed to random mutagenesis. The EuCOMM program modified the GGTC conditional gene trap vector into a conditional gene trap vector, while NorCOMM (which includes the CMHD and MFGC gene trap facilities) developed a novel ‘‘null-first, condition-ready’’ allele. Available targeted genes as well as the status of genes in the targeting pipelines can be followed using the links supplied in Table 14.2.

A Wider Context for Gene Trap Mutagenesis

285

5.3. New tools The combination of multiple gene trap and gene targeting alleles provides allelic series for a sizable percentage of genes. In addition, many of the insertions have been generated using vectors encoding recombination sites to excise the gene trap vector for rescue, recombination sites for conditional mutagenesis, or specifically designed for RMCE. These clones can be exploited for more than mutation analysis, providing there are readily available RMCE vectors for genome modification of the postgene-trapped or -gene-targeted loci. For effective use of these resources, tool boxes of recombinase site flanked cassettes (fluorescent reporters, recombinases, Diphtheria Toxin A, etc.) need to be made available. Indeed, a large-scale toolbox is scheduled to be released for the NorCOMM gene targeting vector at the end of 2010. A number of approaches have been developed for RMCE. The Floxin (or flanked lox site insertion) strategy (Singla et al., 2010) was developed by the Reiter lab for RMCE based on the structure of the pGTLxf and pGTLxr gene trap vectors and incorporates hetero-specific lox sites on either side of the SA. These vectors have been used by both BayGenomics and SIGTR, to make over 24,000 ES cell clones RMCE compatible. The utility of this Floxin was recently demonstrated by Singla et al. (2010), who have used the Floxin technology for replacing the bgeo reporter with eGFP and expressing full length cDNAs fused to C-terminal tandem affinity purification (TAP) tags for a variety of genes. Furthermore, they demonstrate faithful expression levels and subcellular distribution of the TAPtagged proteins. We recently used a Floxin-generated TAP-tagged Suz12 cell line to identify endogenous Suz12 interacting proteins by mass spectroscopy in undifferentiated ES cells. These data corroborated our own coimmunoprecipitation analysis in wild-type ES cells demonstrating that the Polycomb-like 2 (Pcl2) protein interacts with Suz12 in the Polycomb Repressive Complex 2 (PRC2) (Walker et al., 2006). Thus, using the Floxin and other systems, the gene trap and gene targeting resources described can be utilized not only for the analysis of mutagenized alleles but also cell biology and biochemical analyses as well.

6. Gene Discovery and Annotation It has been 10 years since the first comprehensive assembly of the genome was completed. The current level of transcriptomics is based on cDNA identified as ESTs from expression libraries and predicted or open reading frames from genome sequence. However, this annotation is limited by two factors, our definition of a gene as a translated sequence and our

286

Joshua M. Brickman et al.

capacity to predict intron and exon junctions. The ability of gene trap vectors to identify transcripts based solely on the existence of an active promoter and splicing has led to the identification of increasing numbers of novel transcripts. Moreover, global gene expression data sets have used microarray technology that was based on this annotation, whereas gene trapping is a random process. Gene trap sequence tags do not always fit with canonical predicted or established reading frames. These sequences, most commonly identified with polyA trap vectors were generally dismissed as ‘‘nongenic’’ (Nord et al., 2007). However, RT PCR expression analysis of wild-type ES cells using primers designed to detect such RACE products indicated that these represent genuine transcripts that are not present in the EST databases (Chen et al., 2004; Roma et al., 2007; Tsakiridis et al., 2009). Moreover, some of these sequences have been shown to be expressed during development and in a tissue-specific manner (Chen et al., 2004; Matsuda et al., 2004; Roma et al., 2007). The level of conservation observed in some of these transcription units suggests that in a number of cases these transcripts represent both novel exon sequences in already described genes or entirely new genes (Roma et al., 2007; Tsakiridis et al., 2009). In addition, a large number of these novel sequences appear to have no clear coding sequence and therefore could represent regulatory RNAs (Roma et al., 2007). Of these new sequences in 2007, Roma et al. (2007) suggested that only 59% are predicted with ab inititio predictions made by GENSCAN. However, we have observed that an increasing number of these sequences appear in GENSCAN predictions with time (A. Tsakiridis and J. M. Brickman, unpublished). This suggests refinements to the GENSCAN algorithm based upon new training sets, which may include gene trap data. While the transition from fixed array-based platforms to transcriptome sequencing may begin to uncover some of these noncanonical transcripts, even direct sequencing is still a population-based assay. At the level of expression analysis, the total set of genes that are trapped at a high frequency with SA-type vectors are expressed at reasonable levels in ES cells, but show remarkable little overlap with ES cell expression data sets (Roma et al., 2007). Moreover, while the ES cell expression data set represents large genes that are expressed at reasonable levels, the expression of a large number of these highly trapped genes is heterogeneous (Tanaka et al., 2008; Tsakiridis et al., 2007). That this heterogeneity has functional significance has recently been suggested by studies that show heterogeneous expression by key ES cell lineage determinants—such as Nanog (Chambers et al., 2007), Stella (Hayashi et al., 2008), and Rex1 (Toyooka et al., 2008). We have made similar observations about the expression of the early endoderm transcript, Hex and been able to purify the Hex expressing subpopulation of ES cells and show that they represent a reversible, stable selfrenewing, lineage biased component of ES cell cultures (Canham et al., 2009).

A Wider Context for Gene Trap Mutagenesis

287

The significance of these observations has previously been obscured by a dependence on population-based measures. However, gene trapping generates single cell resolution data and the heterogeneity observed in gene trapping may provide important insights into the molecular basis for ES cell pluripotency and early lineage determination (Silva and Smith, 2008; Tanaka et al., 2008). Moreover, as SA-type technology is based on selection, gene trapping is an effective means of identifying genes that are highly expressed in a small subpopulation of ES cells and provides a means by which they can be selectively expanded.

7. Hypothesis-Driven Screens While recent attention has focused on laboratories and consortia performing high throughput sequence based insertional mutagenesis, hypothesis-driven screens previously predominated the gene trap field. Since reporter genes are encoded in most gene trap vectors, gene trapping lends itself to hypothesis-driven screens based upon changes in reporter expression. Large-scale in vivo gene trap expression screens have been successfully performed (Leighton et al., 2001; Wurst et al., 1995); however, most investigators have opted to utilize the in vitro differentiation capacity of ES cells or the ability of ES cells to respond to physiological stimuli (Bautch et al., 1996; Doetschman et al., 1985; Nakano et al., 1994) as a surrogate assay for in vivo expression or gene function. Examples of expression-based screens can be found beginning in the mid-1990s—including induction screens identifying retinoic acid responsive genes, homeobox targets, and ionizing radiation-regulated genes (Forrester et al., 1996; Mainguy et al., 2000; Vallis et al., 2002) and developmental screens for genes regulated in hematopoiesis, angiogenesis, and neurogenesis (Carroll et al., 2001; Hirashima et al., 2004; Kuhnert and Stuhlmann, 2004; Roberts et al., 2004; Stanford et al., 1998). These screens were generally performed using vectors encoding b-galactosidase as the reporter. New gene trap vectors encoding sensitive fluorescent reporter systems (Tanaka et al., 2008; Tsakiridis et al., 2007), particularly when combined with high content imaging, can take advantage of the opportunity to analyze gene transcription in individual cells as well as populations (Tanaka et al., 2008). The contributions of gene trapping have generally fit well under the heading of this volume, ‘‘Techniques in Mouse Development,’’ because gene trapping has principally been used in ES cells. However, the future of gene trapping may lie outside of mouse developmental biology or at least outside of mouse ES cell mutagenesis. For example, a couple of laboratories have exploited gene trapping in rodent NSCs to identify genes downstream oxidative stress, various differentiation agonists, and targets of extracellular

288

Joshua M. Brickman et al.

matrix signaling (Moritz et al., 2008; Scheel et al., 2005). Gene trapping has been performed in human ES cells (Dhara and Benvenisty, 2004) and could be exploited in many other developmentally important cell lines including trophoblast stem (TS) cells (Tanaka et al., 1998) and extraembryonic endoderm (XEN) cells (Kunath et al., 2005). The use of strategies that induce homozygosity would enable functional as well as expression screens (Wang et al., 2009). Similarly, we performed a loss-of-function screen to identify tumor suppressor genes involved in glioblastoma multiforme (GBM) (Kamnasaran et al., 2007), the most common and aggressive adult primary brain tumor. This was performed in nontransformed primary astrocytes derived from a mouse model of GBM in which v12Ha-Ras is expressed by the GFAP promoter. We knew that additional mutations were required to drive tumor formation between 3 and 4 months of age and that until this time point explanted astrocytes could not form colonies in soft-agar. Nontransformed neonatal GFAP: v12Ha-Ras astrocytes were infected with retroviral-delivered SA-bgeo vector and plated in soft-agar. Soft-agar colonies were isolated from gene trap insertions in the Gata6 gene. We found that the transformed Gata6-trapped astrocyte clones could form GBM tumors upon syngeneic intracranial implantation. More importantly, we found that while only 10% low-grade human astrocytomas lacked GATA6 expression, 85% of human GBMs lacked GATA6 expression. Furthermore, overexpression of GATA6 in GBM cell lines extinguished their capacity to generate brain tumors in transplanted Nod/Scid mice, demonstrating that this gene trap approach isolated a novel human brain tumor suppressor gene (Kamnasaran et al., 2007).

8. Protocols 8.1. ES cell transfection, colony picking, and expansion All of the gene trap screens performed by the Brickman lab utilized E14TG2a ES cells, while the Stanford lab used R1 ES cells. The cell lines used are up to the individual investigator. For simplicity, we describe the cellular protocols currently being used by the Brickman lab. The most recent Stanford lab protocols can be found online (http://www.cmhd.ca/ genetrap/protocols.html). E14TG2a ES cells are maintained as described previously (Livigni et al., 2009; Morrison and Brickman, 2006; Tsakiridis et al., 2007, 2009; Zamparini et al., 2006) in ES cell medium (1 Glasgow Minimal Essential Medium supplemented with 10% fetal calf serum, 0.25% sodium bicarbonate, 0.1% nonessential amino acids, 2 mM L-glutamine, 1 mM sodium pyruvate, 0.1 mM b-mercaptoethanol, 100 U/ml LIF). We have employed successfully both plasmid and retroviral gene trap vectors

A Wider Context for Gene Trap Mutagenesis

289

and the advantages and disadvantages of both the approaches are discussed in Chapter 13 of this volume. While retroviral-based vectors appear to be more biased toward insertional hotspots compared to plasmid ones (see Section 3), their use ensures the integration of a single intact vector. For electroporation, we routinely use 100 g of linearized plasmid DNA for 107 cells in 800 l of PBS. The parameters we employ are 3 F/0.8 kV for a BioRad Gene Pulser and the time constant should be 0.1. Electroporated cells are resuspended in 19 ml of warm ES cell medium and split into 20 10-cm dishes. For retroviral infection, we first transfect our gene trap vector pGep-IV3 into PhoenixTM cells (2–3  106) by lipofection and after culture for 2–3 days the supernatant is collected and filtered through a 0.22-m filter. Supernatant diluted at 1:10 in ES cell medium containing polybrene (4 g/ml) is then added (4 ml) to E14TG2a ES cells grown in 10 cm dishes at subconfluency in polybrene-supplemented ES cell medium. After 24 h (for electroporation) or 48 h (retroviral infection), the medium is replaced with fresh medium supplemented with the appropriate antibiotic for selection. To avoid batch-to-batch variation, we perform a kill curve analysis in unmodified E14TG2a ES cells to determine the optimal antibiotic concentration for selection. In most cases, we employ 200–250 g/ml for G418 and 150–180 g/ml for hygromycin. After 5–10 days of selection, visible colonies appear. For colony picking, we replace the medium with 5 ml of PBS and single clones are picked manually with a micropipette and transferred into 96-well U-bottom plates containing 20 l of PBS per well. Individual clones are then trypsinized by adding 30 l of trypsin per well and incubating for 5 min at 37  C. The trypsin is then quenched with 150 l of selective medium and cells are triturated into single cell suspensions, which are then transferred to a 96-well flat-bottomed plate. After reaching confluency, clones in each well are split into a 96-well plate for subsequent freezing and a 24-well plate for further expansion. We routinely use clones grown in 24-well plates for RNA isolation and clones in 6-well plates for genomic DNA (gDNA) extraction.

8.2. Gene trap vectors for trapping and targeting The design of our gene trap vectors is based on the combination of different SA-modules with a polyA trap cassette consisting of the human b-actin promoter driving the expression of the nonmutant neomycin resistance gene and the rabbit exon 2/intron 2 SD with the human GM-CS AU-rich element (ARE) (Xu et al., 1997) cloned into the SD’s intron. The combination of the strong human b-actin promoter with the nonmutant neo has been shown to facilitate the selection of insertions that are not severely biased toward the 30 -most intron of target genes while the presence of the ARE appears to reduce background due to nonproductive splicing

290

Joshua M. Brickman et al.

events (Tsakiridis et al., 2009). To facilitate convenient modular vector assembly, our polyA trap cassette is flanked with AscI/PacI restriction sites. Different SA-modules we have employed in combination with this polyA trap cassette include the Engrailed-2 SA-eGFP-bgeo or bhygro and human adenovirus type 3/5 SA-IRES/Gtx-Venus cassettes. For targeted trapping, we have employed the SA-IRES/Gtx-Venus containing polyA trap vector. Our vector construction strategy routinely includes the addition of unique restriction sites to flank the 50 and 30 ends of the vector and the insertion through a polylinker of the same sites into plasmids containing the homology arms for the gene of interest. Targetting arms varied from 2 to 5 kb of homologous sequence and can be incorporated via a modular cloning strategy. Vector electroporation and clone selection is performed as described in section above. Correct targeting events are verified by Southern blotting.

8.3. Splinkerette The advantage of RACE PCR is that it provides an assessment of the fusion transcript. This is particularly informative in terms of mutagenicity if 50 RACE is used. However, the problem with RACE is that it is costly and labor intensive and does not identify the genomic insertion site, a necessary prerequisite for developing a genotyping protocol. Thus, it is ideal to use both DNA- and RNA-based sequence identification strategies to identify the trapped genes. We recommend performing a genomic first screen and then validating clones of interest with RACE. We have found that splinkerette PCR (Devon et al., 1995) works much more robustly, especially in high-throughput than inverse PCR or plasmid rescue. As with most genomic approaches, splinkerette PCR is likely to work more robustly with retrovirus-generated clones as the 50 and 30 ends of the vector are welldefined when using retroviruses. Splinkerette PCR is a linker-mediated genomic PCR-based technique which features a hairpin loop in the linker oligos leading to a decrease in end-repair priming and thus greater specificity (Devon et al., 1995). Splinkerette PCR was first used for identifying gene trap insertions by the GGTC (Horn et al., 2007) and we have adapted their protocols for the CMHD vectors and provide a general protocol below for the validation of the insertion site of individual clones rather than high-throughput (96-well) screening. The step-by-step protocol for the CMHD vectors can be found at http://www.cmhd.ca/genetrap/ protocols.html. gDNA is isolated using standard protocols followed by two isopropanol precipitations. The gDNA is washed in 70% ethanol with the DNA pellet airdried for 15–20 min at room temperature. The gDNA is resuspended in 50–100 ml TE depending on the amount of cells used for the isolation and incubated at 55  C for 2 h to help dissolve the pellet and then stored at 4  C

A Wider Context for Gene Trap Mutagenesis

291

until use. For our vectors, we have optimized restriction digestion using ApoI. Nine microliters of gDNA is digested in a total volume of 20 ml and 5U ApoI at 50  C for 2.5 h followed by 20 min at 80  C. During the digestion, the splinkerette adaptor can be prepared from synthesized oligonucleotides. Add 1.5 ml each of the 10 mM oligonucleotides (SpAa_ApoI, CGAAGAGTAACCGTTGCTAGGAGAGACCGTGGCTGAATGAGACTGGTG TCGACACTAGTGG; SpBb_ApoI, AATTCCACTAG TGTCGACACCAGTCTCTAATTTTTTTTTTCAAAAAAA) to 0.5 ml of SuRE buffer M (Roche) and 6.5 ml of H2O. Incubate the mixture at 97  C for 5 min, then allow the primers to anneal while the temperature reaches room temperature gradually. Add 1 ml of annealed splinkerette adaptor, 3 ml of 10 DNA ligation buffer, 1 ml of T4-DNA ligase (400U), and 5 ml of H2O to the 20 ml of the ApoI-digested DNA, mix, and allow ligation to occur overnight at 16  C. The next day, purify the ligated product using a PCR purification kit, which will remove excess adaptors and salts. Two successive PCR amplifications are then performed each for the 50 splinkerette amplification and for the 30 splinkerette amplification, which are used to identify the 50 and 30 integration sequences, respectively. In the first round PCR, 5 ml of the ligated gDNA is used in a 25-ml PCR reaction. The PCR primers include a vector-specific primer and the Sp0F primer corresponding to splinkerette sequence (CGAAGAGTAACCGTTGCTAGGAGAGACC). We use 33 cycles but the annealing temperature is dependent upon the vector-specific primers used. Using 1 ml of a 1:4 dilution of the first round PCR reaction, a 35-cycle nested PCR reaction is performed in 25 ml reaction volumes. The nested primer for the splinkerette sequence (primer Sp1F) is GTGGCTGAATGAGACTGGTGTCGAC.

ACKNOWLEDGMENTS We thank Lauryl Nutter for helpful comments and Colin McKerlie for Table 14.2. Recent work related to gene trapping has been supported by grants from the Canadian Institutes of Health Research (MOP-74528 & FRN-36653) and Genome Canada (NorCOMM) to W. L. S. and from the Medical Research Council (G0701428), Wellcome Trust (062965), and the Scottish Funding Council to J. M. B. W. L. S. is a Canada Research Chair in Stem Cell Bioengineering and Functional Genomics, J. M. B. is a Senior Non-Clinical Fellow of the Medical Research Council.

REFERENCES Bautch, V. L., et al. (1996). Blood island formation in attached cultures of murine embryonic stem cells. Dev. Dyn. 205, 1–12. Bronchain, O. J., Hartley, K. O., and Amaya, E. (1999). A gene trap approach in Xenopus. Curr. Biol. 9(20), 1195–1198.

292

Joshua M. Brickman et al.

Canham, M., et al. (2010). Functional Heterogeneity of Embryonic Stem Cells Revealed through Translational Amplification of an Early Endodermal Transcript. PLoS Biol. 8(5), e1000379. doi:10.1371/journal.pbio.1000379. Carroll, P., et al. (2001). Sorting nexin-14, a gene expressed in motoneurons trapped by an in vitro preselection method. Dev. Dyn. 221(4), 431–442. Chambers, I., et al. (2007). Nanog safeguards pluripotency and mediates germline development. Nature 450(7173), 1230–1234. Chen, W. V., et al. (2004). Identification and validation of PDGF transcriptional targets by microarray-coupled gene-trap mutagenesis. Nat. Genet. 36(3), 304–312. Cobellis, G., et al. (2005). Tagging genes with cassette-exchange sites. Nucleic Acids Res. 33(4), e44. Collins, F. S., Rossant, J., and Wurst, W. (2007). A mouse for all reasons. Cell 128(1), 9–13. Devon, R. S., Porteous, D. J., and Brookes, A. J. (1995). Splinkerettes-improved vectorettes for greater efficiency in PCR walking. Nucleic Acids Res. 23(9), 1644–1645. Dhara, S. K., and Benvenisty, N. (2004). Gene trap as a tool for genome annotation and analysis of X chromosome inactivation in human embryonic stem cells. Nucleic Acids Res. 32(13), 3995–4002. Doetschman, T. C., et al. (1985). The in vitro development of blastocyst-derived embryonic stem cell lines: Formation of visceral yolk sac, blood islands and myocardium. J. Embryol. Exp. Morphol. 87, 27–45. Floss, T., and Schnutgen, F. (2008). Conditional gene trapping using the FLEx system. Methods Mol. Biol. 435, 127–138. Forrester, L. M., et al. (1996). An induction gene trap screen in embryonic stem cells: Identification of genes that respond to retinoic acid in vitro. Proc. Natl. Acad. Sci. USA 93(4), 1677–1682. Friedel, R. H., et al. (2005). From the cover: Gene targeting using a promoterless gene trap vector (‘‘targeted trapping’’) is an efficient method to mutate a large fraction of genes. Proc. Natl. Acad. Sci. USA 102(37), 13188–13193. Friedel, R. H., et al. (2007). EUCOMM—The European conditional mouse mutagenesis program. Brief. Funct. Genomic. Proteomic. 6(3), 180–185. Friedrich, G., and Soriano, P. (1991). Promoter traps in embryonic stem cells: A genetic screen to identify and mutate developmental genes in mice. Genes Dev. 5(9), 1513–1523. Gaiano, N., et al. (1996). Insertional mutagenesis and rapid cloning of essential genes in zebrafish. Nature 383(6603), 829–832. Gossler, A., et al. (1989). Mouse embryonic stem cells and reporter constructs to detect developmentally regulated genes. Science 244(4903), 463–465. Hansen, J., et al. (2003). A large-scale, gene-driven mutagenesis approach for the functional analysis of the mouse genome. Proc. Natl. Acad. Sci. USA 100(17), 9918–9922. Hayashi, K., et al. (2008). Dynamic equilibrium and heterogeneity of mouse pluripotent stem cells with distinct functional and epigenetic states. Cell Stem Cell 3(4), 391–401. Hicks, G. G., et al. (1997). Functional genomics in mice by tagged sequence mutagenesis. Nat. Genet. 16, 338–344. Hirashima, M., et al. (2004). Gene-trap expression screening to identify endothelial-specific genes. Blood 104(3), 711–718. Horn, C., et al. (2007). Splinkerette PCR for more efficient characterization of gene trap events. Nat. Genet. 39(8), 933–934. Ishida, Y., and Leder, P. (1999). RET: A poly A-trap retrovirus vector for reversible disruption and expression monitoring of genes in living cells. Nucleic Acids Res. 27(24), e35. Kamnasaran, D., et al. (2007). GATA6 is an astrocytoma tumor suppressor gene identified by gene trapping of mouse glioma model. Proc. Natl. Acad. Sci. USA 104(19), 8053–8058.

A Wider Context for Gene Trap Mutagenesis

293

Kawakami, K., et al. (2004). A transposon-mediated gene trap approach identifies developmentally regulated genes in zebrafish. Dev. Cell 7(1), 133–144. Kuhnert, F., and Stuhlmann, H. (2004). Identifying early vascular genes through gene trapping in mouse embryonic stem cells. Curr. Top. Dev. Biol. 62, 261–281. Kunath, T., et al. (2005). Imprinted X-inactivation in extra-embryonic endoderm cell lines from mouse blastocysts. Development 132(7), 1649–1661. Leighton, P. A., et al. (2001). Defining brain wiring patterns and mechanisms through gene trapping in mice. Nature 410(6825), 174–179. Li, M., et al. (1998). Generation of purified neural precursors from embryonic stem cells by lineage selection. Curr. Biol. 8(17), 971–974. Lin, Q., et al. (2006). Mutagenesis of diploid mammalian genes by gene entrapment. Nucleic Acids Res. 34(20), e139. Livigni, A., et al. (2009). Differentiation of embryonic stem cells into anterior definitive endoderm. Curr. Protoc. Stem Cell. Biol. Chapter 1: p. Unit 1G 3. Mainguy, G., et al. (2000). An induction gene trap for identifying a homeoprotein-regulated locus. Nat. Biotechnol. 18(7), 746–749. Maquat, L. E. (2004). Nonsense-mediated mRNA decay: Splicing, translation and mRNP dynamics. Nat. Rev. Mol. Cell Biol. 5(2), 89–99. Matsuda, E., et al. (2004). Expression profiling with arrays of randomly disrupted genes in mouse embryonic stem cells leads to in vivo functional analysis. Proc. Natl. Acad. Sci. USA 101(12), 4170–4174. Mitchell, K. J., et al. (2001). Functional analysis of secreted and transmembrane proteins critical to mouse development. Nat. Genet. 28(3), 241–249. Moritz, S., et al. (2008). An induction gene trap screen in neural stem cells reveals an instructive function of the niche and identifies the splicing regulator sam68 as a tenascin-C-regulated target gene. Stem Cells 26(9), 2321–2331. Morrison, G. M., and Brickman, J. M. (2006). Conserved roles for Oct4 homologues in maintaining multipotency during early vertebrate development. Development 133(10), 2011–2022. Mountford, P., Zevnik, B., Dowel, A., Nichols, J., Li, M., Dani, C., Robertson, M., Chambers, I., and Smith, A. (1994). Dicistronic targeting constructs: reporters and modifiers of mammalian gene expression. Proc. Natl. Acad. Sci. USA 91(10), 4303–4307. Nakano, T., Kodama, H., and Honjo, T. (1994). Generation of lymphohematopoietic cells from embryonic stem cells in culture. Science 265, 1098–1101. Niwa, H., et al. (1993). An efficient gene-trap method using poly A trap vectors and characterization of gene-trap events. J. Biochem. 113(3), 343–349. Nord, A. S., et al. (2006). The International Gene Trap Consortium Website: A portal to all publicly available gene trap cell lines in mouse. Nucleic Acids Res. 34, D642–D648. (database issue). Nord, A. S., et al. (2007). Modeling insertional mutagenesis using gene length and expression in murine embryonic stem cells. PLoS One 2(7), e617. O’Kane, C. J., and Gehring, W. J. (1987). Detection in situ of genomic regulatory elements in Drosophila. Proc. Natl. Acad. Sci. USA 84(24), 9123–9127. Osipovich, A. B., et al. (2004). Activation of cryptic 30 splice sites within introns of cellular genes following gene entrapment. Nucleic Acids Res. 32(9), 2912–2924. Osipovich, A. B., Singh, A., and Ruley, H. E. (2005). Post-entrapment genome engineering: First exon size does not affect the expression of fusion transcripts generated by gene entrapment. Genome Res. 15(3), 428–435. Pietsch, J., et al. (2006). lessen encodes a zebrafish Trap100 required for enteric nervous system development. Development 133(3), 395–406.

294

Joshua M. Brickman et al.

Roberts, D. M., et al. (2004). A vascular gene trap screen defines RasGRP3 as an angiogenesis-regulated gene required for the endothelial response to phorbol esters. Mol. Cell. Biol. 24(24), 10515–10528. Roma, G., et al. (2007). A novel view of the transcriptome revealed from gene trapping in mouse embryonic stem cells. Genome Res. 17(7), 1051–1060. Romito, A., et al. (2010). Lack of sik1 in mouse embryonic stem cells impairs cardiomyogenesis by down-regulating the cyclin-dependent kinase inhibitor p57kip2. PLoS One 5(2), e9029. Salminen, M., Meyer, B. I., and Gruss, P. (1998). Efficient poly A trap approach allows the capture of genes specifically active in differentiated embryonic stem cells and in mouse embryos. Dev. Dyn. 212(2), 326–333. Scheel, J. R., et al. (2005). Quantitative analysis of gene expression in living adult neural stem cells by gene trapping. Nat. Methods 2(5), 363–370. Schnutgen, F., et al. (2005). Genomewide production of multipurpose alleles for the functional analysis of the mouse genome. Proc. Natl. Acad. Sci. USA 102(20), 7221–7226. Shigeoka, T., Kawaichi, M., and Ishida, Y. (2005). Suppression of nonsense-mediated mRNA decay permits unbiased gene trapping in mouse embryonic stem cells. Nucleic Acids Res. 33(2), e20. Silva, J., and Smith, A. (2008). Capturing pluripotency. Cell 132(4), 532–536. Singla, V., et al. (2010). Floxin, a resource for genetically engineering mouse ESCs. Nat. Methods 7(1), 50–52. Skarnes, W. C. (2005). Two ways to trap a gene in mice. Proc. Natl. Acad. Sci. USA 102(37), 13001–13002. Skarnes, W. C., Auerbach, B. A., and Joyner, A. L. (1992). A gene trap approach in mouse embryonic stem cells: The lacZ reported is activated by splicing, reflects endogenous gene expression, and is mutagenic in mice. Genes Dev. 6(6), 903–918. Skarnes, W. C., et al. (1995). Capturing genes encoding membrane and secreted proteins important for mouse development. Proc. Natl. Acad. Sci. USA 92(14), 6592–6596. Skarnes, W. C., et al. (2004). A public gene trap resource for mouse functional genomics. Nat. Genet. 36(6), 543–544. Soriano, P., Friedrich, G., and Lawinger, P. (1991). Promoter interactions in retrovirus vectors introduced into fibroblasts and embryonic stem cells. J. Virol. 65(5), 2314–2319. Stanford, W. L., et al. (1998). Expression trapping: Identification of novel genes expressed in hematopoietic and endothelial lineages by gene trapping in ES cells. Blood 92, 4622–4631. Stanford, W. L., Cohn, J. B., and Cordes, S. P. (2001). Gene-trap mutagenesis: Past, present and beyond. Nat. Rev. Genet. 2(10), 756–768. Stanford, W. L., et al. (2006). Gene trapping in embryonic stem cells. In ‘‘Stem Cell Tools and Other Experimental Protocols,’’ (R. Lanza and I. Klimanskaya, eds.). Stryke, D., et al. (2003). BayGenomics: A resource of insertional mutations in mouse embryonic stem cells. Nucleic Acids Res. 31(1), 278–281. Sutherland, H. G., et al. (2001). Large-scale identification of mammalian proteins localized to nuclear sub-compartments. Hum. Mol. Genet. 10(18), 1995–2011. Tanaka, S., et al. (1998). Promotion of trophoblast stem cell proliferation by FGF4. Science 282(5396), 2072–2075. Tanaka, T. S., et al. (2008). Development of a gene-trap vector with a highly sensitive fluorescent protein reporter system for expression profiling. Genesis 46(7), 347–356. Taniwaki, T., et al. (2005). Characterization of an exchangeable gene trap using pU-17 carrying a stop codon-beta geo cassette. Dev. Growth Differ. 47(3), 163–172. Tate, P., et al. (1998). Capturing novel mouse genes encoding chromosomal and other nuclear proteins. J. Cell Sci. 111(Pt 17), 2575–2585.

A Wider Context for Gene Trap Mutagenesis

295

To, C., et al. (2004). The Centre for Modeling Human Disease Gene Trap resource. Nucleic Acids Res. 32, D557–D559. (database issue). Toyooka, Y., et al. (2008). Identification and characterization of subpopulations in undifferentiated ES cell culture. Development 135(5), 909–918. Tsakiridis, A., et al. (2007). A novel triple fusion reporter system for use in gene trap mutagenesis. Genesis 45(6), 353–360. Tsakiridis, A., et al. (2009). Expression-independent gene trap vectors for random and targeted mutagenesis in embryonic stem cells. Nucleic Acids Res. 37(19), e129. doi: 10.1093/nar/gkp640. Valenzuela, D. M., et al. (2003). High-throughput engineering of the mouse genome coupled with high-resolution expression analysis. Nat. Biotechnol. 21(6), 652–659. Vallis, K. A., et al. (2002). dentification of radiation-responsive genes in vitro using a gene trap strategy predicts for modulation of expression by radiation in vivo. Radiat. Res. 157, 8–18. Walker, E., et al. (2010). Polycomb-like 2 associates with PRC2 and regulates transcriptional networks during mouse embryonic stem cell self-renewal and differentiation. Cell Stem Cell 6, 153–166. Wang, W., Bradley, A., and Huang, Y. (2009). A piggyBac transposon-based genome-wide library of insertionally mutated Blm-deficient murine ES cells. Genome Res. 19(4), 667–673. Wurst, W., et al. (1995). A large-scale gene-trap screen for insertional mutations in developmentally regulated genes in mice. Genetics 139, 889–899. Xu, N., Chen, C. Y., and Shyu, A. B. (1997). Modulation of the fate of cytoplasmic mRNA by AU-rich elements: Key sequence features controlling mRNA deadenylation and decay. Mol. Cell. Biol. 17(8), 4611–4621. Yenofsky, R. L., Fine, M., and Pellow, J. W. (1990). A mutant neomycin phosphotransferase II gene reduces the resistance of transformants to antibiotic selection pressure. Proc. Natl. Acad. Sci. USA 87(9), 3435–3439. Zambrowicz, B. P., et al. (1998). Disruption and sequence identification of 2,000 genes in mouse embryonic stem cells. Nature 392(6676), 608–611. Zambrowicz, B. P., et al. (2003). Wnk1 kinase deficiency lowers blood pressure in mice: A gene-trap screen to identify potential targets for therapeutic intervention. Proc. Natl. Acad. Sci. USA 100(24), 14109–14114. Zamparini, A. L., et al. (2006). Hex acts with beta-catenin to regulate anteroposterior patterning via a Groucho-related co-repressor and Nodal. Development 133(11), 3709–3722.

C H A P T E R

F I F T E E N

Mouse Mutagenesis with the Chemical Supermutagen ENU Frank J. Probst and Monica J. Justice Contents 298 298 299

1. Introduction 1.1. A brief history of mouse mutagens 1.2. ENU’s mechanism of action 1.3. Recovering ENU-induced mutations with various breeding schemes 2. Materials and Methods 2.1. ENU preparation 2.2. Injecting the animals 2.3. Inactivating the ENU 2.4. Breeding the mutagenized males 2.5. Screening for abnormal phenotypes 2.6. Mapping mutant phenotypes and identifying mutations 3. Conclusion Acknowledgments References

300 304 304 305 306 306 307 308 308 310 310

Abstract The generation and analysis of germline mutations in the mouse is one of the cornerstones of modern biological research. The chemical supermutagen N-ethyl-N-nitrosourea (ENU) is the most potent known mouse mutagen and can be used to generate point mutations throughout the mouse genome. The progeny of ENU-mutagenized males can be screened for autosomal dominant phenotypes, or they can be used to generate multigeneration pedigrees to screen for autosomal recessive traits. The introduction of balancer chromosomes into the breeding scheme can allow for the selective capture of mutations in a specific chromosomal region. More recent work has demonstrated that the use of animals that already have a mutation of interest can lead to the successful isolation of additional mutations that modify the original mutant

Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas, USA Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77015-4

#

2010 Elsevier Inc. All rights reserved.

297

298

Frank J. Probst and Monica J. Justice

phenotype. Further, modern molecular techniques ensure that mutations can be readily identified. We describe here the procedures for mutagenizing male mice with ENU and explain the various types of screens that can be performed for different kinds of induced mutations. The currently published research on ENU mutagenesis in the mouse has only scratched the surface of what is possible with this powerful technique, and further work is certain to deepen our knowledge of the role of the individual components of the mouse genome and the myriad relationships between them.

1. Introduction 1.1. A brief history of mouse mutagens The ability to generate and maintain pathologic mutations in the house mouse (Mus musculus) is one of the most powerful tools available to modern scientists. The study of mouse mutations is generally either genotype-driven— in which a specific mutation is engineered in a particular gene, and the resulting animals are analyzed for abnormalities—or phenotype-driven—in which a population of mice are screened for individuals with abnormal phenotypes, and the subsequent analysis focuses on identifying and analyzing the underlying genetic changes that are responsible for the abnormalities. Phenotype-driven mouse mutation screens typically begin with a mutagenesis step in order to increase the number of DNA lesions in the population being screened. Early studies on the ability of various mutagens to generate germline mutations in the mouse focused on the use of a T (test) stock. The original T-stock mice were homozygous for seven different autosomal recessive, viable mutations affecting coat color and ear shape. When a T-stock female is bred to a wild-type male, the progeny will usually be heterozygous at all seven loci and will therefore appear wild-type. However, if the male parent transmits a second recessive mutation at one of the seven loci, then the offspring will have an abnormal coat color or ear shape. Analyzing the number of abnormal progeny allows for a determination of the mutation rate in the male germline at each of the seven loci. This type of experiment is known as a specific-locus test. The specific-locus test was initially used shortly after World War II to demonstrate that X-ray irradiation of male mice can increase the observed mutation rate in their progeny by over 20-fold when compared to nonirradiated mice, which was a substantially greater increase than expected, based on previous work with Drosophila (Russell, 1951). Furthermore, it was demonstrated that in the mouse, unlike in Drosophila, radiation given in a single acute dose is far more mutagenic than chronic exposure to radiation, even when the total dose of radiation is the same (Russell et al., 1958).

299

Mouse Mutagenesis with ENU

The first experiments using the specific-locus test with chemical mutagens were somewhat disappointing. Most chemicals given to male mice did not result in a large increase in the number of progeny with specific-locus mutations (reviewed in Ehling, 1978). Even procarbazine, a powerful chemotherapeutic agent used to treat cancer, was found to induce mutations at only one-third the rate of X-rays (Ehling and Neuhauser, 1979), and neither ethylmethane sulfonate (EMS) nor diethylnitrosamine (DEN), both powerful mutagens in Drosophila, caused a significant increase in the specific-locus mutation rate in mice (Ehling and Neuhauser-Klaus, 1989; Russell and Kelley, 1979). The results with the chemical N-ethyl-N-nitrosourea (ENU; Fig. 15.1), on the other hand, were stunning. When male mice were given a single dose of ENU, they were initially found to have temporary sterility. After 2–3 months, however, many of the animals recovered their fertility. (This prolonged period of infertility suggests that ENU has its principal effect on spermatagonial stem cells as opposed to postmeiotic cells.) The offspring of these animals revealed a mutation rate five times higher than that seen with X-rays (Russell et al., 1979). Furthermore, the administration of multiple doses of ENU over a period of several weeks yielded a mutation rate 12 times higher than X-rays, 36 times higher than procarbazine, and 200 times higher than the spontaneous mutation rate. At this rate, an average of one mutation can be isolated per locus per every 700 gametes (Hitotsumachi et al., 1985; Russell et al., 1982a,b). ENU was therefore deemed a ‘‘supermutagen’’ and declared the chemical mutagen of choice for generating new mouse mutations (Russell et al., 1979).

1.2. ENU’s mechanism of action ENU causes mutations by alkylating DNA. The ethyl group of ENU (Fig. 15.1) can be transferred to a variety of different target atoms in DNA, including the N-1, N-3, and N-7 groups of adenine; the O2 and N-3 groups of cytosine; the N-3, O6, and N-7 groups of guanine; the O2, N-3, and O4 groups of thymine; and the phosphate groups of the DNA backbone (Singer, 1976; Sun and Singer, 1975). In practice, ENU usually causes point mutations (i.e., single base-pair changes) in the mouse O

O

N

C N

CH3

NH2

CH2

Ethyl group

Figure 15.1 Chemical structure of N-ethyl-N-nitrosourea (ENU).

300

Frank J. Probst and Monica J. Justice

germline, though small deletions are occasionally reported. A–T to T–A transversions are the most common base-pair changes, but all possible point mutations can occur ( Justice et al., 1999; Kohler et al., 1991). The fact that ENU is a point mutagen means that it not only generates null mutations (in which no functional gene product is produced) but also missense mutations and regulatory mutations that can have a variety of different effects on the gene product. These changes can lead to too little or too much protein, a less active or a superactive protein, misexpression of the gene in certain tissues, or even a completely new function for the protein. Thus, in contrast to most targeted mouse gene inactivations (commonly known as ‘‘knockout’’ mice), which have an ‘‘all or none’’ effect on the gene, ENU can produce an allelic series of mutations in a gene, each with a different phenotype, which can provide far more information about the function of a gene than a single null allele.

1.3. Recovering ENU-induced mutations with various breeding schemes ENU can be used to generate both dominant and recessive mutations, but different breeding schemes must be used to detect the different kinds of mutations. The detection of autosomal dominant phenotypes is relatively straightforward. Male mice are injected with ENU, allowed to recover their fertility, and then bred to healthy female mice (the G0 generation). As a result of the ENU treatment, the G0 male mice have germline mosaicism for a number of different ENU-induced mutations. Autosomal dominant mutations caused by the ENU will be apparent in the progeny of these animals (the G1 generation), so the G1 animals can be screened for phenotypes of interest, and mutant animals can be indentified and bred for further study (Fig. 15.2A). Large-scale mutagenesis programs for mutant phenotypes using this scheme have identified hundreds of new autosomal dominant mouse mutations (Hrabe de Angelis et al., 2000; Justice, 2000; Nolan et al., 2000). Genome-wide screens for autosomal recessive mutations are more laborious and require more animal breeding. ENU-mutagenized male mice are first bred to healthy female mice (the G0 generation). The progeny of this cross (the G1 generation) will be heterozygous carriers for any mutations caused by the ENU, so recessive mutations will not yield an abnormal phenotype in this generation. In order to breed these mutations to homozygosity, G1 animals (typically only the males, because they have greater breeding capacity) must be bred to healthy mice to produce G2 animals. If the G1 animal carries an autosomal recessive mutation, it will be transmitted to 50% of the G2 progeny. The G2 animals can then be either backcrossed to the original G1 animal or intercrossed to one another to produce the G3 generation. The G3 animals have the potential to be homozygous for any mutation induced by the ENU, so autosomal recessive mutations can yield a

301

Mouse Mutagenesis with ENU

A

B ENU mutagenesis

ENU mutagenesis ♂

G0

♀

X

G1

♂

X

X

♀

G3 X

D G2

♀

X

G2

ENU mutagenesis

G1

♂

G1

C

G0

G0

ENU mutagenesis G0

♂

X

♀

G1 G3

Figure 15.2 Breeding schemes for the isolation of various kinds of ENU-induced mutations. (A) In an autosomal dominant screen, a male mouse is mutagenized with ENU and bred to a healthy female mouse (the G0 generation). For each mouse, the two bars under the animal represent the two copies of any given autosome. A germline mutation in the ENU-mutagenized male is shown on one of the two chromosomes. If this mutation leads to an autosomal dominant phenotype (in this example, a white coat color), it can be detected in the G0 male’s progeny (the G1 generation). (B) Autosomal recessive screens require more breeding. G1 animals will be heterozygous for any ENU-induced mutations and will therefore not show an autosomal recessive phenotype. Mutations must be bred to homozygosity by breeding G1 animals (typically only the males, who can produce more progeny) to produce a second generation of animals (the G2 generation). Half of the G2 animals will have inherited

302

Frank J. Probst and Monica J. Justice

any ENU-induced mutation in the G1 animal. The G2 animals can then be crossed back to the original G1 animal (usually a father/daughter cross) or intercrossed to one another (a brother/sister cross) to produce G3 animals. ENU-induced autosomal recessive mutations will manifest themselves in this generation (shown here again as a white coat color). However, because only half of the G2 animals will inherit any given mutation from the G1 generation, some mutations can be lost if they are not transmitted to those G2 animals that are selected for further mating. Furthermore, embryonic lethal mutations will only be discovered if numerous G2 females are sacrificed and examined for abnormal embryos. (C) These problems can be overcome via an autosomal recessive screen using a balancer chromosome, which will trap all ENU-induced autosomal recessive mutations in a region of interest. In this example, the balancer chromosome (shown as a bar with arrowheads, indicating inversion breakpoints) is engineered to confer an autosomal dominant agouti coat color, shown here as a gray animal. In the G1 generation, only agouti-colored animals (who inherited the balancer from their mother) are selected for further breeding. These animals are bred to animals that are heterozygous for the balancer chromosome and also heterozygous for a second autosomal dominant marker on the opposite chromosome, such as the rex (Re) mutation, which produces wavy fur. Animals in the G2 generation with straight fur (indicating absence of the rex mutation) must have inherited a mutagenized chromosome from their father and a balancer chromosome from their mother. (In this example, the balancer chromosome has been engineered for embryonic lethality in the homozygous state, so this genotype is not seen among the live-born progeny.) Brother/sister matings between G2 animals with straight agouti fur can then be used to produce the G3 generation. Animals in this generation are either heterozygous for the balancer and a mutagenized segment of DNA for the region of interest (inherited en bloc from the G1 generation, since the balancer chromosome suppresses the recovery of recombinants that occur between the inversion breakpoints), or they are homozygous for a mutagenized segment of DNA. Any autosomal recessive phenotypes caused by mutations in the region of interest (shown here again as a white coat color) will be apparent in this generation. If all of the animals in this generation have an agouti coat color, then all of them must be heterozygous for the balancer chromosome. This indicates the presence of an autosomal recessive embryonic lethal mutation. Pregnant G2 animals can then be analyzed to determine precisely when and how such a mutation disrupts embryogenesis. (D) In a modifier screen, in contrast to the other types of screens, the G0 generation already has a known mutation of interest. In this example, both the males and females in the G0 generation are homozygous for a mutation that causes a white coat color (shown by the diamonds). When male animals are mutagenized and bred to female animals, all of the progeny will be homozygous for the original mutation of interest and should have a white coat color. If the ENU causes a new autosomal dominant mutation that modifies the white coat color phenotype (shown here as a gray animal), this will be detectable in the G1 generation. (Note that for simplicity, the modifier mutation shown here is on the same chromosome as the original mutation, but modifiers located anywhere in the genome can be isolated.) Screens for suppressor mutations may prove to be extremely valuable for understanding disease modification. Screens for enhancers exploit nonallelic noncomplementation, where a ‘‘sensitizing’’ mutation allows other mutations to be seen as semidominant phenotypes only in the presence of the first mutation. Such mutations may result in an autosomal recessive phenotype without the presence of the ‘‘sensitizing’’ mutation.

Mouse Mutagenesis with ENU

303

phenotype in this generation (Fig. 15.2B). Such screens have been valuable in dissecting genes and pathways required for embryonic development (Fernandez et al., 2009; Garcia-Garcia et al., 2005; Herron et al., 2002; Kasarskis et al., 1998; Zohn et al., 2005). However, one of the major challenges encountered in these genome-wide screens for autosomal recessive mutations is the isolation and maintenance of embryonic lethal mutations. Since homozygous mutants will not be seen in the live-born progeny, detecting such mutations usually requires dissecting of a number of pregnant G2 female mice and observing that a fraction of the embryos are either malformed or deceased in utero. Once this observation has been made, numerous additional animals must be bred to further analyze and maintain the mutation. The breeding scheme for isolating embryonic lethal autosomal recessive mouse mutations can be made much more efficient by restricting the analysis to certain chromosomal regions through the use of balancer chromosomes. A balancer chromosome is an engineered construct that ideally has three traits: (1) one or more DNA inversions to suppress the recovery of recombination products when a recombination event occurs within the inversion during meiosis, (2) a dominant phenotype that allows for visual genotyping of animals inheriting the balancer chromosome in subsequent generations, and (3) either an autosomal recessive lethal mutation or an autosomal recessive mutation with an obvious phenotype to remove animals from the population that are homozygous for the balancer chromosome (Hentges and Justice, 2004). The introduction of balancer chromosomes into a mutagenesis breeding scheme allows for the capture of all induced mutations that are linked to the chromosomal region of interest, even if these mutations produce a lethal phenotype (Fig. 15.2C) (Kile et al., 2003). Embryos harboring these mutations can then be analyzed to determine precise effect of the mutation on embryonic development (Boles et al., 2009a,b; Hentges et al., 2006). If the desire is to capture autosomal recessive alleles at a specific locus, the search can be accomplished in two different ways. First, an ENUmutagenized male mouse can be bred to a female mouse carrying a known mutation at the locus of interest (as either a heterozygote or a homozygote), and the offspring can be screened for animals that inherited the known mutation from the mother and a new ENU-induced mutation from the father (Cordes and Barsh, 1994; Ebersole et al., 1996; Shumacher et al., 1996). Second, sequence analysis of the locus of interest can be performed on a panel of genomic DNAs from the G1 males of a large number of ENU-treated fathers. Sperm samples from all of these males have been isolated and stored, so when a heterozygous mutation is detected in the panel, the sperm sample can be requested, and the mutation can be recovered and bred to homozygosity (Quwailid et al., 2004).

304

Frank J. Probst and Monica J. Justice

A more recent and highly promising use of ENU mutagenesis in the mouse is the generation of mutations that modify a phenotype produced by a well-characterized mutant locus. This type of experiment is known as a modifier screen. All such screens involve mutagenizing and/or breeding to mice that already carry a mutation of interest, with the goal of the screen being to identify mutations that alter the original mutant phenotype (Fig. 15.2D). The first published example of such a screen in the mouse was for suppressors of thrombocytopenia (low numbers of platelets). Male mice homozygous for a null allele in the gene for the thrombopoietin receptor (Mpl / ), which causes thrombocytopenia, were mutagenized with ENU and then bred to Mpl / females. Platelet counts were performed on progeny of this cross to establish several dominant mutant lines that carried a presumed mutation that suppressed the thrombocytopenia phenotype. Further work revealed that two of these lines had inherited hypomorphic mutations in the Myb gene (Carpinelli et al., 2004). A third line had inherited a mutation in the gene for p300, which is known to interact with c-Myb (Kauppi et al., 2008). Subsequent screens have identified modifiers of a GFP transgene with variegated expression (Blewitt et al., 2005), Delta1-dependent Notch signaling (Rubio-Aliaga et al., 2007), neural crest cell development (Buac et al., 2008; Matera et al., 2008), and the growth hormone and TGF-beta signaling pathways (Mohan et al., 2008). Clearly, ENU mutagenesis in the mouse has been a critical tool in the generation of new, biologically important mouse mutants, but it is still a relatively new research tool and will undoubtedly produce many more interesting animal models in the years to come. With the recent advances that have been made in both DNA sequence capture technology and nextgeneration sequencing, coupled with the availability of a mouse reference genome, it is now easier than ever to identify the specific DNA lesions that have been induced by ENU and are yielding phenotypes of interest.

2. Materials and Methods 2.1. ENU preparation Note: ENU is both mutagenic and carcinogenic, and it should be handled with extreme caution. It should only be used inside an efficient chemical hood, and all handlers should wear gloves, lab coats, and masks. Fortunately, ENU has a very short half-life (a few minutes) in alkaline conditions. In the event of a chemical spill, the area should be flooded with 0.1 M KOH to inactivate the ENU. ENU is very sensitive to light, humidity, and pH. It should therefore be prepared fresh prior to each use, and it should be used within 3 h of preparation.

Mouse Mutagenesis with ENU

305

1. Prepare phosphate/citrate buffer: 0.1 M dibasic sodium phosphate, 0.05 M sodium citrate, adjust the pH to 5.0 with phosphoric acid. Filter sterilize using a 0.45 mm filter. 2. Remove the 1 g ENU ISOPAC (Sigma-Aldrich catalog number N3385) from the freezer and allow it to warm to room temperature in the dark. 3. In an efficient chemical hood, insert an 18-gauge needle into the ENU bottle as a vent. Using a second needle, inject the bottle with 10 ml of 95% ethanol. Remove both needles and gently agitate the bottle until the ENU goes into solution. When completely dissolved, the solution will be yellow but clear. It can take up to 10 min for the ENU to dissolve completely. 4. Reinsert the 18-gauge needle into the ENU bottle as a vent and use a second needle to add 90 ml of phosphate/citrate buffer. Mix thoroughly. Note: If the amount of ENU to be injected into each animal is small, then 5 ml of concentrated ENU can be removed from the stock bottle prior to adding the phosphate/citrate buffer, and the remaining 5 ml can be diluted with 95 ml of phosphate/citrate buffer. If concentrated ENU is removed from the stock bottle at this step, it should be inactivated by adding it to at least 50 ml of 0.1 M KOH and exposing it to light for at least 24 h. 5. Dilute 400 l of the ENU solution with 1.6 ml phosphate/citrate buffer (a 1:5 dilution) in a disposable plastic cuvette. Set up a blank containing 40 l of 95% ethanol and 1.96 ml phosphate/citrate buffer in a similar disposable plastic cuvette. Measure the OD398 nm of the ENU solution on a spectrophotometer. An OD398nm of 0.72 corresponds to an ENU concentration of 1 mg/ml. Therefore, you should calculate the concentration (in mg/ml) of your ENU stock solution by multiplying by 5 (the dilution factor) and dividing by 0.72. Note: Determining the concentration of ENU in the stock vial is very important, because 1 g vials can contain from 0.7 to 1.3 g of ENU.

2.2. Injecting the animals Note: The optimum dose of ENU varies significantly among mouse strains. Optimum doses have previously been determined for a number of different inbred mouse strains. For C57BL/6J animals, the optimum dose is a total of 300 mg/kg given in three 100 mg/kg fractions administered at 1-week intervals. Animals should be 8–12 weeks of age at the time of the first injection. C57BL/6J animals will have a sterile period lasting 90–105 days from the time of the last injection. The length of the sterile period, like the optimum ENU dose, also varies from strain to strain (Davis et al., 1999;

306

Frank J. Probst and Monica J. Justice

Weber et al., 2000). Researchers working with a previously unstudied strain of mice will need to optimize the dose of ENU for that particular strain. 1. Weigh each animal immediately prior to each injection. Animals tend to lose weight after ENU treatment, so it is particularly important to reweigh animals when giving fractionated doses. 2. The appropriate ENU dose for the animal depends on (a) the ENU concentration in the stock bottle, as calculated above, (b) the weight of the animal, and (c) the strain of animals being used. The appropriate dose should be given to each animal intraperitoneally using a 1 cc tuberculin syringe with a 26-gauge, 3/8 in. needle. Animals should then be placed in a fresh cage with clean bedding. The animals may appear wobbly for about 30 min after the injection, due to the alcohol content of the injection. 3. Keep the mice in an efficient chemical hood for at least 24 h after each injection. Do not change or handle any of the bedding during this time frame.

2.3. Inactivating the ENU 1. When the injections are complete, vent the ENU stock bottle with an 18 gauge needle and fill the bottle with 0.1 M KOH with a second needle. Leave the bottle in an efficient chemical hood, exposed to light, for at least 24 h. 2. Treat all equipment and gloves that came in contact with ENU with 0.1 M KOH. This includes rinsing all needles and syringes with 0.1 M KOH. 3. Discard the contents of the ENU stock bottle into a liquid chemical waste container, rinse the bottle with water, and again discard the contents into a liquid chemical waste container. The bottle should be discarded into a waste container approved for glass. The remainder of the solid waste should be discarded in the appropriate waste containers, including sharps containers for the needles and syringes.

2.4. Breeding the mutagenized males ENU-treated male mice typically go through a period of infertility that lasts for several weeks after the last ENU injection, and some animals will never recover their fertility. For this reason, it is strongly advised that all mutagenized animals be tested for fertility by placing them in cages with multiple female mice approximately 8–10 weeks after the last ENU injection. Only males that successfully impregnate a female should be used for further experiments.

Mouse Mutagenesis with ENU

307

Fertile males can then be introduced into a breeding scheme that is appropriate for the kind of mutation that is being screened for (Fig. 15.2). Mutagenized males are expected to have germline mosaicism for many different ENU-induced mutations but may carry repeat mutations or clusters, because ENU is a premeiotic mutagen. In order to avoid oversampling the gametes of a single mouse (and isolating the same induced mutation multiple times), males should be discarded after a certain number of gametes have been sampled (50 offspring per mutagenized male for dominant mutation screens, and 30 offspring per mutagenized male for recessive mutation screens). To generate these offspring as efficiently as possible, rotation matings are strongly recommended. In rotation matings, each male is placed with two new females at the beginning of each week. Mouse gestation takes approximately 3 weeks, and mice can be weaned at 3 weeks of age, for a total of 6 weeks from the time of conception to the time of weaning. Thus, we recommend rotating each mutagenized male through seven sets of females at 1-week intervals, and then starting this whole cycle again at the end of 7 weeks. Female breeders should be retired at 6–9 months of age and replaced with younger animals. Mutagenized males will have a markedly reduced lifespan as a result of the ENU treatment, and if rotation matings are not used, it is very likely that the male will die prior to producing the desired number of offspring.

2.5. Screening for abnormal phenotypes Like the breeding scheme, the phenotype screening that is performed will vary from experiment to experiment, depending on the overall goals of the research. Phenotypes that produce a visible abnormality are obviously the easiest to detect and are seen in virtually all ENU mutagenesis screens. More subtle phenotypes must be ascertained with more sophisticated analyses, such as biochemical testing of blood and urine, imaging, histologic analysis, or behavioral testing. The ideal phenotype screen is inexpensive, rapid, easy to perform, and reproducible. Mice that are scored as abnormal in the first round of screening can be assessed with further testing or immediately bred to assess for transmission of the phenotype to later generations. For a review of the challenges of accurate phenotyping, see Justice (2008) and Brown et al. (2009). Even the most rugged phenotyping screens will sometimes identify animals that fail to transmit the phenotype to successive generations. There are several possible explanations for this phenomenon. First and foremost, the mutant phenotype may represent a simple outlier as opposed to a true mutant phenotype. Second, the phenotype may have been the result of an epigenetic change rather than a base-pair change, as ENU can yield epigenetic changes in the genome (including DNA methylation and histone modifications) as well as point mutations. Third, the phenotype may

308

Frank J. Probst and Monica J. Justice

be penetrant on some genetic backgrounds but not on others. This should always be a consideration if potential mutants are immediately bred to a different mouse strain to initiate genetic mapping of the mutation. Absence of the phenotype in such a cross does not confirm that a particular mutant phenotype is not heritable—if this occurs, the original abnormal animal should be crossed to an animal with the same genetic background to determine if the phenotype can be transmitted on the original background. If this is the case, then special consideration must be given to the particular strains that are subsequently used to maintain and map the mutation.

2.6. Mapping mutant phenotypes and identifying mutations Once a particular phenotype has been identified and confirmed to be heritable, the underlying mutation responsible for the phenotype must be determined. This is typically done by the traditional positional cloning approach of breeding the mutation to a different strain (or strains) of mice and analyzing the offspring to determine which genetic markers consistently segregate with the abnormal phenotype. Candidate genes in the region of interest can then be sequenced to identify the causative mutation. This was previously a laborious process that could take years, but the completion of the mouse genome, the availability of single nucleotide polymorphism (SNP) panels, and the falling cost of DNA sequencing have allowed most mutations to now be identified over a period of weeks to months of focused research (Fig. 15.3). In the past, candidate gene analysis typically progressed by sequencing gene after gene until a causative mutation was identified. More recently, as DNA sequencing has become less and less expensive, it has become cost-effective to simply sequence all of the exons in the critical interval harboring the mutation (Boles et al., 2009b). In the near future, it is likely that DNA capture technology and next-generation sequencing will allow researchers to simply sequence the entire genomic interval of interest to identify the mutations responsible for their mutant phenotypes.

3. Conclusion The extreme mutagenicity of ENU on mouse gametes was not discovered until the late 1970s (Russell et al., 1979), and the first large-scale genome-wide mouse ENU mutagenesis screens for autosomal dominant mutations were not published until 2000 (Hrabe de Angelis et al., 2000; Justice, 2000; Nolan et al., 2000). Since then, there has been an explosion in research using ENU mutagenesis in the mouse, with the publication of genome-wide screens for autosomal recessive mutations (Fernandez et al., 2009; Garcia-Garcia et al., 2005; Herron et al., 2002; Kasarskis et al., 1998;

A

F1

X

F2

X

Nonrecombinant progeny

Recombinant progeny

B Nonrecombinant interval

Gene 1

Gene 2

Gene 3

Gene 4

Gene 5

Classic technique

Modern technique

Sequence individual candidate genes exon-by-exon until the mutation is identified

Capture and sequence the entire nonrecombinant interval

Figure 15.3 Mapping and identifying mutations of interest. In order to identify the mutation responsible for a particular phenotype, the mutation must first be localized to a specific segment of the genome. This is most often done with a mapping cross (A). This example shows how this is accomplished for an autosomal recessive mutation causing a white coat color. Mice homozygous for the mutation are bred to a genetically different strain of mice to produce F1 animals. Each F1 animal will have inherited one of each pair of chromosomes from the mutant animal and the other from the other strain of mice. They will therefore be heterozygous for both the mutation and for numerous genetic markers scattered throughout the genome. When F1 animals are intercrossed to one another, they produce the F2 generation. Animals in the F2 generation can then be analyzed to determine which genetic markers are being coinherited with the mutation. This is known as a genome scan. Once a particular segment of the genome has been identified, the recombinant progeny—animals that inherited a meiotic crossover event very close to the mutation—can be used to determine the narrowest possible region that the mutation must lie in, which is known as the nonrecombinant interval (B). Once a nonrecombinant interval has been established, the genes in the interval can be identified by examining the published sequence of the mouse genome, typically by using one of the publicly available genome browsers, such as the UCSC Genome Browser or the Ensembl Genome Browser. The classic technique for identifying a mutation was to design polymerase chain reaction (PCR) primers to the exons of each gene in the region and then sequence the PCR products. This was typically done one gene at a time, until the mutation was identified. In the near future, researchers will likely use sequence capture technology to isolate the DNA from a mutant animal that covers the nonrecombinant interval, followed by next-generation sequencing to establish the sequence of the entire nonrecombinant interval.

310

Frank J. Probst and Monica J. Justice

Zohn et al., 2005), screens for autosomal recessive mutations linked to balancer chromosomes (Boles et al., 2009a; Hentges et al., 2006; Kile et al., 2003), and screens for dominant modifiers of other genetic mutations (Blewitt et al., 2005; Buac et al., 2008; Carpinelli et al., 2004; Kauppi et al., 2008; Matera et al., 2008; Mohan et al., 2008; Rubio-Aliaga et al., 2007). A large number of mouse ENU mutagenesis screens are currently ongoing, many of which have generated very promising preliminary results, so we anticipate a bright future for the role of ENU-induced mouse mutations in biological research in the years to come.

ACKNOWLEDGMENTS Frank J. Probst, M.D., Ph.D., holds a Career Award for Medical Scientists from the Burroughs Wellcome Fund. Monica J. Justice, Ph.D., is supported by the Rett Syndrome Research Trust and the National Institutes of Health R01 CA115503.

REFERENCES Blewitt, M. E., Vickaryous, N. K., Hemley, S. J., Ashe, A., Bruxner, T. J., Preis, J. I., Arkell, R., and Whitelaw, E. (2005). An N-ethyl-N-nitrosourea screen for genes involved in variegation in the mouse. Proc. Natl. Acad. Sci. USA 102, 7629–7634. Boles, M. K., Wilkinson, B. M., Maxwell, A., Lai, L., Mills, A. A., Nishijima, I., Salinger, A. P., Moskowitz, I., Hirschi, K. K., Liu, B., Bradley, A., and Justice, M. J. (2009a). A mouse chromosome 4 balancer ENU-mutagenesis screen isolates eleven lethal lines. BMC Genet. 10, 12. Boles, M. K., Wilkinson, B. M., Wilming, L. G., Liu, B., Probst, F. J., Harrow, J., Grafham, D., Hentges, K. E., Woodward, L. P., Maxwell, A., Mitchell, K., Risley, M. D., et al. (2009b). Discovery of candidate disease genes in ENU-induced mouse mutants by large-scale sequencing, including a splice-site mutation in nucleoredoxin. PLoS Genet. 5, e1000759. Brown, S. D., Wurst, W., Kuhn, R., and Hancock, J. M. (2009). The functional annotation of mammalian genomes: The challenge of phenotyping. Annu. Rev. Genet. 43, 305–333. Buac, K., Watkins-Chow, D. E., Loftus, S. K., Larson, D. M., Incao, A., Gibney, G., and Pavan, W. J. (2008). A Sox10 expression screen identifies an amino acid essential for Erbb3 function. PLoS Genet. 4, e1000177. Carpinelli, M. R., Hilton, D. J., Metcalf, D., Antonchuk, J. L., Hyland, C. D., Mifsud, S. L., Di Rago, L., Hilton, A. A., Willson, T. A., Roberts, A. W., Ramsay, R. G., Nicola, N. A., et al. (2004). Suppressor screen in Mpl / mice: c-Myb mutation causes supraphysiological production of platelets in the absence of thrombopoietin signaling. Proc. Natl. Acad. Sci. USA 101, 6553–6558. Cordes, S. P., and Barsh, G. S. (1994). The mouse segmentation gene kr encodes a novel basic domain-leucine zipper transcription factor. Cell 79, 1025–1034. Davis, A. P., Woychik, R. P., and Justice, M. J. (1999). Effective chemical mutagenesis in FVB/N mice requires low doses of ethylnitrosourea. Mamm. Genome 10, 308–310. Ebersole, T. A., Chen, Q., Justice, M. J., and Artzt, K. (1996). The quaking gene product necessary in embryogenesis and myelination combines features of RNA binding and signal transduction proteins. Nat. Genet. 12, 260–265.

Mouse Mutagenesis with ENU

311

Ehling, U. H. (1978). Specific-locus mutations in mice. In ‘‘Chemical Mutagens: Principles and Methods for Their Detection, Vol. 5,’’ (A. Hollaender and F. J. de Serres, eds.), pp. 233–256. Plenum Press, New York. Ehling, U. H., and Neuhauser, A. (1979). Procarbazine-induced specific-locus mutations in male mice. Mutat. Res. 59, 245–256. Ehling, U. H., and Neuhauser-Klaus, A. (1989). Induction of specific-locus mutations in male mice by ethyl methanesulfonate (EMS). Mutat. Res. 227, 91–95. Fernandez, L., Marchuk, D. A., Moran, J. L., Beier, D. R., and Rockman, H. A. (2009). An N-ethyl-N-nitrosourea mutagenesis recessive screen identifies two candidate regions for murine cardiomyopathy that map to chromosomes 1 and 15. Mamm. Genome 20, 296–304. Garcia-Garcia, M. J., Eggenschwiler, J. T., Caspary, T., Alcorn, H. L., Wyler, M. R., Huangfu, D., Rakeman, A. S., Lee, J. D., Feinberg, E. H., Timmer, J. R., and Anderson, K. V. (2005). Analysis of mouse embryonic patterning and morphogenesis by forward genetics. Proc. Natl. Acad. Sci. USA 102, 5913–5919. Hentges, K. E., and Justice, M. J. (2004). Checks and balancers: Balancer chromosomes to facilitate genome annotation. Trends Genet. 20, 252–259. Hentges, K. E., Nakamura, H., Furuta, Y., Yu, Y., Thompson, D. M., O’Brien, W., Bradley, A., and Justice, M. J. (2006). Novel lethal mouse mutants produced in balancer chromosome screens. Gene Expr. Patterns 6, 653–665. Herron, B. J., Lu, W., Rao, C., Liu, S., Peters, H., Bronson, R. T., Justice, M. J., McDonald, J. D., and Beier, D. R. (2002). Efficient generation and mapping of recessive developmental mutations using ENU mutagenesis. Nat. Genet. 30, 185–189. Hitotsumachi, S., Carpenter, D. A., and Russell, W. L. (1985). Dose-repetition increases the mutagenic effectiveness of N-ethyl-N-nitrosourea in mouse spermatogonia. Proc. Natl. Acad. Sci. USA 82, 6619–6621. Hrabe de Angelis, M. H., Flaswinkel, H., Fuchs, H., Rathkolb, B., Soewarto, D., Marschall, S., Heffner, S., Pargent, W., Wuensch, K., Jung, M., Reis, A., Richter, T., et al. (2000). Genome-wide, large-scale production of mutant mice by ENU mutagenesis. Nat. Genet. 25, 444–447. Justice, M. J. (2000). Capitalizing on large-scale mouse mutagenesis screens. Nat. Rev. Genet. 1, 109–115. Justice, M. J. (2008). Removing the cloak of invisibility: Phenotyping the mouse. Dis. Models Mech. 1, 109–112. Justice, M. J., Noveroske, J. K., Weber, J. S., Zheng, B., and Bradley, A. (1999). Mouse ENU mutagenesis. Hum. Mol. Genet. 8, 1955–1963. Kasarskis, A., Manova, K., and Anderson, K. V. (1998). A phenotype-based screen for embryonic lethal mutations in the mouse. Proc. Natl. Acad. Sci. USA 95, 7485–7490. Kauppi, M., Murphy, J. M., de Graaf, C. A., Hyland, C. D., Greig, K. T., Metcalf, D., Hilton, A. A., Nicola, N. A., Kile, B. T., Hilton, D. J., and Alexander, W. S. (2008). Point mutation in the gene encoding p300 suppresses thrombocytopenia in Mpl / mice. Blood 112, 3148–3153. Kile, B. T., Hentges, K. E., Clark, A. T., Nakamura, H., Salinger, A. P., Liu, B., Box, N., Stockton, D. W., Johnson, R. L., Behringer, R. R., Bradley, A., and Justice, M. J. (2003). Functional genetic analysis of mouse chromosome 11. Nature 425, 81–86. Kohler, S. W., Provost, G. S., Fieck, A., Kretz, P. L., Bullock, W. O., Sorge, J. A., Putman, D. L., and Short, J. M. (1991). Spectra of spontaneous and mutagen-induced mutations in the lacI gene in transgenic mice. Proc. Natl. Acad. Sci. USA 88, 7958–7962. Matera, I., Watkins-Chow, D. E., Loftus, S. K., Hou, L., Incao, A., Silver, D. L., Rivas, C., Elliott, E. C., Baxter, L. L., and Pavan, W. J. (2008). A sensitized mutagenesis screen identifies Gli3 as a modifier of Sox10 neurocristopathy. Hum. Mol. Genet. 17, 2118–2131.

312

Frank J. Probst and Monica J. Justice

Mohan, S., Baylink, D. J., and Srivastava, A. K. (2008). A chemical mutagenesis screen to identify modifier genes that interact with growth hormone and TGF-beta signaling pathways. Bone 42, 388–395. Nolan, P. M., Peters, J., Strivens, M., Rogers, D., Hagan, J., Spurr, N., Gray, I. C., Vizor, L., Brooker, D., Whitehill, E., Washbourne, R., Hough, T., et al. (2000). A systematic, genome-wide, phenotype-driven mutagenesis programme for gene function studies in the mouse. Nat. Genet. 25, 440–443. Quwailid, M. M., Hugill, A., Dear, N., Vizor, L., Wells, S., Horner, E., Fuller, S., Weedon, J., McMath, H., Woodman, P., Edwards, D., Campbell, D., et al. (2004). A gene-driven ENU-based approach to generating an allelic series in any gene. Mamm. Genome 15, 585–591. Rubio-Aliaga, I., Soewarto, D., Wagner, S., Klaften, M., Fuchs, H., Kalaydjiev, S., Busch, D. H., Klempt, M., Rathkolb, B., Wolf, E., Abe, K., Zeiser, S., et al. (2007). A genetic screen for modifiers of the delta1-dependent notch signaling function in the mouse. Genetics 175, 1451–1463. Russell, W. L. (1951). X-ray-induced mutations in mice. Cold Spring Harb. Symp. Quant. Biol. 16, 327–336. Russell, W. L., and Kelley, E. M. (1979). Ineffectiveness of diethylnitrosamine in the induction of specific-locus mutations in mice. Genetics 91(Suppl.), s109–s110. Russell, W. L., Russell, L. B., and Kelly, E. M. (1958). Radiation dose rate and mutation frequency. Science 128, 1546–1550. Russell, W. L., Kelly, E. M., Hunsicker, P. R., Bangham, J. W., Maddux, S. C., and Phipps, E. L. (1979). Specific-locus test shows ethylnitrosourea to be the most potent mutagen in the mouse. Proc. Natl. Acad. Sci. USA 76, 5818–5819. Russell, W. L., Hunsicker, P. R., Carpenter, D. A., Cornett, C. V., and Guinn, G. M. (1982a). Effect of dose fractionation on the ethylnitrosourea induction of specific-locus mutations in mouse spermatogonia. Proc. Natl. Acad. Sci. USA 79, 3592–3593. Russell, W. L., Hunsicker, P. R., Raymer, G. D., Steele, M. H., Stelzner, K. F., and Thompson, H. M. (1982b). Dose–response curve for ethylnitrosourea-induced specificlocus mutations in mouse spermatogonia. Proc. Natl. Acad. Sci. USA 79, 3589–3591. Shumacher, A., Faust, C., and Magnuson, T. (1996). Positional cloning of a global regulator of anterior–posterior patterning in mice. Nature 383, 250–253. Singer, B. (1976). All oxygens in nucleic acids react with carcinogenic ethylating agents. Nature 264, 333–339. Sun, L., and Singer, B. (1975). The specificity of different classes of ethylating agents toward various sites of HeLa cell DNA in vitro and in vivo. Biochemistry 14, 1795–1802. Weber, J. S., Salinger, A., and Justice, M. J. (2000). Optimal N-ethyl-N-nitrosourea (ENU) doses for inbred mouse strains. Genesis 26, 230–233. Zohn, I. E., Anderson, K. V., and Niswander, L. (2005). Using genomewide mutagenesis screens to identify the genes required for neural tube closure in the mouse. Birth Defects Res. A Clin. Mol. Teratol. 73, 583–590.

C H A P T E R

S I X T E E N

Phenotype-Driven Mouse ENU Mutagenesis Screens Tamara Caspary Contents 314 315 316 316 317 317 319 321 321 322 323 323 324 325 326 326 326

1. Introduction 2. Screen Design 3. Screen Execution 3.1. Mutagenesis 3.2. Materials 3.3. Procedure 3.4. Breeding 3.5. Screening 3.6. Mutants! 3.7. Establishing independent lines 4. Gene Identification 4.1. Mapping 4.2. Sequencing 4.3. Proving the variant is causal 4.4. Cryopreservation of lines 5. The Future for Mouse Forward Genetics References

Abstract In the past decade, forward genetic screens in the mouse have come into their own as a practical method for exploring the genetic basis of many biological processes. By looking directly for disruption in a process of interest, genetic screens have always been powerful, but completion of the genome sequence has made mouse forward genetic screens practical, as well. The sequenced genome means we can map and sequence more efficiently than before, so small focused screens are now within the reach of even small labs. N-Ethyl-Nnitrosourea (ENU) is the preferred mutagen in forward genetic screens, because it is extremely potent in the premeiotic male germ line, where it induces point mutations. This last point is crucial, as point mutations lead to all classes of mutations (e.g., null, hypomorphs, neomorphs, antimorphs, and hypermorphs), Department of Human Genetics, Emory University School of Medicine, Atlanta, Georgia, USA Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77016-6

#

2010 Elsevier Inc. All rights reserved.

313

314

Tamara Caspary

which is why forward genetic screens can implicate a gene in a particular process when a targeted deletion may not. Point mutations often mimic human disease states, yielding highly relevant animal models. Since mammals reproduce, lactate, behave, develop, and protect themselves from infection differently from other vertebrates, mammalian forward genetic screens are uniquely informative. In fact, in the past decade, forward genetics has uncovered mutations demonstrating that certain genes exist only in mammals, that specific mechanisms function only in mammals, and that particular biological processes may exist only in mammals; hence screens focused on these processes have identified unsuspected genes. As powerful as the approach is, forward genetics remains a method for the committed; the process of screening requires organization and tenacity. This chapter is intended to help those who are ready to make the commitment by providing practical advice. To this end I detail the issues surrounding screen design and screen execution, as well as mutation identification and confirmation.

1. Introduction Forward genetic screens using potent alkylating agents such as Nethyl-N-nitrosourea (ENU) have uncovered fundamental principles of biology in many model organisms, including the laboratory mouse, which joined the growing list in the past decade. While chemically induced mutants have been around for some time, the process of creating them was so laborious that it was usually limited to experts at large facilities like Harwell, Oak Ridge National Laboratory, or The Jackson Labs. Completion of the mouse genome sequence, however, has transformed forward genetic screens, making gene identification far more straightforward; this in turn has resulted in small labs being able to perform focused screens for mutants that disrupt whatever biological process they study. While still somewhat the domain of experts, the technological advances in targeted resequencing (also called next-generation sequencing) promise to further facilitate gene identification, enabling any lab, no matter their level of expertise, to use ENU mutagenesis to uncover novel genes involved in the fundamental biology in which they are interested. The purpose of this chapter is to guide you through the process. Because screen design is critical, I will begin by discussing the decisions that need to be made prior to mutagenizing any mice; good decisions at this stage will accelerate downstream steps in the screen enough that the time spent upfront is well justified. Next, I will go into detail on each step involved in executing a screen: mutagenesis, breeding, screening, finding mutants, and establishing independent lines. Finally, I will talk about gene identification: the mapping and sequencing, followed by the methods you can use to verify a causative mutation.

Mouse ENU Screens

315

2. Screen Design First and most critically, you must clearly define the phenotype of interest and choose the phenotype for which you can screen; these may or may not be identical. At a practical level, the screen itself must be simple and reproducible enough that it can achieve ‘‘repetitive consistency,’’ meaning that different individuals in your lab should be able to perform the same assay in different lines and agree as to whether or not there is a phenotype. For purely practical reasons, this might mean that you decide to look initially for a general phenotype, and then perform a secondary screen only in lines that meet the primary screen criteria. Intrinsic to this decision, also, is whether you will screen for dominant or recessive traits. Many factors can influence the phenotype for which you decide to screen. For example, the viability and fertility of the mutants determines how you will establish the line to recover the mutation of interest. If the mutants are inviable or sterile, you will breed the parents to propagate the mutation. Furthermore, the phenotype must be robust, meaning it cannot be within the normal variation seen in wild-type animals. Phenotypes that do not cross such a threshold will result in false-positive lines. This issue is especially relevant as you decide which mouse strain to mutagenize and what mouse strain you will use in the mapping backcross; the mapping backcross identifies where the mutation must lie and depends on polymorphisms between the two strains being used. Thus, the mapping strain should be chosen to maximize genotypic divergence while minimizing phenotypic variation. Cost may also influence your choice of phenotype. Any screen that can be performed prior to weaning the pups reduces cost considerably. Obviously, it is not possible to screen for adult-onset phenotypes in such young mice, but consider whether a marker can be used that anticipates the adult phenotype; you might be able to screen initially for aberrance of the marker, and then screen only those lines for the actual phenotype. Markers can be useful in many screens, especially since there are so many genetically modified lines with fluorescently tagged proteins. If you are using a reporter line, you should mutagenize a distinct strain and cross the mutagenized mice to the reporter line so that you incorporate the marker in the screen and do not segregate it away from mutations of interest. When using previously generated lines, two issues tend to arise: genetic background and husbandry. Many genetically modified lines are on mixed genetic backgrounds, which can complicate mapping, although this is less of a problem now that the single nucleotide polymorphism (SNP) databases for individual mouse strains are denser. Genetic background will come to matter even less as targeted resequencing matures. For now, however, it

316

Tamara Caspary

remains a consideration. As to husbandry, if you use a genetically modified line in the screen, you will have to perform the breeding to supply the screen yourself, increasing the need for cage space and complicating coordination. This will also be the case when using a line to sensitize the background to generate a robust visible phenotype. The success of your screen depends in large measure on the decisions you make as you ponder all these issues before starting the mutagenesis. Note that these are general guidelines and considerations; for your screen, details such as drawing out each cross and calculating the number of mice and cages that will be needed for each cross will reveal the particulars that your screen requires to be successful.

3. Screen Execution In performing the actual screen, the overall scheme involves injecting male mice with ENU, waiting about 12 weeks while they become sterile and regain fertility, breeding the mice one or two generations (depending on whether you are looking for dominant or recessive phenotypes), performing a mapping backcross for chromosomal and fine mapping of the mutation, and finally sequencing candidate genes in a defined interval for potential causative mutations. Below I describe each step in detail.

3.1. Mutagenesis ENU is preferred in mouse forward genetics because it is so effective and so well documented (Beier, 2000; Chen et al., 2000; Coghill et al., 2002; Concepcion et al., 2004; Gondo et al., 2009; Russell et al., 1979). The most effective doses have been empirically determined and published for many strains and hybrid backgrounds ( Justice et al., 2000; Weber et al., 2000). In general, the data show that multiple small doses are more effective than a single large dose, although plenty of investigators have used a single dose successfully (Hitotsumachi et al., 1985). In my lab, we use the C57/BL6 strain because its genome is fully sequenced, facilitating gene identification later. However, other strains are fine, since you can always sequence to validate ENU-induced changes. The number of mice you need to mutagenize depends on how many lines you hope to screen, since many mutagenized mice will not regain fertility or will die before mating. As a rule of thumb, we typically inject twice as many males as we anticipate breeding. It is crucial that the males are mutagenized as they become fertile (8–10 weeks), so that adequate numbers of offspring can be recovered and all the induced mutations screened.

Mouse ENU Screens

317

Before you even begin, it is critical to get the paperwork taken care of. ENU is a potent mutagen, (Chemical Abstract Number (CAS) registry # 759-73-9), so you will need to obtain institutional safety approval in addition to animal protocol approval. Each mutagenesis session will generate several types of waste that will need to be disposed of appropriately. These include liquid chemical waste (
300 ml of inactivated ENU solution), solid waste (one large biohazard trash bag per injection day), and animal cage waste (one large biohazard bag per injection day). Once all approvals are in place, gather the following items and follow the detailed protocol below (modified from Justice, 2000; McDonald and Beier, 2003; Salinger and Justice, 2008).

3.2. Materials Chemical fume hood Signs reserving the hood for at least 24 h on each injection day Disposable lab coats, gloves, and masks Aluminum foil Plastic bags (large biohazard bags for trash and small for solid waste) Glass liquid waste container Scale for weighing mice Syringes (plastic, 1, 10, and 60 cc) 95% nondenatured ethanol Phosphate/citrate buffer: [100 mM dibasic sodium phosphate, 50 mM sodium citrate (14.2 g Na2HPO4 and 14.7 Na3C6H5O7 g in 1 l H2O), adjust pH to 5.0, sterilize, store indefinitely] N-Ethyl-N-nitrosourea (Sigma N3385, 1 g ISOPAC container; store in freezer until use) One of the following inactivating solutions: Potassium hydroxide: [0.1 M KOH (5.6 g KOH in 1 l of H2O)] or Alkaline sodium thiosulfate: [0.1 M NaOH (4 g/l), 20% sodium thiosulfate (200 g/l)] Squirt or spray bottle for inactivating solution Mice: males of appropriate strain at 8 weeks of age, acclimated to mouse room for at least 1 week

3.3. Procedure Two people are needed on injection days for the procedure to run smoothly. One person should be well qualified to perform the intraperitoneal (i.p.) injections of ENU.

318

Tamara Caspary

1. Tag each male and record weight to the nearest gram. 2. Calculate and record how much ENU solution (injection volume, IV) will be injected into each mouse through the following formula: IV ðmlÞ ¼

ðfinal concentration of ENUÞ  ðmouse body weightÞ 10 mg=ml ENU

For example, for a final concentration of 100 mg/kg ENU in a 20-g mouse: IV ðmlÞ ¼

ð0:1 mg=gÞ  ð20 gÞ ¼ 0:2 ml 10 mg=ml ENU

3. Using a 10-cc syringe and a 16-gauge needle, inject 10 ml of 95% ethanol in the manufacturer’s bottle containing the ENU. The ENU should dissolve immediately. If it does not, gently swirl the bottle in the warmth of your hands until it does. 4. Place an 18-gauge needle in the top of the ISOPAC bottle for venting. Then use a 60-cc syringe with an 18-gauge needle to inject 90 ml of phosphate/citrate buffer. Mix thoroughly. (Note: Many protocols would now have you determine the concentration of the ENU in solution using spectrophotometry. While this undoubtedly gives an accurate measurement, we have found it is not necessary and calculate the concentration based on the 1 g label on the vial. Given the hazards of handling ENU, as well as its extreme sensitivity to light and humidity, we proceed directly to injecting the mice.) 5. Working in the chemical fume hood, inject each mouse intraperitoneally with the appropriate dose of ENU. Transfer each mouse into a clean cage, leaving the old cage in hood. We change the needle between each cage. (Note: Due to the ethanol, the animals will appear uncoordinated and lethargic for
30 min after injection. They should be monitored to ensure they return to a normal level of activity. It is fine to house multiple males together.) 6. Spray or squirt inactivating solution on all equipment that has come into contact with the ENU solution. Dispose of all waste in appropriate containers. Inject at least 50 ml of inactivating solution into the ISOPAC container. Take off personal safety protection (gloves, disposable gowns, masks) and place all items in the bag. Put everything in the hood with the light ON for at least 24 h. Post signs warning all other workers away from the hood for 24 h. 7. After 24 h, ENU is inactive and all used reagents can be disposed of appropriately (solutions in chemical hazardous waste, empty ISOPAC container in glass waste, etc.).

Mouse ENU Screens

319

8. After 24 h change the mouse cages housing the mutagenized males. Spray the bedding with inactivating solution and put in biohazard bag. Spray cages and load into dishwasher. There is very little danger of ENU exposure at this point; however, if facility staff are performing this step, please be sure the person doing this is aware of the situation and wearing personal protective gear.

3.4. Breeding The only sure-fire way to know that the mutagenesis was effective is to obtain mutant lines, but there are several milestones to watch for that can serve as indicators. These include the mutagenized males becoming sterile within about 2 weeks after the final ENU injection, regaining fertility at about 10–12 weeks, and then succumbing to complications from the somatic mutations induced by the ENU within about 4–8 months (strains other than C57/BL6 may vary). We monitor these events by setting the males up with individual females 3 weeks after the final ENU injection and checking for copulation plugs. If a few plugs result in pups, we retest the relevant males and only include them in the screen if they retest as sterile. However, if the plugs prove to be productive across the population, this indicates the ENU was not effective, and it is best to just start over with new males. At about 8 weeks after the final ENU injection, we remove the test females from the cages of the sterile males and the following week add two wild-type females for the first cross of the screen. Within a month or two, litters from these crosses begin to be born. We use two females to ensure that we can propagate the mutations before the males die, so that all induced mutations can be screened. The initial crosses serve two purposes: (1) to propagate the mutations so that they can be screened and (2) to initiate the mapping cross. The exact crossing scheme depends on whether you are performing a screen for dominant or recessive mutations and whether you need to incorporate specific lines to sensitize the background or enable you to see a marked phenotype (Fig. 16.1). Dominant screens are the simplest: in just one generation, you will identify autosomal or X-linked mutations. Recessive screens generally require you to perform two or three generations of crosses (see Note in Fig. 16.1 legend for the exception). In the two-generation (G2) version, you will cross to a deletion or genetically modified strain so you can screen the G2 animals. In the three-generation (G3) version, you will backcross G2 females to their F1 fathers and examine the mutations in the G3 animals. It is only necessary to keep the mice that are essential for the screen. Since the germ line is repopulated by a limited number of spermatogonia in the mutagenized males, only 8–10 F1 progeny from each mutagenized male should be bred to avoid isolating the same mutations repeatedly

320

Tamara Caspary

ENU +A +A

⌾

+B +B

*A +B

F1

⌾

One-generation dominant screen

~B +B

⌾

~B G2

*A +B

*A

+B +B

or

+B +B

Backcross G2 females with their fathers Two-generation recessive screen G3

*A *A Three-generation recessive screen

Figure 16.1 Overview of breeding schemes. Wild-type (þ) male mice of strain A (A) are mutagenized with ENU and crossed to wild-type mice of strain B (B), also referred to in the text as the mapping strain. ENU mutations are designated by an asterisk (*). For a one-generation dominant screen, the resulting progeny are assayed for the phenotype of interest. In a two-generation recessive screen, the F1 mice are crossed to a defined mutant allele (designated as
) and the resulting progeny screened. This can reveal new alleles of the original defined mutation, as well as interacting factors (nonallelic, noncomplementation). (Note: If the phenotype is viable and fertile, new alleles of a mutation can be recovered in a one-generation cross by crossing mutagenized males to
B/þB mice). For the three-generation recessive screen, the F1 males are crossed to wild-type mice of strain B, and the female progeny backcrossed to their fathers. Note that this is a blind cross, as only half the female G2 animals will carry a mutagenized chromosome. The resulting G3 progeny are then screened to see whether 25% display a phenotype of interest (*A/*A).

( Justice, 2000; Russell, 2004). For example, when we do a three-generation cross for recessive mutations, we discard all F1 females and all extra (>8) F1 males. As we only breed G2 females back to their F1 fathers, we also discard all G2 males until we see a phenotype of interest. At that point we begin to wean and keep the G2 males, so that we can identify carrier males and establish the line (see more in Section 3.7). This minimizes mouse cage costs during the screen, while ultimately enabling us to recover lines.

Mouse ENU Screens

321

3.5. Screening To ensure ‘‘repetitive consistency’’ and keep the mouse room under control, it is best to minimize the number of people who perform the actual screening. Typically, we limit the number of people to two to three. This lets us communicate possible phenotypes effectively and take individual days off. Although the mice are mutagenized and initially mated on the same days, there is a natural spread to the productivity of each cross. This means that the screening distributes across days, but also that you must continually screen to prevent a backlog. This is of critical importance in a mouse room with limited space (or to researchers with a limited budget). As the screening proceeds, decisions to dispose of lines that do not display the phenotype of interest must be made. We discard lines when we have not seen the phenotype of interest in three litters that are expected to carry the appropriate genotype. Thus, in a dominant screen, we examine three F1 litters; in a recessive screen, where only half the G2 animals carry the mutations to be screened, we examine six litters. The unbiased nature of forward genetic screens makes them ideal for collaboration. The general thought is that since the mutagenesis and breeding are already being done, then it is efficient to look for more than one phenotype. This is absolutely true, as long as all parties agree to maintain the screening schedule. If the extra phenotype is something that anyone can look for while searching for the initial phenotype of interest, this can go quite smoothly. However, if the other phenotype requires that a second lab do separate tests, they must do so expeditiously. Alternatively, the mice can be physically transferred to the second lab if the first lab does not find their phenotype of interest. This latter solution also works well if the second lab is interested in a phenotype at a distinct time point. The important thing to keep in mind is that any backlog in calling lines results in an increased burden in the mouse room.

3.6. Mutants! Seeing the phenotype you have hoped to see for the first time is exciting; seeing it again in a second litter from the same line is exhilarating, as it means the phenotype is genetic. This is when you know the screen has worked and, simply put, it is amazing how well it works. In the course of the screen, you will likely run across the phenotype you are interested in sporadically, because you are looking for it. But the genetic ones are the ones you want. At a minimum, you want to see the phenotype twice in two separate litters before you can have much hope that it is genetic. In practice, you are likely to see it more frequently. Despite these guidelines about how many litters to examine and how many times to see a phenotype before establishing a line, we do break the

322

Tamara Caspary

rules from time to time. For instance, some phenotypes may not be so clearcut and can generate suspicion, but nothing conclusive. In such cases, screen more litters and determine whether the phenotype is interesting and amenable to study. It can be helpful to decide up front how much further to screen, so that months later you are not still stuck trying to decide. As with any screen in any organism, as the number of mutants rises and it becomes clear the screen worked, you will be less tempted to keep the problematic ones.

3.7. Establishing independent lines Most screens depend on one obligate carrier, often the F1 male. Until more carrier animals are identified, however, we refrain from considering the line to be a definite mutant line. We identify carrier animals through blind crosses, crosses with animals that may or may not carry the mutation of interest; if the cross results in the phenotype of interest, then the test animal is a carrier. For a one-generation dominant screen or two-generation recessive screens, the test animal needs to be crossed to a wild-type or a deletion animal, respectively. For a three-generation recessive screen for adult phenotypes, the test animal can be crossed to a known carrier. The crosses are more involved when you examine embryonic phenotypes. Since the females must be sacrificed so the embryos can be harvested, the identification of new carrier males requires crosses between two test animals. The G2 animals each have a 50% chance of carrying the mutation, which means one out of every four litters of a G2 intercross will display the mutant phenotype. Simultaneously, testing multiple G2 males a single time yields the highest probability of finding carriers (as opposed to retesting individual G2 males with additional G2 females), and thus is the most effective way to identify carrier males with which to establish the line. At a practical level, the number of sexually mature females will always be the limiting resource in the screen. In the mouse room, setting one G2 female up with one G2 male and discarding any cages that do not result in the phenotype of interest efficiently minimizes cost while establishing the mutant line. One practical note about keeping track of the lines: We designate each male we mutagenize with a unique letter (or combination of letters) and add a number to that letter to identify each resulting F1 male. That way it is clear during the screening which lines are related to one another. This is useful when similar phenotypes arise, because if they originate from the same mutagenized male, they may in fact represent the same genetic alteration. This identification scheme also makes it easy to know when eight F1 males have been screened.

Mouse ENU Screens

323

4. Gene Identification 4.1. Mapping Gene identification depends on mapping the mutation to a locus to pinpoint its location. This is possible, thanks to meiotic recombination between the alleles of the mutagenized strain and the mapping strain. Analyzing markers that are polymorphic between the two strains distinguishes the alleles and allows you to monitor the part of the genome that segregates with the phenotype. Either SNPs or simple sequence length polymorphisms (SSLPs) are used. In both cases, PCR primers surrounding the polymorphism can be used to amplify the DNA at a particular marker. SSLPs are detectable simply by the size differences between the two strains when run on a gel, whereas the SNPs can be detected by restriction digestion that distinguishes the alleles. Alternatively, array-based SNP panels can be used. Mutant lines must first be mapped to a chromosome. This should be done as soon as DNAs representing around a dozen opportunities for meiotic recombination are available from affected individuals. Knowing the chromosome on which the mutation is found lets you genotype the mice, drastically reducing mouse cage space and cost. The DNAs gathered need to be analyzed with low-density markers spaced evenly across the genome. When doing this ‘‘in house,’’ we have found that 41 markers are sufficient. This translates to three markers on the largest mouse chromosomes (1–3) and two markers per chromosome for the smaller ones (4–19). Alternatively, the job can be outsourced to a core facility that uses Illumina’s low-density SNP panel or equivalent. One panel that has been optimized solely for the purpose of mapping ENU-induced mutations is the 768 SNP panel, which is available for fee-for-service through the Partners Center for Personalized Genomic Medicine (PCPGM). Whichever method you choose, the genotypes at any particular marker will enable you to locate the mutation to a chromosome (and sometimes to a defined interval on the chromosome). As an aside, if sperm cryopreservation is an available option, this is a good time to start setting aside males for the procedure. Successful screens often lead to the identification of more lines than a single lab can study immediately. Freezing sperm can also serve as an insurance policy against loss of the line. In addition to protection against mouse room calamities, frozen sperm allow a line to be recovered if experimental mistakes are made during the mapping or mutation identification phase. The next step you will take is fine mapping the mutation. There are two goals in the course of fine mapping: (1) to increase the opportunities for meiotic recombination and (2) to increase the polymorphic marker density

324

Tamara Caspary

within the mutation interval. The first goal is achieved by continuously crossing in the mapping strain each generation, and the second by efficient use of the many mouse resources available, which means you can use markers others have developed or make new markers. Once you have confirmed the markers are polymorphic between the strains in your screen, you can define the smallest region in which the mutation must lie by analyzing all recombination events. The DNA samples from recombinant animals will get reanalyzed again and again with dense markers across the interval, until the region is small enough that you are ready to start sequencing genes.

4.2. Sequencing Sequencing technology is clearly poised to transform screens in the next decade, as next-generation resequencing technology is improved and becomes cheaper. Sequencing has always been essential to identify a causative mutation; what will change are decisions about when to start sequencing. The mapping is done to reduce the region in which the mutation must lie to a manageable interval. To date, that means an interval with a number of candidate genes that can be PCR-sequenced. For extensively spliced genes, amplicons to both cDNAs and to individual exons often get analyzed, since a third of all ENU mutations disrupt normal splicing of genes. This method is extremely effective, having revealed hundreds of ENU alleles so far. Targeted sequencing enables analyses of much larger amplicons and is clearly effective in identifying ENU-induced causal mutations. For example, the Megf8 ENU mutant was the first ENU allele identified through targeted sequencing (Zhang et al., 2009). This required the generation of a BAC library from the mutant DNA, and the minimal tiling path across a 2.2-Mb interval was sequenced. Subsequently, a number of ENU alleles that all mapped to chromosome 11 were identified using targeted resequencing of PCR products (Boles et al., 2009). New methods of preparing the mutant DNA samples promise to facilitate the incorporation of resequencing into mouse ENU screens. Ultimately, methods that will sequence all the exons and splice sites on a given chromosome will be available, virtually eliminating the need for fine mapping ENU-induced mutations. In any case, the initial sequencing is usually performed on a sample from an individual affected animal. You can include a control sample from the mutagenized strain (but a nonmutagenized individual), or the database may also substitute for this control. The mouse genome sequence database represents C57/BL6 sequence, and SNP data is constantly being added for other strains (highlighted at genomic location at www.ensembl.org/Mus_musculus/Info/Index). In order to validate any change, it is best to compare PCR-based Sanger sequencing of DNA (or cDNA) from two distinct

Mouse ENU Screens

325

mutant animals and at least one control sample. Any base pair changes are potentially causative and should be analyzed for their predicted effect. In general, the functional ENU-induced changes alter amino acids or splicing of genes and are fairly obvious. At this point we like to develop a genotyping assay that detects the ENU-induced change directly. This enables us to analyze all the DNAs in the screen to determine whether the mutation segregates with the phenotype. Sometimes the base pair change creates or destroys a restriction site so we design a restriction fragment length polymorphism (RFLP) assay. If no restriction site exists, we create one with a primer. We design a primer in which the final base pairs juxtaposed to the ENU-induced change are in turn changed to create a restriction site that requires the ENU-induced change. The ENU-induced base pair itself is within the amplicon and not on the primer. When the primer is used to amplify the mutant allele, the restriction enzyme recognizes the product and cuts it, whereas the wildtype allele is not cut (or vice versa). This same method can be used to detect SNP markers in the mapping stage, as well. Such assays let us quickly interrogate all recombinants in the screen to see whether the mutation does in fact segregate with the phenotype of interest. Perfect segregation is a powerful indicator that we have identified the causative mutation.

4.3. Proving the variant is causal ENU induces mutations fairly randomly throughout the genome, so confirming that a variant exists and segregates with the phenotype of interest still leaves open the possibility that another linked mutation plays a functional role in causing the phenotype. Again, in the near future sequencing technology will likely enable us to identify what (if any) the linked mutations are. Even then, though, each mutant line will present its own unique data that will determine what is necessary to prove the variant is causal. The gold standard tests to prove causality are genetic: transgenic rescue or failure to complement a known mutant allele of the gene in question. However, the available reagents will largely decide which approach is best for you. For example, if the mutation is in a gene for which there are existing antibodies, a Western blot could reveal that the mutant line has no stable protein. If there is an existing mouse mutant in the gene, the alleles should fail to complement. A harder situation arises when a completely uncharacterized gene is suspected. If the gene is in a pathway for which there are reagents in the downstream products, they might be useful. But when there are no reagents, proving allelism can be time consuming. The most popular methods are raising an antibody against the protein or making a mutant mouse. This latter possibility is aided by the concerted efforts of American and European consortiums, which are making available alleles in all mouse genes through gene traps and targeting (Collins et al., 2007a,b).

326

Tamara Caspary

In fact, the National Institutes of Health Mouse Knockout Project (NIH KOMP) and the European Conditional Mouse Mutagenesis Program (EUCOMM) accept nominations to target specific conditional mutations in embryonic stem (ES) cells using their high-throughput methods. The targeted ES cells are then available to any researcher in the world. This can save a great deal of time and money in getting the appropriate mutant allele and provides a generally useful reagent for the future.

4.4. Cryopreservation of lines An additional benefit of the mapping cross is that it eliminates the unlinked, nonfunctional ENU-induced mutations (‘‘cleans up’’ the line in genetics vernacular). Once the mutant line is congenic on the background of the mapping strain, we cryopreserve sperm from established male carriers. This allows us to eliminate lines we are not actively studying from the mouse room, while preserving the line for researchers in the future. All alleles are entered into the database of alleles maintained by The Jackson Labs (http:// www.informatics.jax.org/, through the submit data link).

5. The Future for Mouse Forward Genetics In the past decade, the completion of the mouse genome sequence propelled mouse forward genetic screens into a tractable method for mouse geneticists. In the coming decade, technological advances promise to make such lines more easily available to all researchers, no matter their expertise. This will hinge primarily on next-generation resequencing which is already facilitating gene identification. As it is easier to find causative mutations, it is likely that repositories for the mutant lines will expand as has been true for the knockout projects. As this occurs I see forward genetic screens truly complementing reverse genetic approaches in the toolbox researchers use to uncover mammalian biology.

REFERENCES Beier, D. R. (2000). Sequence-based analysis of mutagenized mice. Mamm. Genome 11, 594–597. Boles, M. K., et al. (2009). Discovery of candidate disease genes in ENU-induced mouse mutants by large-scale sequencing, including a splice-site mutation in nucleoredoxin. PLoS Genet. 5, e1000759. Chen, Y., et al. (2000). Genotype-based screen for ENU-induced mutations in mouse embryonic stem cells. Nat. Genet. 24, 314–317. Coghill, E. L., et al. (2002). A gene-driven approach to the identification of ENU mutants in the mouse. Nat. Genet. 30, 255–256.

Mouse ENU Screens

327

Collins, F. S., et al. (2007a). A new partner for the international knockout mouse consortium. Cell 129, 235. Collins, F. S., et al. (2007b). A mouse for all reasons. Cell 128, 9–13. Concepcion, D., et al. (2004). Mutation rate and predicted phenotypic target sizes in ethylnitrosourea-treated mice. Genetics 168, 953–959. Gondo, Y., et al. (2009). Next-generation gene targeting in the mouse for functional genomics. BMB Rep. 42, 315–323. Hitotsumachi, S., et al. (1985). Dose-repetition increases the mutagenic effectiveness of N-ethyl-N-nitrosourea in mouse spermatogonia. Proc. Natl. Acad. Sci. USA 82, 6619–6621. Justice, M. J. (2000). Mutagenesis of the mouse germline. In ‘‘Mouse Genetics and Transgenics: A Practical Approach,’’ (I. J. Jackson and C. M. Abbott, eds.), pp. 185–216. Oxford University Press, Oxford. Justice, M. J., et al. (2000). Effects of ENU dosage on mouse strains. Mamm. Genome 11, 484–488. McDonald, J. D., and Beier, D. (eds.) (2003). ENU Mutagenesis in the Mouse, John Wiley & Sons. Russell, L. B. (2004). Effects of male germ-cell stage on the frequency, nature, and spectrum of induced specific-locus mutations in the mouse. Genetica 122, 25–36. Russell, W. L., et al. (1979). Specific-locus test shows ethylnitrosourea to be the most potent mutagen in the mouse. Proc. Natl. Acad. Sci. USA 76, 5818–5819. Salinger, A. P., and Justice, M. J. (2008). Mouse mutagenesis using N-ethyl-N-nitrosourea (ENU). Cold Spring Harb. Protoc. doi:10.1101/pdb.prot4985. Weber, J. S., et al. (2000). Optimal N-ethyl-N-nitrosourea (ENU) doses for inbred mouse strains. Genesis 26, 230–233. Zhang, Z., et al. (2009). Massively parallel sequencing identifies the gene Megf8 with ENU-induced mutation causing heterotaxy. Proc. Natl. Acad. Sci. USA 106, 3219–3224.

C H A P T E R

S E V E N T E E N

Using ENU Mutagenesis for Phenotype-Driven Analysis of the Mouse Rolf W. Stottmann and David R. Beier Contents 330 331 332 334 335 335 335 336 336 339 340 343 345 345

1. Introduction 2. ENU Screen Design 2.1. Developmental phenotypes 2.2. Metabolic phenotypes 2.3. Physiological phenotypes 2.4. Cellular phenotypes 2.5. Behavioral phenotypes 2.6. Suppressor/enhancer phenotypes 3. ENU Treatment 4. Mutant Ascertainment 5. Mutation Identification 6. Mutation Validation 7. Summary References

Abstract The use of mutagenesis in invertebrates to generate phenotypic variants has a long and productive history. Despite the conclusive demonstration by Russell and colleagues in the 1970s that the chemical N-ethyl-N-nitrosourea (ENU) is a highly effective mutagen in mice, the application of phenotypic-driven mutagenesis as a method to study mammalian biology proceeded slowly. With the development of tools for genomic analysis, the task of positional cloning ENUinduced mutations has become quite feasible, and this approach has recently been widely applied and highly productive. It has specifically lived up to its theoretical utility as means to provide insight into the biological roles of genes that is not biased by presumptions of their function. While the power of this approach is indisputable, the effort necessary for its success remains Division of Genetics, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77017-8

#

2010 Elsevier Inc. All rights reserved.

329

330

Rolf W. Stottmann and David R. Beier

substantial, requiring careful attention to aspects including ENU treatment, mouse husbandry, screen assay design, genetic mapping, positional cloning, and mutation validation. In this chapter we discuss practical aspects of implementing a phenotype-driven analysis of an ENU-mutagenized mouse population.

1. Introduction Treatment with N-ethyl-N-nitrosourea (ENU) is an effective tool for generating DNA sequence variation with low morbidity and/or mortality in several model organisms. The first studies of mutagenesis using ENU as a mutagen in the mouse were some of the most informative; in these Russell et al. (1979) characterized the effectiveness of ENU using the specific locus test to quantify its mutation rate. These investigations determined that intraperitoneal (i.p.) injection of ENU could efficiently generate germ cell mutations without causing systemic morbidity. Analysis of the timing of the effects suggested that it is the spermatogonial stem cells that are sensitive to this agent. Additional studies determined that the efficiency of mutagenesis could be optimized by use of a fractionated dose (Russell et al., 1982). In this fashion, frequencies of 6–15
10 4 mutations per locus per gamete can be obtained, greater than that obtained using radiation mutagenesis and 1000-fold higher than the spontaneous rate of mutation of 0.5
10 6. Characterization of ENU mutations in Drosophila revealed that the molecular basis of these mutants were single-base changes; many recent studies have confirmed that this is the case in mammalian cells as well (e.g., Coghill et al., 2002). Russell’s pioneering studies were then followed by Bode’s elegant characterization of the breeding strategy for uncovering recessive mutations (see Fig. 17.1) and the practical

G0

ENU

G1

G2

G3

Figure 17.1 Three-generation strategy for generating recessive mutations.

Phenotype-Driven ENU Screens

331

demonstration of its utility by generating mouse models of hyperphenylalaninemia (McDonald and Bode, 1988; McDonald et al., 1990). However, despite the very clear demonstration of the efficacy of ENU, it was not readily adopted as a means to study mammalian biology, primarily due to the difficulty at the time of identifying loci that were mutated as a consequence of a single-base change. Fortunately, the technology was maintained and applied by investigators in a few laboratories; notably, that of Dove, who used it to generate the widely used Min mouse model of APC (Moser et al., 1990), Favor (Favor et al., 1991), and Guenet (Montagutelli et al., 1994; Tutois et al., 1991). As development of the tools for genomic analysis made it feasible to pursue the positional cloning of ENU-induced mutant loci, there was an increased appreciation of the potential utility of this approach for phenotype-driven analysis of mouse biology. The past 15 years have seen a remarkable resurgence in the application of ENU mutagenesis to all manner of biological processes in the mouse. These have been done both as large consortium efforts with broad phenotypic aims (Hrabe de Angelis et al., 2000; Nolan et al., 2000a) and by individual investigators focusing on specific problems (Herron et al., 2002; Kasarskis et al., 1998). The design and outcomes of these diverse efforts are the subject of numerous reviews (e.g., Acevedo-Arozena et al., 2008; Cook et al., 2006; Georgel et al., 2008; Godinho and Nolan, 2006). The aim of this chapter is to facilitate the consideration of using ENU mutagenesis investigation of mammalian biology by discussing the practical aspects of this approach.

2. ENU Screen Design Treatment with a mutagenic dose of ENU results in a period of infertility. Recovery from this occurs after 8–12 weeks, and the treated mice (designated G0 for Generation 0) are then bred. Dominantly or semidominantly inherited phenotypes can be assayed in their first generation (G1) progeny (Fig. 17.1). A number of screens have focused on this population, as it is logistically feasible to very screen large numbers of mice (Hrabe de Angelis et al., 2000; Nolan et al., 2000a). The productivity of this approach can be further enhanced by analyzing each mouse using a battery of tests (Gailus-Durner et al., 2005; Nolan et al., 2000b). However, because most mutations in the mouse have no or only subtle effects in heterozygotes, it is frequently necessary to employ a multigeneration breeding scheme to uncover recessive mutations. This is particularly the case for developmental mutations: consider that haploinsufficiency for Sonic hedgehog (Shh) causes holoprosencephaly in humans (Roessler et al., 1996), but heterozygous mice appear essentially normal even though homozygous embryos have profound patterning defects (Chiang et al., 1996).

332

Rolf W. Stottmann and David R. Beier

Additionally, recessive screens can identify mutations that result in lethality. Also, establishing that an observed phenotype is heritable and monogenic is facilitated by the design of a recessive screen, as it would be expected (and in fact is required) that multiple affected mice are obtained from a G1 male. While the husbandry demands make it less feasible to test large numbers of G1 mice than in a screen for dominant mutations, this is offset by the fact that each G1 male carries many mutations. Russell’s studies demonstrate that the per-locus mutation frequency of optimized fractionated dose ENU treatment is about 1/1000. The nature of the specific locus test that Russell employed would uncover primarily mutations that were effectively null. Assuming 25,000 genes in the mouse genome, one can therefore conclude a G1 male carries 25 null mutations, and potentially many more with hypomorphic effects. Finally, by performing the screen as an outcross, genetic informativeness is introduced such that the mutations can be mapped as they are obtained (as opposed to establishing a separate mapping cross after mutation ascertainment). The general strategy of husbandry described by Bode is the most commonly used means for recovering recessive mutations; this is discussed in more detail below (McDonald and Bode, 1988). There is indisputable evidence that ENU mutagenesis is efficient (with respect to generating single nucleotide changes), that these mutations can be readily mapped (assuming reasonable penetrance) and that they can ultimately be identified using straight-forward approaches of positional cloning. Thus, the only unknown in determining the success of a mutagenesis experiment is the likelihood that a heritable phenotypic variant will be identified. At one level, this is a function of the number of genes that are required for the generation or maintenance of ‘‘normal,’’ with respect to the phenotype in question. If, for example, it is a single gene, the probability of ascertaining this in a standard mutagenesis analysis is low. If, however, tens to hundreds of genes are required to generate or maintain the wild-type state, the likelihood of their discovery in a modest-size screen is quite reasonable. Given the latter scenario, the key determinant for success is whether the phenotype to be screened is amenable to reasonably high-throughput analysis (as even a modest-size screen still requires the analysis of hundreds to thousands of progeny) and whether the assay will unambiguously discriminate between normal variation and a truly mutant phenotype (as the latter will still be rare relative to the number of normal mice). With this general principle in mind, ENU mutagenesis can be applied for investigation of a wide variety of biological processes, as discussed in the following sections.

2.1. Developmental phenotypes ENU mutagenesis has arguably been most successful in discovering genes required for normal development. This is likely due to a number of factors: the biological complexity of the process (such that there are many target loci),

Phenotype-Driven ENU Screens

333

the relative simplicity and qualitative nature of the assay (which may often require only inspection), and the fact that compensatory changes in gene expression are unlikely to ‘‘correct’’ a developmental defect and thus mask its occurrence (as compared to physiological/ metabolic parameters, as discussed below). Furthermore, the present wealth of knowledge regarding the genetic basis of development often allows investigators to rapidly integrate novel findings into existing paradigms. This is crucial, as the entire premise of a phenotype-driven approach is that genes that are previously entirely uncharacterized or would not necessarily be presumed to play a role in a specific biological process will be identified. The ability to assign mechanism to genes of unexpected consequence can rapidly enlarge our understanding of functional interactions. However, integrating a novel gene with limited functional annotation into a mechanistic understanding of the developmental phenotype will often be a significant task. This success is best exemplified by the role that mutagenesis has had in delineating the importance of primary cilia in mediating Hedgehog signaling. Much of this insight has been obtained from studies by Anderson and colleagues of mutants ascertained at midgestation with morphological defects (Caspary et al., 2007; Garcia-Garcia et al., 2005; Huangfu et al., 2003; Liem et al., 2009; Ocbina and Anderson, 2008; Weatherbee et al., 2009). These were examined for abnormalities in neural tube patterning, and then more specifically by genetic and biochemical analysis for perturbation of the Hedgehog pathway. Unexpectedly, many of these proved to be genes previously implicated in flagellar function in Chlamydomonas, and, by extension, cilial function in vertebrates. Genes uncovered in mutagenesis studies by other investigators have contributed to understanding this pathway as well (Endoh-Yamagami et al., 2009; Ermakov et al., 2009; May et al., 2005; Tran et al., 2008). A method to increase the sensitivity of developmental phenotyping is to highlight a structure of interest beyond what one can assess by inspection. This can be done in a tissue-specific fashion in situ using immunohistochemistry or RNA hybridization (Mar et al., 2005). However, this will require a significant increase in effort as well as the amount of time necessary for determining which animals have a phenotype of interest. The latter is not a trivial concern as a major challenge to an efficient screen is the speed with which a determination can be made that a given line is worthy of interest. One way to significantly decrease the assay time for tissue-specific analysis is to use a genetic model that carries either a transgene or ‘‘knock-in’’ allele that expresses a reporter in an informative pattern. While a modest amount of additional effort is required to genotype carriers for use during the husbandry, this can be readily justified by the marked increase in sensitivity (Zarbalis et al., 2004). Ideally, these reporters are from inbred strains with a uniform background, although some genetic

334

Rolf W. Stottmann and David R. Beier

heterogeneity can be tolerated. In the use of a reporter strain, investigators must be aware of the possibility that an ENU mutation could affect the expression of the transgene/‘‘knock-in’’ itself without directly changing the nature of the tissue normally expressing the transgene (i.e., a regulatory mutation affecting the reporter locus). As such, an independent method to validate the presumptive phenotype of any potential mutations would be useful (e.g., immunohistochemical analysis of a protein expressed in the same spatiotemporal domain). Most reporter alleles take advantage of either green fluorescent protein (GFP) or beta-galactosidase (bgal). If GFP is the reporter of choice, investigators should be aware that the phenotype analysis will likely have to be done immediately upon embryo or tissue harvesting. This requires significant periods of access to a fluorescent microscope and may be difficult to do in addition to dissections and tissue preparation in any given work session. If bgal reporters are used, the tissue must be lightly fixed and the X-gal stain added. However, the analysis of the bgal expression pattern can be done in a separate work session as the substrate is stable for some time. Additionally, bgal staining can be seen more clearly in thicker blocks of tissue whereas the GFP is more useful in early embryos or transparent organs. If a reporter allele is to be used, a test cross to show that the reporter can indeed highlight a phenotype of interest is a desirable early step in evaluation of the allele. Crossing of the reporter with a previously characterized mutation known to perturb the tissue of interest should allow the investigator to ascertain putative homozygous mutants by examination for altered expression of the reporter. These can then be verified by genotyping to test the sensitivity and reliability of the reporter.

2.2. Metabolic phenotypes These have been the focus of a number of screening efforts, with a mixed history of success. These screens have been particularly compelling given the relevance of many of the screened phenotypes to human disease. Further, highly sensitive assays have been developed for many clinically relevant traits, and the importance of these for diagnostic testing has facilitated their translation into technically simple and/or automated tests that are amenable for high-throughput analysis. Application of these assays to mutagenesis screens has yielded a wealth of mutations affecting basic metabolic processes, which often serve as models of corresponding human diseases (Aigner et al., 2009a). Interestingly, some traits appear less amenable to perturbation using mutagenesis (Aigner et al., 2009b). This is likely due either to the fact that such perturbations are not tolerated, leading to early lethality, or that compensatory metabolic changes make allelic variants difficult to detect.

Phenotype-Driven ENU Screens

335

2.3. Physiological phenotypes This class is similar to the metabolic class, with respect to medical relevance and the existence of sensitive assays that have been adapted for analysis of rodent models. However, there are some important differences. Firstly, these assays are often time, labor, and resource-intensive, making them less amenable to analysis of large numbers of mice. Secondly, physiological traits may show considerable normal variation, making it more difficult to distinguish new mutants. This may be further compounded during the mapping analysis, as strain-specific effects may result in additional variation in intercross or backcross progeny. Despite this, there have been notable efforts to develop systematic high-throughput protocols to achieve efficiency and consistency (Hardisty-Hughes et al., 2010).

2.4. Cellular phenotypes One area of considerable success is the utilization of cell-based assays for screening. While the extraction and preparation of the cells themselves requires effort, the assays themselves are frequently amenable to large-scale analysis. The identification of variants in immune function has been particularly productive (Cook et al., 2006; Georgel et al., 2008). This is due in part to the routine use of cell-based assays as part of immunological investigation, the relative ease of sample preparation, and the fact that, for analyses using peripheral blood, the mutant mice can continue to be bred after phenotyping. However, even more complex cell-based assays, such as those analyzing genome stability and DNA replication, have been successful as well (Shima et al., 2003). One caveat for this approach is that the rationale for organismal mutagenesis to obtain a cellular phenotype must be carefully considered, given the ability to directly knockdown gene expression in cells using RNAi technology (Nybakken et al., 2005).

2.5. Behavioral phenotypes There has been considerable interest in obtaining behavioral mutants using mouse mutagenesis, as modeling human cerebral function is not readily done in lower organisms. However, behavioral assays have unique challenges with respect to throughput and reproducibility (Crabbe et al., 1999; Mandillo et al., 2008). That said, the identification and characterization of the clock mutant by Takahashi and colleagues is notable both for its importance to our understanding of the molecular basis of circadian rhythm, and its success at a time when genomic tools for positional cloning were not yet well developed (Vitaterna et al., 1994). The key to this experiment was having a sensitive and highly reproducible quantitative assay, as well as a

336

Rolf W. Stottmann and David R. Beier

relatively narrow range of normal variation. Thus, a well-designed assay for behavioral traits is amenable to application in a mutagenesis screen.

2.6. Suppressor/enhancer phenotypes A strategy well proven in lower organisms that has been productively translated to mouse mutagenesis is the generation of modifier loci. This is done by including mice carrying a known mutation in a mutagenesis experiment, and then screening for loci that enhance or diminish the phenotypic effect of the mutant locus (Kauppi et al., 2008; Matera et al., 2008). The fact that unlinked genetic loci affect phenotypic expression in mice has long been evident by virtue of the fact that this can vary substantially depending upon strain background. However, efforts to identify the specific modifying loci using genetic analysis have had limited success, due to the difficulty in obtaining a high degree of genetic resolution, coupled with the large number of strain-specific DNA sequence variants. While it may seem at first glance that the effort required to generate modifier loci by de novo mutagenesis is inefficient, the fact that, once obtained, they can be unambiguously identified using positional cloning techniques makes this approach compelling, especially as these techniques are constantly being enhanced as genomic technology advances. A limitation of this approach is that mutagenesis will more likely create loss-of-function mutations, so certain classes of modifying loci (i.e., gain of function or those caused by increased expression) will be less efficiently uncovered. Furthermore, if a novel phenotype is seen, it will be necessary to determine if indeed this is truly dependent on the sensitizing locus by testing whether it segregates with the mutation.

3. ENU Treatment ENU is a potent mutagen and one should take care to use all appropriate personal protective equipment. Injections are best done in a hardducted fume hood and protocols should be discussed with animal facility staff and supervisory personnel. ENU obtained from Sigma comes within an ISOPAC unit; to ensure maximal ENU effectiveness, a fresh ISOPAC should be used for each session of ENU treatment and injections performed within a few hours of preparing the solution. Initially, dissolve the ENU by injecting 10 ml of 95% ethanol with a 10-ml syringe and 16-gauge needle. To relieve pressure in the ISOPAC, vent the container using a second 16gauge needle. Hand-warm the solution and inspect to insure the ENU is fully dissolved—the solution will be completely clear and yellow. Further dilute the ENU with 90 ml of the phosphate/citrate diluent buffer (100 mM

Phenotype-Driven ENU Screens

337

sodium phosphate dibasic (14.2 g/l solution), 50 mM sodium citrate (14.7 g sodium citrate dihydrate per liter solution), adjust pH to 5.0 using phosphoric acid). This dilution is most easily done with two injections of 45 ml. Again, be aware of the pressure inside the ISOPAC and pause at approximately each 10 ml of injection to allow pressure to vent. ENU is light sensitive so the container should be covered in foil. The actual concentration of ENU needs to be determined experimentally to ensure appropriate dosage. Use a 1-ml syringe and 27-gauge needle to remove 200 ml of the ENU solution from the ISOPAC and add 800 ml of diluent buffer (a 1:5 dilution). Prepare a blank solution of 20 ml 95% ethanol and 980 ml diluent buffer and prepare spectrophotometer. Transfer the diluted ENU solution to a disposable spectrophotometer cuvette and perform a wavelength scan from 350 to 450 nm. An effective preparation of ENU will have a peak centered at 398 nm. Measure the absorbance at 398 nm; a typical reading will be 1.3–1.5. Calculate the concentration of ENU with the reference point of 1 mg/ml at OD398 ¼ 0.72. Concentration in mg/ml ¼ (OD at 398 nm/0.72)
5 [to correct for the dilution]. Each mouse should be weighed and the required injection volume calculated according to the following formula: Injection volume ¼ ðmass of animalÞ
ðENU doseÞ
ð0:001Þ=ENU stock concentration The ENU solution is injected intraperitoneally. Shortly after injection, the mice may appear wobbly due to alcoholic solvent of ENU. After injection, inactivate remaining ENU solution by injecting 20–30 ml of alkaline thiosulfate inactivating solution (0.1 M NaOH (4 g/l solution), 20% (w/v) sodium thiosulfate (200 g/l solution)) into the ISOPAC. Leave one syringe in ISOPAC vessel in fume hood to allow gas to escape and leave in the fume hood overnight. Soak all equipment coming into contact with ENU with inactivating solution and dispose of as biohazardous materials. Mice should stay within the fume hood 24–48 h to allow for the ENU to become inactivated before returning to standard colony housing. Any ENU not absorbed by the tissues will be excreted and disposed of with all bedding. Before returning treated animals, transfer to a new cage and mix old bedding with the inactivating solution before disposal. An important consideration is the choice of mouse strain to be mutagenized. The ideal strain is one which is known to breed well, able to tolerate ENU treatment, and does not have strain-specific effects on the tissue of interest. Fortunately, work has been done in this area to determine which strains are most tolerant of ENU treatments ( Justice et al., 2000; Weber et al., 2000). Popular choices include A/J, C3H, and C57BL/6. In our lab, we generally use A/J and outcross to FVB to take advantage of

338

Rolf W. Stottmann and David R. Beier

the tolerance of A/J to ENU treatment and the high fecundity of FVB. We have avoided mutagenizing B6, as we often introduce the mutation into this background by serial backcross, and the A/J alleles linked to the mutant locus provide a convenient means to identify carriers prior to mutation identification. However, the possibility of mutation identification using high-throughput sequencing, which we discuss below, may make this less necessary. Male mice treated with a mutagenic dose of ENU will lose fertility, which will return 8–12 weeks after treatment as the gamete cell lineage is repopulated. The long-term survival of these animals is compromised as compared to untreated animals due to somatic mutations induced by ENU, so treated animals have a somewhat limited time to pass on the mutations. In our experience, however, survival is sufficient to create an adequate number of offspring (Fig. 17.2). We generally treat 20 G0 mice and aim to obtain 75–100 G1 males. While there are standardized doses of ENU likely to be mutagenic that can be employed (Weber et al., 2000), we have had success using an empirical strategy to determine the most effective treatment regimen. In this approach, we treat G0 mice with a range of doses (e.g., with i.p. injections of 3
90, 3
95, or 3
100 mg ENU/g body weight), test injected males for recovery of fertility, and then use those treated with the highest doses for generation of G1 progeny.

120

Percent survival

100 80 60 40 20 0

1

4

7

10 13 16 19 22 25 28 31 34 37 40 43 46 49 Weeks postinjection

Figure 17.2 Percent survival of G0 male mice treated with ENU over time. Results of two mutagenesis experiments are shown (solid and broken lines).

Phenotype-Driven ENU Screens

339

When breeding the mutagenized (G0) males, it is important to try and preserve offspring from as many different productive G0 males as possible. This is necessary in order to capture as broad a range of ENU mutations as possible and not to limit this to the set of mutations present in the germ line from any given mutagenized male, which is repopulated by a finite (and potentially small) cohort of spermatagonial stem cells. The G1 progeny generated by breeding the G0 males are either examined directly (for detection of mutations which have their effects as heterozygotes) or the males are further bred, such that each of these is the ‘‘founder’’ of an individual line or family. As noted, genetic informativeness for mapping purposes can be introduced by mating G0 or G1 males with females from a different inbred strain. Also, screens requiring utilization of specific mutant loci (transgenic reporters, sensitizing loci) can be introduced at these breeding steps. An alternative to the breeding strategy described above is to intercross G1 mice and then intercross their G2 progeny. The third generation is then examined for mutants. This approach has the virtue of introducing twice as many mutants into a pedigree (because both G1 parents are mutagenized) (Silver et al., 2007). However, as any two G2 mice have only a 25% chance of both being heterozygous, a large number of mating pairs are necessary to comprehensively test a single family for the presence of a mutation of interest (compared to the three to four litters required for a conventional analysis, as discussed below).

4. Mutant Ascertainment As previously noted, the appropriate population for screening to identify recessive mutations are the G3 progeny of (G1
G2) mice (Fig. 17.1). If the phenotype of interest is not likely to be compatible with postnatal survival, one would sacrifice pregnant G2 females on day 18 of pregnancy (immediately before birth in most strains) or earlier and examine the G3 fetuses. The embryos to be analyzed can be staged to the investigator’s preference by examinations of G2 matings for vaginal plugs (‘‘timed pregnancies’’). Abnormalities that may be readily identified by this embryonic analysis include neural tube defects (both exencephaly and spina bifida), limb or other skeletal defects, facial clefting, eye anomalies such as micropthalmia or open eyelids, or paleness (secondary to anemia). Further, an internal exam can be performed to identify abnormalities in patterning and specific defects in organogenesis, or to obtain tissue for histological studies. For phenotypes that are compatible with postnatal survival, the G2 female can either be left with its G1 mate, as it will go into estrus after giving birth, facilitating the generation of additional G3 mice for analysis, or the G1 mating can be renewed after the G3 progeny are weaned.

340

Rolf W. Stottmann and David R. Beier

An important consideration is the estimation of how many litters from any G1 parent should be analyzed. For a highly penetrant phenotype, one can recommend that at least three litters from each G1 male be examined. The rationale for this is as follows: If one obtains at least three G2 daughters from each G1 male, there is an 87.5% chance that at least one of them is a heterozygote like the father. (For each G2 female resulting from a cross between a heterozygous mutant (m/þ) G1 male with a þ/þ female, there is a 50% chance that she will be m/þ. With three such G2 females, the likelihood of at least one heterozygote being produced is unity less the probability that none will be produced, or 1
0.53 ¼ 0.875.) If you examine a litter of eight from a cross between a G1 male m/þ and a G2 m/þ female, you will have a 90% chance of producing at least one homozygous m/m G3 animal. (This figure again is obtained by subtracting from 1 the probability that none will be produced; i.e., 1  0.758 ¼ 0.90). Therefore, if by looking at three G2 females you have a 87.5% likelihood of successfully producing at least one heterozygote and by looking at eight G3 animals you have a 90% likelihood of producing at least one m/m G3 animal (given that the mother is a heterozygote), then the overall likelihood of identifying a mutation carried by the G1 male is the product of these two probabilities, or nearly 80% (0.875
0.90 ¼ 0.79). Thus, an analysis of three litters provides reasonable confidence one will identify a fully penetrant mutation, and examination of additional litters adds only modest power. That this approach may be biased to ascertainment of highly penetrant mutations may be considered an advantage, as these will be more efficiently mapped and characterized. Note that this calculation does not factor in the use of transgenic or mutant lines in the analysis; this will reduce power (unless every mouse in the screen is a carrier) and appropriate correction should be made. When an abnormal phenotype of interest is identified, the G1 male should be crossed to additional G2 daughters in order to validate the phenotype and begin the mapping process. A phenotype found in a single litter may be due to multigenic or somatically determined causes. However, because of the multigeneration nature of the cross, a phenotype found reproducibly in G3 litters obtained from different G2 mothers is statistically likely to be monogenic, which is supported by our empirical experience. Once an abnormal phenotype is found in multiple G3 litters, the cross should be maintained and expanded in order to obtain sufficient affected mice for further analysis.

5. Mutation Identification Once a mutant line has been established, it is necessary to pursue genetic mapping in order to facilitate positional cloning (although rapid developments in sequencing technology may make this optional, as discussed below).

Phenotype-Driven ENU Screens

341

If a mutant screen has been performed in an inbred strain, it will be necessary to set up a standard mapping cross; for example, crossing the mutant to a different strain and intercrossing heterozygous carriers. If the mutant is not viable or fertile, progeny testing is required to identify the carrier mice. However, as noted above, introduction of a different strain into the screening protocol will introduce genetic informativeness, such that the mutant locus can be mapped using the same mice in which the mutation was identified. This is because, for a recessive mutation, all of the affected mice will be homozygous for the mutagenized parental background at the causal locus, while the (unlinked) remainder of the genome will segregate randomly. Indeed, the breeding strategy described by Bode adds an additional generation to that used in a standard mapping cross, which, if done as an outcross, further decreases the representation of mutagenized parental alleles, and makes identification of the mutant locus more efficient. Specifically, in any single mouse, the likelihood that a recessive mutant locus is homozygous for the parental mutagenized background is 100%, while the likelihood of an unlinked locus being homozygous is only 12.5%. Taking advantage of this, we have been able to map mutants with as few as three affected mice (in which the likelihood of a single locus being homozygous for the parental mutagenized background by random chance is 0.2%). In any case, the technical task of genetic mapping (at least to moderate resolution) has been hugely simplified by the development of automated genome-wide SNP analysis methods (Moran et al., 2006). There exists sufficient diversity between most inbred strains that a fixed panel provides ample marker density (i.e., selection of specific markers is unnecessary), and map resolution is usually limited more by the number of mice tested than by SNP informativeness. These assays are routinely performed at genotyping centers and core facilities, and the per-sample cost is modest for the number of loci tested. A primary goal of genetic mapping is to identify and exclude remutations in well-characterized genes. This can be done by examining the recombinant interval for genes known to have mutant phenotypes that are comparable to the mapped mutation. This obviously will not be conclusive, as it would require comprehensive knowledge of the mutant phenotypes of large numbers of loci. In addition to the literature, there are resources to facilitate this approach, such as the phenotype query function of the Mouse Genome Database: http://www.informatics.jax.org. The reason to rapidly identify remutations is to avoid the unnecessary commitment of resources to high-resolution mapping crosses. This is not to say that remutations in known genes are of no interest, as the determination of the specific base change can potentially be informative with respect to gene structure– function relationships (Bialek et al., 2004). A problem for genetic mapping analysis is that initial localization via a genome-wide SNP panel often defines a moderate-size chromosomal interval

342

Rolf W. Stottmann and David R. Beier

(30 Mb in our experience), and ‘‘bench-top’’ technologies for finemapping using SNPs are inefficient. We have developed a Web-based tool, SNP2RFLP, which can extract region-specific SNPs from the dbSNP database and identify those that create restriction fragment length polymorphisms (RFLPs) (Beckstead et al., 2008; http://genetics.bwh.harvard.edu/snp2rflp). This information can then be used to develop an RFLP assay and further refine the region containing the mutation of interest. Alternatively, microsatellite markers can be identified using genome databases (such as the UCSC genome browser; http://genome.ucsc.edu/), and allele variation can be queried using resources such as MGI (http://www.informatics.jax.org). Finally, it may be simplest to identify SNPs in the region using database or browser queries, and sequence these directly in the set of mice for which the recombinant interval remains indeterminate. A tool for positional cloning that has been perhaps underutilized in ENU mutagenesis studies is the use of microarray analysis as a means to identify candidate genes for mutational analysis. This is probably the case because one might not expect the single-base changes induced by ENU to affect mRNA expression. However, many ENU-induced mutations affect splicing, and ENU-induced mutations in coding regions can result in nonsense-mediated decay of mRNA. This approach has the virtue that it can be employed even when the mapping resolution is moderate, is low cost, and may allow one to identify a candidate locus without the husbandry required for a high-resolution analysis. Since the recombinant interval containing the mutation is known, one can examine specifically expression differences in that region; that is, genomewide thresholds for significance do not need to be applied. Furthermore, while an initial characterization of a microarray dataset would focus on differential mRNA expression to identify the causal locus, this expression data could also become useful when trying to determine the molecular mechanism underlying the phenotype at later stages of analysis. One caveat for this approach is that the expression differences may exist in the recombinant interval that are strainspecific, and unrelated to the mutation. These will be discovered if one compares mutant and wild-type littermates from an outcross, as they will be discordant for strain background around the mutant locus. Consequently, presumptive differential expression should be validated using comparison to wild-type mice of the mutagenized parental background. Once a mutant locus has been localized, it is necessary to identify sequence variants in candidate genes. This is an area of such rapidly evolving technology that one can be sure what is written here will be dated by the time it is published. Historically, one selected candidate genes based on examination of the recombinant interval and inferences from data on gene expression and function. This has been recently facilitated by databases of tissue- or temporal-specific expression data, usually obtained from microarray analysis. Genes of interest would then be sequenced, using either genomic templates and analysis of exons and their flanking splice junctions,

Phenotype-Driven ENU Screens

343

and/or cDNA templates and analysis of transcripts. While it is certainly possible that a causal mutation will occur in nontranscribed sequences, there is presently only one report documenting an ENU-induced mutation occurring in a regulatory region (Sagai et al., 2004). There is also a recent report describing an ENU-induced mutation in a microRNA seed region (Lewis et al., 2009). There is clearly bias in the conclusion that exons and their flanking regions are likely to contain the mutated nucleotide, as nontranscribed regions are rarely examined in detail. However, in the characterization of ENU mutants mapped to loci where known candidates exist, presumptively causal sequence variants can usually be found in coding regions or splice sites (Hart et al., 2005). Using the above approach, a necessary task was to narrow the recombinant interval as much as possible to constrain the number of candidate genes. This would typically require the analysis of hundreds of intercross or backcross progeny, which was both costly and slow. With the development of means for efficient sequencing of exons, generally by hybridization capture, as well as improved gene annotation, high-resolution mapping is less compelling. That is, it is feasible to sequence the entire predicted exon complement from any specified region (or the entire genome). Furthermore, there are sufficient SNPs in coding regions that this analysis can be used to delineate the map position directly from the sequence characterization of a pooled sample (in which all mice will be homozygous for the mutagenized parental background in the region of the mutant locus); that is, a separate mapping analysis may be unnecessary (but perhaps still advisable, given the low cost). Further, whole exome sequencing may make it unnecessary to map at all. That is, one can contemplate sequencing a pooled sample of affected mice that were never outcrossed; while there will be many mutations segregating in the population, only one (or a few tightly linked variants) will be homozygous in all affected mice, and will appear as a difference from a reference genome. Finally, it seems inevitable that decreasing costs for whole genome sequencing will eventually make the exon capture step unnecessary, and one will proceed very rapidly from ascertaining mutants to querying their entire genome for mutations. A caution for this utopian vision is the computational task required for managing the data from these analyses will be immense, and the technical limitations that will qualify its application (such as error rates and coverage issues) still need to be addressed.

6. Mutation Validation A significant issue with respect to chemical mutagenesis is the fact that, for many mutant loci, only the single ENU-induced allele may exist. Thus, a validation step is necessary to prove that the presumptive mutant gene is

344

Rolf W. Stottmann and David R. Beier

indeed causal for the observed phenotype. This will increasingly be an issue to the extent that advances in sequencing technologies will not require that the mutant locus be resolved to high resolution by genetic mapping. Given that the mutation rate of ENU treatment is 0.5–2 Mb 1, sequencing will uncover many variants within recombinant intervals of moderate resolution (Boles et al., 2009). While causality can be suggested by functional studies, a formal proof requires either correcting the defect by complementation or identifying independent alleles. The former can be achieved by BAC transgenesis; however, this is expensive and time-consuming. Further, if it is necessary to create novel mouse strains, it will generally be more useful to generate a conditional or reporter-tagged allele by homologous recombination, despite the additional work this may entail. This is particularly the case for hypomorphic mutations, since it will be of interest to determine the null phenotype. Recent developments in gene targeting technologies have made it practical to propose to do this on significant numbers of mutant lines. The generation of targeted mutants is being hugely facilitated by international mouse knockout projects such as KOMP, EuCOMM, and NorCOMM that are on target to generate their combined goal of 14,000 mutant ES cells (Int’l_Mouse_Knockout_Consortium, 2007). Genes of interest can be queried using the IKMC (http://www.knockoutmouse.org/) database, and cells and mice are available from mutant respositories (e.g., Mutant Mouse Regional Resource Centers; http://www.mmrrc.org/). However, not all genes will be targeted, many will not have conditional alleles, and germ-line potential of clones, while high, is not 100%. To address this, we have pursued a strategy of embryonic expression of ‘‘transient’’ transgenic RNAi as a means to validate candidate mutations as causal. In this case, ‘‘transient’’ means that the microinjected embryos are directly examined, and not grown to generate stable transgenic lines. This is a common strategy for the in vivo analysis of enhancers, as it is often necessary to examine many deletion variants. This approach is theoretically ideal for mutation validation, as one could evaluate knockdown phenotypes within several weeks of microinjection, obviating the need for the establishment of permanent knockout lines. The major obstacles are the variability of expression knockdown, and the efficiency of generating mice expressing the dsRNA. The former is possibly less of a problem, as one can screen for dsRNA constructs that generate efficient knockdown by testing them in vitro. Further, some variation in knockdown could be highly informative, as many ENU-induced mutations are hypomorphic, and a complete null may prove to have a more severe phenotype that is not readily comparable to the ENU mutant. However, this makes the latter problem of efficiency all the more compelling, as it is necessary to generate reasonably large numbers of knockdown embryos in order to assess the range of phenotypes in a population.

Phenotype-Driven ENU Screens

345

Given this, the traditional strategy of microinjection to make transient transgenic mice, which may yield 10–20% positive embryos, is unattractive. We have taken advantage of the observation that transposable elements such as Sleeping Beauty or PiggyBac can facilitate transgenesis. We have found that injection of a vector carrying PiggyBac transposon that expresses an shRNA driven by the U6 Pol III promoter, along with transposase mRNA can achieve a high degree of transgenesis (> 50%) along with efficient gene knockdown and phenotypic reproducibility (unpublished data).

7. Summary Phenotype-driven analysis in the mouse has been enabled by the pioneering efforts of a small cohort of mouse geneticists coupled with the more recent application of genomic analysis to this approach by investigators with specific interests in mammalian biology. The technique has lived up to its promise as a means to provide novel and unexpected insight into the role of specific genes in a myriad of biological processes. Continuing progress in genomic technology makes this method even more compelling, and it will serve as a productive complement to genotype-driven strategies employing, for example, knockout mice. As many of the components of a mutagenesis approach have been well proven, the key to its success is primarily the selection of a phenotype screen that is definitive and applicable to high-throughput analysis. In addition, careful attention to the logistic task of characterizing large numbers of mice and pursuing genetic analysis in the subset of phenotypic interest will facilitate a productive effort.

REFERENCES Acevedo-Arozena, A., Wells, S., Potter, P., Kelly, M., Cox, R. D., and Brown, S. D. (2008). ENU mutagenesis, a way forward to understand gene function. Annu. Rev. Genomics Hum. Genet. 9, 49–69. Aigner, B., Rathkolb, B., Klaften, M., Sedlmeier, R., Klempt, M., Wagner, S., Michel, D., Mayer, U., Klopstock, T., de Angelis, M. H., and Wolf, E. (2009a). Generation of N-ethyl-N-nitrosourea-induced mouse mutants with deviations in plasma enzyme activities as novel organ-specific disease models. Exp. Physiol. 94, 412–421. Aigner, B., Rathkolb, B., Klempt, M., Wagner, S., Michel, D., de Angelis, M. H., and Wolf, E. (2009b). N-Ethyl-N-nitrosourea mutagenesis produced a small number of mice with altered plasma electrolyte levels. J. Biomed. Sci. 16, 53. Beckstead, W. A., Bjork, B. C., Stottmann, R. W., Sunyaev, S., and Beier, D. R. (2008). SNP2RFLP: a computational tool to facilitate genetic mapping using benchtop analysis of SNPs. Mamm. Genome 19, 687–690. Bialek, P., Kern, B., Yang, X., Schrock, M., Sosic, D., Hong, N., Wu, H., Yu, K., Ornitz, D. M., Olson, E. N., Justice, M. J., and Karsenty, G. (2004). A twist code determines the onset of osteoblast differentiation. Dev. Cell 6, 423–435.

346

Rolf W. Stottmann and David R. Beier

Boles, M. K., Wilkinson, B. M., Wilming, L. G., Liu, B., Probst, F. J., Harrow, J., Grafham, D., Hentges, K. E., Woodward, L. P., Maxwell, A., Mitchell, K., Risley, M. D., et al. (2009). Discovery of candidate disease genes in ENU-induced mouse mutants by large-scale sequencing, including a splice-site mutation in nucleoredoxin. PLoS Genet. 5, e1000759. Caspary, T., Larkins, C. E., and Anderson, K. V. (2007). The graded response to Sonic Hedgehog depends on cilia architecture. Dev. Cell 12, 767–778. Chiang, C., Litingtung, Y., Lee, E., Yong, K. E., Corden, J. L., Westphal, H., and Beachy, P. A. (1996). Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 383, 407–413. Coghill, E. L., Hugill, A., Parkinson, N., Davison, C., Glenister, P., Clements, S., Hunter, J., Cox, R. D., and Brown, S. D. (2002). A gene-driven approach to the identification of ENU mutants in the mouse. Nat. Genet. 30, 255–256. Cook, M. C., Vinuesa, C. G., and Goodnow, C. C. (2006). ENU-mutagenesis: Insight into immune function and pathology. Curr. Opin. Immunol. 18, 627–633. Crabbe, J. C., Wahlsten, D., and Dudek, B. C. (1999). Genetics of mouse behavior: Interactions with laboratory environment. Science 284, 1670–1672. Endoh-Yamagami, S., Evangelista, M., Wilson, D., Wen, X., Theunissen, J. W., Phamluong, K., Davis, M., Scales, S. J., Solloway, M. J., de Sauvage, F. J., and Peterson, A. S. (2009). The mammalian Cos2 homolog Kif7 plays an essential role in modulating Hh signal transduction during development. Curr. Biol. 19, 1320–1326. Ermakov, A., Stevens, J. L., Whitehill, E., Robson, J. E., Pieles, G., Brooker, D., Goggolidou, P., Powles-Glover, N., Hacker, T., Young, S. R., Dear, N., Hirst, E., et al. (2009). Mouse mutagenesis identifies novel roles for left-right patterning genes in pulmonary, craniofacial, ocular, and limb development. Dev. Dyn. 238, 581–594. Favor, J., Neuhauser-Klaus, A., and Ehling, U. H. (1991). The induction of forward and reverse specific-locus mutations and dominant cataract mutations in spermatogonia of treated strain DBA/2 mice by ethylnitrosourea. Mutat. Res. 249, 293–300. Gailus-Durner, V., Fuchs, H., Becker, L., Bolle, I., Brielmeier, M., Calzada-Wack, J., Elvert, R., Ehrhardt, N., Dalke, C., Franz, T. J., Grundner-Culemann, E., Hammelbacher, S., et al. (2005). Introducing the German Mouse Clinic: Open access platform for standardized phenotyping. Nat. Methods 2, 403–404. Garcia-Garcia, M. J., Eggenschwiler, J. T., Caspary, T., Alcorn, H. L., Wyler, M. R., Huangfu, D., Rakeman, A. S., Lee, J. D., Feinberg, E. H., Timmer, J. R., and Anderson, K. V. (2005). Analysis of mouse embryonic patterning and morphogenesis by forward genetics. Proc. Natl. Acad. Sci. USA 102, 5913–5919. Georgel, P., Du, X., Hoebe, K., and Beutler, B. (2008). ENU mutagenesis in mice. Methods Mol. Biol. 415, 1–16. Godinho, S. I., and Nolan, P. M. (2006). The role of mutagenesis in defining genes in behaviour. Eur. J. Hum. Genet. 14, 651–659. Hardisty-Hughes, R. E., Parker, A., and Brown, S. D. (2010). A hearing and vestibular phenotyping pipeline to identify mouse mutants with hearing impairment. Nat. Protoc. 5, 177–190. Hart, A. W., McKie, L., Morgan, J. E., Gautier, P., West, K., Jackson, I. J., and Cross, S. H. (2005). Genotype–phenotype correlation of mouse pde6b mutations. Invest. Ophthalmol. Vis. Sci. 46, 3443–3450. Herron, B. J., Lu, W., Rao, C., Liu, S., Peters, H., Bronson, R. T., Justice, M. J., McDonald, J. D., and Beier, D. R. (2002). Efficient generation and mapping of recessive developmental mutations using ENU mutagenesis. Nat. Genet. 30, 185–189. Hrabe de Angelis, M. H., Flaswinkel, H., Fuchs, H., Rathkolb, B., Soewarto, D., Marschall, S., Heffner, S., Pargent, W., Wuensch, K., Jung, M., Reis, A., Richter, T., et al. (2000). Genome-wide, large-scale production of mutant mice by ENU mutagenesis. Nat. Genet. 25, 444–447.

Phenotype-Driven ENU Screens

347

Huangfu, D., Liu, A., Rakeman, A. S., Murcia, N. S., Niswander, L., and Anderson, K. V. (2003). Hedgehog signalling in the mouse requires intraflagellar transport proteins. Nature 426, 83–87. Int’l_Mouse_Knockout_Consortium (2007). A mouse for all reasons. Cell 128, 9–13. Justice, M. J., Carpenter, D. A., Favor, J., Neuhauser-Klaus, A., Hrabe de Angelis, M., Soewarto, D., Moser, A., Cordes, S., Miller, D., Chapman, V., Weber, J. S., Rinchik, E. M., et al. (2000). Effects of ENU dosage on mouse strains. Mamm. Genome 11, 484–488. Kasarskis, A., Manova, K., and Anderson, K. V. (1998). A phenotype-based screen for embryonic lethal mutations in the mouse. Proc. Natl. Acad. Sci. USA 95, 7485–7490. Kauppi, M., Murphy, J. M., de Graaf, C. A., Hyland, C. D., Greig, K. T., Metcalf, D., Hilton, A. A., Nicola, N. A., Kile, B. T., Hilton, D. J., and Alexander, W. S. (2008). Point mutation in the gene encoding p300 suppresses thrombocytopenia in Mpl/ mice. Blood 112, 3148–3153. Lewis, M. A., Quint, E., Glazier, A. M., Fuchs, H., De Angelis, M. H., Langford, C., van Dongen, S., Abreu-Goodger, C., Piipari, M., Redshaw, N., Dalmay, T., MorenoPelayo, M. A., et al. (2009). An ENU-induced mutation of miR-96 associated with progressive hearing loss in mice. Nat. Genet. 41, 614–618. Liem, K. F., Jr., He, M., Ocbina, P. J., and Anderson, K. V. (2009). Mouse Kif7/Costal2 is a cilia-associated protein that regulates Sonic hedgehog signaling. Proc. Natl. Acad. Sci. USA 106, 13377–13382. Mandillo, S., Tucci, V., Holter, S. M., Meziane, H., Banchaabouchi, M. A., Kallnik, M., Lad, H. V., Nolan, P. M., Ouagazzal, A. M., Coghill, E. L., Gale, K., Golini, E., et al. (2008). Reliability, robustness, and reproducibility in mouse behavioral phenotyping: A cross-laboratory study. Physiol. Genomics 34, 243–255. Mar, L., Rivkin, E., Kim, D. Y., Yu, J. Y., and Cordes, S. P. (2005). A genetic screen for mutations that affect cranial nerve development in the mouse. J. Neurosci. 25, 11787–11795. Matera, I., Watkins-Chow, D. E., Loftus, S. K., Hou, L., Incao, A., Silver, D. L., Rivas, C., Elliott, E. C., Baxter, L. L., and Pavan, W. J. (2008). A sensitized mutagenesis screen identifies Gli3 as a modifier of Sox10 neurocristopathy. Hum. Mol. Genet. 17, 2118–2131. May, S. R., Ashique, A. M., Karlen, M., Wang, B., Shen, Y., Zarbalis, K., Reiter, J., Ericson, J., and Peterson, A. S. (2005). Loss of the retrograde motor for IFT disrupts localization of Smo to cilia and prevents the expression of both activator and repressor functions of Gli. Dev. Biol. 287, 378–389. McDonald, J. D., and Bode, V. C. (1988). Hyperphenylalaninemia in the hph-1 mouse mutant. Pediatr. Res. 23, 63–67. McDonald, J., Shedlovsky, A., and Dove, W. (1990). Investigating inborn errors of phenylalanine metabolism by efficient mutagenesis of the mouse germline. In Banbury Report 34: Biology of Mammalian Germ Cell Mutagenesis Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Montagutelli, X., Lalouette, A., Coude, M., Kamoun, P., Forest, M., and Guenet, J. L. (1994). aku, a mutation of the mouse homologous to human alkaptonuria, maps to chromosome 16. Genomics 19, 9–11. Moran, J. L., Bolton, A. D., Tran, P. V., Brown, A., Dwyer, N. D., Manning, D. K., Bjork, B. C., Li, C., Montgomery, K., Siepka, S. M., Vitaterna, M. H., Takahashi, J. S., et al. (2006). Utilization of a whole genome SNP panel for efficient genetic mapping in the mouse. Genome Res. 16, 436–440. Moser, A. R., Pitot, H. C., and Dove, W. F. (1990). A dominant mutation that predisposes to multiple intestinal neoplasia in the mouse. Science 247, 322–324.

348

Rolf W. Stottmann and David R. Beier

Nolan, P. M., Peters, J., Strivens, M., Rogers, D., Hagan, J., Spurr, N., Gray, I. C., Vizor, L., Brooker, D., Whitehill, E., Washbourne, R., Hough, T., et al. (2000a). A systematic, genome-wide, phenotype-driven mutagenesis programme for gene function studies in the mouse. Nat. Genet. 25, 440–443. Nolan, P. M., Peters, J., Vizor, L., Strivens, M., Washbourne, R., Hough, T., Wells, C., Glenister, P., Thornton, C., Martin, J., Fisher, E., Rogers, D., et al. (2000b). Implementation of a large-scale ENU mutagenesis program: Towards increasing the mouse mutant resource. Mamm. Genome 11, 500–506. Nybakken, K., Vokes, S. A., Lin, T. Y., McMahon, A. P., and Perrimon, N. (2005). A genome-wide RNA interference screen in Drosophila melanogaster cells for new components of the Hh signaling pathway. Nat. Genet. 37, 1323–1332. Ocbina, P. J., and Anderson, K. V. (2008). Intraflagellar transport, cilia, and mammalian Hedgehog signaling: Analysis in mouse embryonic fibroblasts. Dev. Dyn. 237, 2030–2038. Roessler, E., Belloni, E., Gaudenz, K., Jay, P., Berta, P., Scherer, S. W., Tsui, L. C., and Muenke, M. (1996). Mutations in the human Sonic Hedgehog gene cause holoprosencephaly. Nat. Genet. 14, 357–360. Russell, W., Kelly, E., Hunsicker, P., Bangham, J., Maddux, S., and Phipps, E. (1979). Specific locus test shows ethylnitrosurea to be the most potent mutagen in the mouse. Proc. Natl. Acad. Sci. USA 76, 5818–5819. Russell, W., Hunsicker, P., Carpenter, D., Cornett, C., and Guinn, G. (1982). Effect of dose-fractionation on the ethylnitrosurea induction of specific-locus mutations in mouse spermatogonia. Proc. Natl. Acad. Sci. USA 79, 3592–3593. Sagai, T., Masuya, H., Tamura, M., Shimizu, K., Yada, Y., Wakana, S., Gondo, Y., Noda, T., and Shiroishi, T. (2004). Phylogenetic conservation of a limb-specific, cisacting regulator of Sonic hedgehog (Shh). Mamm. Genome 15, 23–34. Shima, N., Hartford, S. A., Duffy, T., Wilson, L. A., Schimenti, K. J., and Schimenti, J. C. (2003). Phenotype-based identification of mouse chromosome instability mutants. Genetics 163, 1031–1040. Silver, J. D., Hilton, D. J., Bahlo, M., and Kile, B. T. (2007). Probabilistic analysis of recessive mutagenesis screen strategies. Mamm. Genome 18, 5–22. Tran, P. V., Haycraft, C. J., Besschetnova, T. Y., Turbe-Doan, A., Stottmann, R. W., Herron, B. J., Chesebro, A. L., Qiu, H., Scherz, P. J., Shah, J. V., Yoder, B. K., and Beier, D. R. (2008). THM1 negatively modulates mouse sonic hedgehog signal transduction and affects retrograde intraflagellar transport in cilia. Nat. Genet. 40, 403–410. Tutois, S., Montagutelli, X., Da Silva, V., Jouault, H., Rouyer-Fessard, P., Leroy-Viard, K., Guenet, J. L., Nordmann, Y., Beuzard, Y., and Deybach, J. C. (1991). Erythropoietic protoporphyria in the house mouse. A recessive inherited ferrochelatase deficiency with anemia, photosensitivity, and liver disease. J. Clin. Invest. 88, 1730–1736. Vitaterna, M., King, D., Chang, A., Kornhauser, J., Lowrey, P., McDonald, J., Dove, W., Pinto, L., Turck, F., and Takahashi, J. (1994). Mutagenesis and mapping of a mouse gene, clock, essential for cricadian behavior. Science 264, 716–719. Weatherbee, S. D., Niswander, L. A., and Anderson, K. V. (2009). A mouse model for Meckel syndrome reveals Mks1 is required for ciliogenesis and Hedgehog signaling. Hum. Mol. Genet. 18, 4565–4575. Weber, J. S., Salinger, A., and Justice, M. J. (2000). Optimal N-ethyl-N-nitrosourea (ENU) doses for inbred mouse strains. Genesis 26, 230–233. Zarbalis, K., May, S. R., Shen, Y., Ekker, M., Rubenstein, J. L., and Peterson, A. S. (2004). A focused and efficient genetic screening strategy in the mouse: Identification of mutations that disrupt cortical development. PLoS Biol. 2, E219.

C H A P T E R

E I G H T E E N

Exploration of Self-Renewal and Pluripotency in ES Cells Using RNAi Christoph Schaniel,*,1 Dung-Fang Lee,*,1 Foster C. Gonsalves,†,1 Ramanuj DasGupta,† and Ihor R. Lemischka* Contents 1. Introduction 2. Maintenance of Mouse Embryonic Stem Cells 3. siRNA-Mediated Gene Knockdown 3.1. siRNA transfection 3.2. Preparation of plates for image acquisition 3.3. Automated image scanning and quantitative image analysis 4. Lentivirus-Based shRNA Knockdown System 4.1. Generation of pLKO.puro-IRES-eGFP (pLKO.pig) lentiviral vector 4.2. Design and cloning of shRNA expression cassette into pLKO.pig lentiviral vector 4.3. Generation and concentration of lentiviral particles 4.4. Infection of mouse ESCs 4.5. Quantification of shRNA-mediated gene silencing efficiency 5. Concluding Remarks Acknowledgments References

352 353 354 356 357 357 358 359 360 361 362 363 364 364 364

Abstract Embryonic stem cells (ESCs) have the ability to expand indefinitely in vitro and give rise to cells of all three germ layers as well as germ cells. For these reasons, ESCs hold great promise for biomedicine. In order to harness the potential of pluripotent cells, it is necessary to first understand the molecular mechanisms that control the pluripotent state. The discovery of RNA interference has made * Department of Gene and Cell Medicine, Black Family Stem Cell Institute, Mount Sinai School of Medicine, New York, USA New York University Cancer Institute and the Department of Pharmacology, The Helen and Martin Kimmel Center for Stem Cell Biology, New York University Langone Medical Center, New York 1 These authors contributed equally to the work. {

Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77018-X

#

2010 Elsevier Inc. All rights reserved.

351

352

Christoph Schaniel et al.

such functional analysis, even at high(er) throughput, possible. Here, we describe the methods used for high-throughput siRNA screening by high-content microscopy to identify gene products that regulate mouse ESC fate decision. In addition, we will describe the application of lentivirus-based shRNA knockdown to explore or validate the role of candidate genes in ESC pluripotency.

1. Introduction Embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) have the ability to self-renew and differentiate into cells of all three germ layers. These characteristics make them invaluable for studying the molecular mechanisms of self-renewal and lineage differentiation processes. Over the last several years many groups, including ours, have analyzed and cataloged the gene-expression profiles of pluripotent stem cells. In order to investigate the function of the genes assembled in these lists, one can either pursue the classical strategy of gene deletion or use RNA silencing approaches. Although gene deletion will result in complete null phenotypes, the strategy is hampered by the fact that it is rather timeconsuming and not feasible for high-throughput analyses. The discovery of RNA interference (RNAi), a natural process through which the expression of target genes can get silenced by either suppressing translation or by site-specific cleavage followed by degradation of the target mRNA (Fire et al., 1998; Meister and Tuschl, 2004), has opened the door for high(er)throughput functional genomics analyses in organism previously inaccessible to such systemic genetic perturbations. Several reagents have been developed to induce RNAi. In mammalian cells, the most common are chemically synthesized short double stranded (ds) RNA known as small interfering RNAs (siRNAs), which share 100% sequence complementarity to their target mRNA (Elbashir et al., 2001). siRNA libraries targeting the entire genome, or specific functional gene subsets thereof, can be obtained from several commercial vendors. These libraries/sets contain mostly two to four siRNAs per gene thereby minimizing off-target effects and maximizing knockdown efficiency. As an alternative to these chemical libraries, long dsRNAs, generated by in vitro transcription of full-length cDNA (libraries), can be enzymatically cleaved in vitro by Dicer or RNase III into multiple endo-RNase prepared siRNAs (esiRNA) per gene target (Buchholz et al., 2006; Yang et al., 2002). Once introduced into cells, these (e)siRNAs are unwound by the activated RNA-induced silencing complex (RISC) in an ATP-consuming process. Only one strand of the (e)siRNA, which is selected based on thermodynamic rules, is then incorporated into RISC. This antisense strand then directs recognition and cleavage of the target mRNA sequences by RISC (Hannon and Rossi, 2004). The cleaved

RNAi in ES Cells

353

(e)siRNA target mRNA is subsequently degraded resulting in a loss of gene function (Hannon and Rossi, 2004). Both siRNA- and esiRNA-type silencing approaches can be delivered into cells using a variety of standard transfection protocols and allow for transient, 3–5 days knockdown of the target genes. For stable, long-term silencing of target mRNAs systems encoding for short hairpin RNAs (shRNAs), mimicing microRNAs, the endogenous triggers of the RNAi pathway, have been developed (Brummelkamp et al., 2002; McCaffrey et al., 2002; McManus et al., 2002; Paddison et al., 2002; Paul et al., 2002; Sui et al., 2002; Xia et al., 2002; Yu et al., 2002). The shRNAs can be expressed from polymerase II or III promoters, using constitutive or inducible systems, vary in length from 19 to 29 nucleotides with diverse stem loops structures, and with various degrees of similarity to natural microRNAs. Because these shRNA triggers are encoded by DNA plasmids, they can be delivered to cells in many ways, including standard transient transfection (not feasible for long-term silencing), stable transfection, and transduction using viruses including adenoviruses and retroviruses. In this chapter, we briefly describe the experimental procedures to study gene function in mouse ESCs by siRNA transfection and/or lentiviralbased shRNA methods.

2. Maintenance of Mouse Embryonic Stem Cells The maintenance of undifferentiated, healthy ESCs is of critical importance to any successful screen. For large-scale screens, it is detrimental that enough cells of the same passage number are frozen away and enough reagents of the same lot number are available to last through the entire experiment from screen setup through optimization to the actual screen. We maintain our ESCs as follows: 1. Precoat 10-cm cell culture plates with 5–7 ml EmbryoMax ESC qualified 0.1% gelatin solution (Millipore; ES-006-B) for at least 15 min prior to use. 2. Mouse CCE and NG4 (CCE cells carrying a green fluorescent protein (GFP) expression cassette driven by the Nanog regulatory loci; Schaniel et al., 2006, 2009) cells are maintained in feeder-free tissue culture condition. ESC culture medium consists of Dulbecco’s modified Eagle medium with high glucose (DMEM high glucose; Invitrogen; Cat. no. 11965) supplemented with 15% qualified fetal bovine serum (FBS; we currently use Stasis stem cell FBS (Gemini Bio-Products; Cat. no. 100-125)), 2 mM L-glutamine (Invitrogen; Cat. no. 25030), 1 mM sodium pyruvate (Invitrogen; Cat. no. 11360), MEM nonessential amino acids (MEM NEAA; Invitrogen; Cat. no. 11140), 100 U/ml penicillin, and 100 mg/ml streptomycin (Pen/Strep; Invitrogen;

354

3. 4. 5. 6. 7. 8.

9.

Christoph Schaniel et al.

Cat. no. 15140), 1
b-mercaptoethanol (BME; preparing 100
BME stock by adding 37 ml BME to 100 ml Dulbecco’s phosphate-buffered saline (DPBS); Sigma; M7522), and 1000 U/ml recombinant mouse LIF (ESGRO; Millipore; ESG1107). Aspirate medium from the cell culture dish and rinse cells with DPBS (Invitrogen; Cat. no. 14287072). Aspirate DPBS and harvest cells by incubation with 1 ml trypsin for 3–5 min to allow detachment of cells from the coated tissue culture plates. Add 10 ml ESC culture medium w/o LIF to inactivate trypsin, pipet gently, and transfer to a 15-ml BD Falcon conical tube (BD Biosciences; Cat. no. 252097). Centrifuge at 300
g for 3–5 min, discard supernatant, and resuspend cells in 10 ml ESC culture medium. Remove gelatin from 10-cm tissue culture plates and replace with 10 ml of ESC culture medium. Add 1 ml of resuspended cells per gelatin-precoated tissue culture dishes with 9 ml ESC culture medium and distribute cells evenly over the plate. Alternatively, count cells using a hemocytometer and add 2–2.5
106 cells per 10-cm plate. Cells are maintained at 37  C in a humidified environment containing 5% CO2. Change medium every day and split cells every two days.

3. siRNA-Mediated Gene Knockdown RNAi using (e)siRNAs is a powerful tool to study gene function. However, the analysis is limited to 3–5 days due to the transient inhibitory nature of (e)siRNAs. In 2009, two genome-wide RNAi screens—one using siRNA and the other using esiRNA—identified several novel regulators defining mouse ESC identity (Ding et al., 2009; Hu et al., 2009). Both these studies relied on flow cytometric analysis of Oct4 promoterdriven GFP expression. Before any screening can proceed, that is, irrespective whether a single gene, several genes, gene families, or the entire genome is to be analyzed, several parameters need to be established and optimized. This part of the screening procedure is the most time-consuming and can take several months to complete. The parameters to establish are (1) choice of (reporter) cell line and read-out, (2) appropriate cell plating density for your particular assay, (3) positive and negative controls to be used, (4) most optimal transfection reagent (depends on siRNA library used and the cell type to be transfected), (5) concentration of siRNA to be used, and (6) concentration of transfection reagent to be used with the optimal concentration of siRNA. Once these parameters are set even a genome-wide screen can be completed within a few weeks.

355

Instances of average intensity of GEP

Instances of average intensity of GEP

RNAi in ES Cells

225 200

0 mM RA

175 150 125 100 75 50 20 0

0

600

100 200 300 400 500 600 700 800 900 1000 Ave intensity of GEP

2 mM RA

500 400 300 200 100 0 0

100 200 300 400 500 600 700 800 900 1000 Ave intensity of GEP

Figure 18.1 Detection of GFP expression using an ArrayScan IV by Cellomics. Control siRNA transfected NG4 cells are treated with 2 mM RA for 2 days. Cells were further fixed with 4% paraformaldehyde, stained with an anti-GFP antibody and FITC-labeled secondary antibody, scanned, and analyzed using Cellomics software.

Here we present a high-throughput, 384-well plate format screening method using NG4 (Nanog-GFP reporter) cells (Schaniel et al., 2006, 2009) to analyze the effects of siRNA treatments on Nanog-promoter activity (see Fig. 18.1 for representative images). We monitor GFP expression, our readout for Nanog-promoter activity, using fully automated high-content and high-throughput microscopy, such as the Arrayscan VTI (Cellomics, Thermo Scientific) or the Image Xpress Ultra Confocal Imaging System (Molecular Devices). For high-throughput analysis, the scanning system should be combined with a plate-moving robot and plate capacity hotel(s) that can accommodate storage of multiple plates to allow for fully automated unattended plate feeding into the imaging device.

356

Christoph Schaniel et al.

3.1. siRNA transfection It should be noted that dispersion of any liquid, including those containing cells, is best accomplished using automatic or semiautomatic liquid handlers and dispensers such as the Perkin Elmer JANUS eight-channel liquid handler, the Thermo Combi liquid dispenser or the WellMate microplate dispenser (Thermo Scientific). A critical point to determine before starting the screen is the capacity of daily plate handling. This will depend largely on the number of siRNAs to be screened, the number of total plates (including replicas), and the capacity of the liquid handler/dispenser and imaging system. One should not forget to take into account the number of new siRNA transfections one is to do the following day(s), the daily media change on plates previously transfected, the time required for preparation of plates for imaging and the image scanning itself. We have established the following protocol that allows us to screen for genes involved in both the activation of the Nanog promoter (surrogate for the maintenance of ESCs) or the repression of Nanog (surrogate for genes required for ESC differentiation). 1. Maintain mouse NG4 ESCs in regular mouse ESC medium as described above. The day before use, harvest NG4 cells and plate at a 1:10 dilution onto gelatin-coated 10-cm tissue culture plates. Grow overnight. 2. On the day of transfection, precoat 384-well tissue culture plates (Corning Life Sciences) with 50 ml/well of EmbryoMax ESC qualified 0.1% gelatin solution at 37  C for 15 min. Aspirate gelatin. 3. Add 5 pmol siRNAs/well. This process is done by automated liquid handlers such as the JANUS MDT (Perkin Elmer) workstation. 4. Dilute 0.15 ml of Lipofectamine 2000 transfection reagent (Invitrogen; Cat. no. 11668-019) in 25 ml DMEM high glucose medium per siRNA, mix and incubate at room temperature for 5 min. 5. Add Lipofectamine 2000 mixture to each well. Note: We use a different tubing to dispense the transfection reagents. Vortex plate gently and centrifuge at 300
g for 1 min, followed by incubation at room temperature for 15 min to allow for formation of siRNA-liposome complexes. 6. In the meantime, harvest mouse ESCs as described above, and resuspend in mouse ESC medium containing 2
FBS but no antibiotics. Count cell number and prepare a cell stock of 30 cells/ml. 7. Dispense 25 ml of cell stock—a total of 750 cells—into each well, vortex gently, centrifuge at 300
g for 1 min, and incubate at 37  C in a humidified environment containing 5% CO2. 8. Twenty-four hours after transfection, carefully aspirate the medium and replace it with fresh medium. (Note: We have used ESC medium minus LIF, supplemented with 0.5 mM retinoic acid (ESC medium þ RA). This allows us to identify genes required for both Nanog-promoter activation as well as repression.) 9. Replace medium again with fresh medium after 24 h.

RNAi in ES Cells

357

3.2. Preparation of plates for image acquisition The process of image acquisition is more time-consuming than the process of siRNA transfection or the daily medium changes. The time needed mainly depends on the imaging system used, the sparsity of cells, and the amount of parameters to be collected (see also below). We have found that the scanning of one 384-well plate can take up to 2 h. Therefore, cells ought to be fixed. Fixation allows for plate storage, scanning and rescanning should something go wrong during the original acquisition without loss in cellular phenotype. The following describes the process of cell fixation. 1. Seventy-two hours after initial transfection, aspirate culture medium and carefully wash cells three times with 50 ml of DPBS. 2. Add 30 ml of 4% paraformaldehyde (Electron Microscopy Sciences; Cat. no. RT-15710) (prepared in DPBS) and fix cells at room temperature for 30 min. 3. Aspirate paraformaldehyde solution and wash cells again three times with 50 ml of DPBS. 4. Permeabilize and block cells with 50 ml of blocking buffer (0.1% Triton X-100, 1% BSA, 10% normal donkey serum in DPBS) at room temperature for 30 min. 5. If needed, the GFP signal-to-background/noise ratio can be increased by staining the cells with an anti-GFP antibody. Otherwise go to step 9. Dilute anti-GFP rabbit antibody (Invitrogen; Cat. no. A-6455) 1:1000 in blocking buffer, add 50 ml/well, and stain at 4  C overnight. 6. Remove primary antibody and wash wells three times with DPBS at room temperature. 7. Dilute anti-rabbit IgG-Alexa488 antibody (from any commercial vendor) 1:1000 in blocking buffer, add 50 ml/well, and incubate at room temperature for 1 h. 8. Remove secondary antibody and wash wells three times with DPBS at room temperature. 9. Add either 50 ml of DAPI (1 mg/ml) or Hoechst 33342 (5 mg/ml) at room temperature for 5–15 min (accurate time needs to be determined; this varies on the cells used for the screen and the imaging system) followed by a wash with DPBS. 10. Add 50 ml DPBS:glycerol (50%:50%) per well, seal plates, and store plates at 4  C until image scanning.

3.3. Automated image scanning and quantitative image analysis Image scanning can be performed on a number of devices equipped to handle high-throughput and high-content microscopy, such as the Arrayscan VTI (Cellomics, Thermo Scientific) or the Image Xpress Ultra

358

Christoph Schaniel et al.

Confocal Imaging System (Molecular Devices). However, before highthroughput and high-content image scanning of any screen trade-offs need to be made regarding the number of optical planes, number of fields per well, channels acquired per exposure, the number of cells required to reasonably determine a significant effect on the cell population, and the time required for acquisition. Since image data are extremely information rich adequate storage capacity in the terabyte range, data backup and database tools are a crucial prerequisite. Automated image scanning is key to reproducible and reliable quantitative measurements in high-throughput and high-content microscopy. Quantitative image analysis is typically done either by image segmentation, by supervised learning algorithms or a combination of the two methods. Most commercial image analysis software, such as ImageXpress (Molecular Devices), MetaXpress (Molecular Devices), or BioApplications (specific modules for distinct applications, Cellomics, Thermo Scientific), come with various modules that allow for image segmentation and pattern recognition, including determination of cell and nuclear boundaries and quantification of fluorescence in predetermined regions. For a more sophisticated discussion of ‘‘computational image analysis for quantitative phenotyping’’ and ‘‘statistical methods for analysis of high-throughput RNAi’’ read the reviews by Conrad and Gerlich, The RNAi Global Initiative, and references therein (Birmingham et al., 2009; Conrad and Gerlich, 2010). Although the rigorous parameter evaluation and optimization process should eliminate that cell boundaries of neighboring cells cannot be easily determined and distinguished, we have found the following method quite useful to work around such a problem. Nuclei, which are surrounded by cytoplasm, can easily be distinguished and marked, therefore identification of cytoplasmic regions around the nucleus can be inferred by using a fixed radius/pixel size spanning from the nuclear boundary. This approach allows for the gathering of fluorescent information of cytoplasmic reporters/markers used for subsequent quantification and analysis. This, however, comes at the cost of reduced signal since one excludes fluorescence in cytoplasmic regions outside this arbitrarily set radius.

4. Lentivirus-Based shRNA Knockdown System The lentiviral-based shRNA technology has two important advantages over siRNA or esiRNA approaches. It allows for long-term gene repression as well as transduction of relative difficult to transfect cells, such as quiescent hematopoietic stem cells or primary neurons. Here, rather than describing large-scale shRNA screening approaches, we focus on shRNA as a tool for functional analysis on a gene-per-gene basis and as a confirmation strategy to validate or disproof hits obtained during siRNA or esiRNA screenings (Ivanova et al., 2006).

359

RNAi in ES Cells

4.1. Generation of pLKO.puro-IRES-eGFP (pLKO.pig) lentiviral vector

Age I

A

Eco RI

The pLKO.puro cloning vector (Fig. 18.2A), generated by the Broad Institute of MIT and Harvard, has been tested in a broad range of cells (Moffat et al., 2006). This vector contains (1) the human U6 promoter, which drives expression of the shRNA transcript; (2) the human phosphoglycerate kinase (hPGK) promoter, which drives expression of the puromycin resistance gene (PuroR); (3) a central polypurine tract (cPPT), which facilitates nuclear import of the vector preintegration complex thereby increasing transduction efficiency. The shRNA duplex is cloned between AgeI and EcoRI restriction sites (Fig. 18.2). This vector is a replicationincompetent lentiviral vector and can be used to knockdown gene expression by direct transfection or by infection after conversion into lentiviral particles. In order to easily evaluate infection efficiency and monitor the

U6

shRNA

cPPT hPGK

PuroR

Nco I

U6

Nco I

Age I

B

Eco RI

pLKO.puro (7032 bp)

shRNA cPPT hPGK

PuroR

pLKO.pig (8346 bp)

IRES eGFP

Nco I

Figure 18.2 Schematic of pLKO.puro and pLKO.pig lentiviral vectors used in our studies. Both pLKO.puro (A) and pLKO.pig (B) vectors contain a U6 promoter driving shRNA, a cPPT, and two multiple cloning sites (AgeI and EcoRI) for inserting annealed shRNA oligos (Fig. 18.1). While the hPGK promoter drives expression of PuroR in pLKO.puro (A), it drives expression of the PuroR-IRES-eGFP cassette in pLKO.pig (B). The AgeI, EcoRI, and NcoI restriction enzyme sites used for shRNA cloning and diagnostic analysis are indicated.

360

Christoph Schaniel et al.

potential phenotypic changes after transfection or infection, we modified the original pLKO.puro plasmid and created pLKO.pig by inserting the internal ribosome entry site (IRES)-enhanced green fluorescent protein (eGFP) cassette after PuroR (Fig. 18.2B). This modification allows us to select transduced cells by PuroR selection and/or monitor eGFP expression.

4.2. Design and cloning of shRNA expression cassette into pLKO.pig lentiviral vector There are several useful web sites to predict a suitable region in your interested gene for shRNA targeting. We opted for the shRNA tool designed by the RNAi consortium (TRC), based at the Broad Institute in Boston. TRC is a collaboration between six academic research institutions and five leading life sciences organizations aimed to generate two shRNAs/ gene targeting all human and mouse genes. The following steps will guide through the design and cloning of shRNA targeting our gene of interest into pLKO.pig. 1. Design your shRNA forward and reverse oligos using the TRC shRNA library database (http://www.broadinstitute.org/rnai/public/ gene/search). You can search either by gene or transcript IDs (NM_ or XM_ sequences) or by official gene symbols. For instance, if you search for Nanog (Official Gene Symbol) or NM_028016 (Reference Transcript ID) for the mouse Nanog gene, the database search will return five shRNAs. Each shRNA duplex consists of two different oligos, a forward oligo of 50 -CCGG-21 bp sense-CTCGAG-21 bp antisenseTTTTTG-30 ) and a reverse oligo (50 -AATTCAAAAA-21 bp senseCTCGAC-21 bp antisense-30 ). The CTCGAG sequence, an XhoI restriction enzyme site, is designed to form a loop structure between the sense and antisense strand of the shRNA. 2. Order oligos from any commercial vendor and resuspend them in RNase/DNase-free molecular grade water (MediaTech Inc., Cat. no. 46-000-CM) at a concentration of 20 mM. Prepare oligo duplex mixture by mixing of 5 ml forward oligo with 5 ml of reverse oligo, 5 ml 10
NEB buffer 2, and 35 ml ddH2O. 3. Incubate oligo duplex mixture in a beaker of boiling water for 5–10 min and then remove the breaker from the heating plate. Allow water to cool down to room temperate slowly in order to allow for the oligos to anneal. It will take several hours to cool down. Alternatively, annealing and cooling can be performed using a Thermocycler, using the cycle ramp step. Use a decline of 1  C/cycle until you reach 25  C with a cycle time of 1 min. 4. Digest pLKO.pig vector with AgeI and EcoRI. Usually, we use 6 mg pLKO.pig plasmid, 7 ml of 10
NEB buffer 1, 6 ml of AgeI (NEB,

RNAi in ES Cells

5.

6.

7. 8. 9. 10.

361

Cat. no. R0522L, 5000 U/ml), 4 ml of EcoRI (NEB; Cat. no. R0101L, 20,000 U/ml), and add ddH2O to 70 ml. Incubate the digestion reaction at 37  C for 1 h 40 min. (Note: A longer incubation time will cause overdigestion and result in lower ligation efficiency.) Run the digested pLKO.pig vector (8 kb) on a 0.8% agarose gel, purify plasmid by QIAquick gel extraction kit (Qiagen; Cat. no. 28706), and resuspend extracted plasmid in 30 ml elution buffer, which is provided in the QIAquick gel extraction kit. Ligate annealed oligos with digested vector by using 2 ml of annealed oligos, 1 ml of digested pLKO.pig plasmid, 2 ml of 10
NEB T4 DNA ligase buffer, 2 ml of NEB T4 DNA ligase (NEB, Cat. no. M0202L, 400,000 cohesive end U/ml), and 13 ml of ddH2O. Incubate the ligation product at room temperate for 2 h. Transform 5 ml of ligation product into 50 ml of DH5a competent cells and plate bacteria on LB agar containing 50 mg/ml ampicillin. Pick up the clones and isolate plasmids by QIAprep spin miniprep kit (Qiagen; Cat. no. 27106). Check plasmid by digesting with EcoRI and NcoI restriction enzyme. Clones containing a correct pLKO.pig-shRNA should have three fragments of 5059, 1980, and 1307 bp. Select positive plasmids for DNA sequencing. The sequence primer used for pLKO.pig is 50 -CAAGGCTGTTAGAGAGATAATTGGA30 , which is the same sequencing primer used for pLKO.puro. (Note: It is important to sequence your shRNA insert using a sequencing protocol that takes the hairpin configuration into consideration. We routinely sequence our shRNA plasmid using the services provided by Genewiz Inc.)

4.3. Generation and concentration of lentiviral particles Generation of lentiviral particles should be done using appropriate protective gear/clothing (lab coat gloves) in a biohazard hood and in accordance with Federal Institutional Biohazard Committee guidelines. All material and reagents that have been in contact with viral particles ought to be soaked in 10% bleach for a minimum of 20 min before being discarded in biohazard bags and autoclaved. 1. Sixteen to 24 h before transfection, plate 2
106 human embryonic kidney (HEK) 293T cells in a 10-cm tissue culture plate using 10 ml regular culture medium without antibiotics (DMEM high glucose supplemented with 10% FBS, 2 mM L-glutamine, and 1 mM sodium pyruvate). 2. On the day of transfection, carefully aspirate medium and add 3 ml fresh Opti-MEM I reduced-serum medium (Opti-MEM; Invitrogen;

362

3.

4. 5. 6. 7.

8. 9.

10.

Christoph Schaniel et al.

Cat. no. 11058-021) gently. (Note: In order to prevent detaching or washing away of 293T cells, do not wash 293T cell with DPBS.) Mix 8 mg pLKO.pig, 6 mg packaging plasmid pCMV-dR8.2 dvpr (Addgene plasmid #8455), and 2 mg VSV-G encoding plasmid pCMV-VSV-G (Addgene plasmid #8454) in 300 ml Opti-MEM. Vortex briefly and incubate DNA mixture at room temperature for 15 min. In the meantime, dilute 22 ml of SuperFect Transfection Reagent (Qiagen; Cat. no. 301305) in 300 ml Opti-MEM and incubate at room temperature for 15 min. Add DNA mixture to SuperFect and incubate at room temperature for 30 min. Add DNA–SuperFect mixture dropwise to culture plate and incubate at 37  C in a humidified environment and 5% CO2. After 3 h add 10 ml of fresh regular culture medium with antibiotics (100 U/ml penicillin and 100 mg/ml streptomycin). After an additional 12–16 h, replace medium with 10 ml fresh ESC culture medium without LIF and grow cells at 37  C, 5% CO2 in a humidified environment for an additional 36–48 h. During this time, cells will produce and secrete lentiviral particles. Collect viral supernatant and remove cell debris by filtration through a 0.45 mm low protein binding filter (Corning Life Sciences, Cat. no. 0976155). Filtered viral particles can be used immediately or concentrated. For concentration, add viral supernatant to the top portion of an Amicon Ultra-15 Centrifuge filter units with ultracel membrane (Millipore; Cat. no. UFC903024) and concentrate viral volume to 0.5 ml by spinning at 1600
g at 4  C for 15–20 min. Use viral particles either right away or store at 80  C until use.

4.4. Infection of mouse ESCs The same safety rules described above apply here. 1. Seed 5
104 CCE cells into a well of the gelatin-precoated 12-well tissue culture plate and grow for 12–16 h in a humidified environment at 37  C/5% CO2. 2. Cells should reach approximately 50–60% confluency in the 12–16 h of grow. Only then aspirate medium and replace with 0.5 ml fresh ESC culture media containing LIF and 16 mg/ml polybrene (officially known as hexadimethrine bromide, Sigma; Cat. no. H9268). 3. Infect cells by addition of 0.5 ml of concentrated viruses. The final polybrene concentration is 8 mg/ml. 4. Twenty-four to 26 h postinfection, replace media with 1 ml fresh ESC culture medium.

363

RNAi in ES Cells

5. The next day (day 2 postinfection), split cells into 1-well of a gelatinprecoated 6-well plate in 2 ml of total ESC medium and continue culture for 2 days. Change medium daily. 6. On day 4 postinfection, split cells at a ratio of 1:10 into 1-well of a 6-well plate and culture for an additional 2 days with daily medium changes. 7. Six days after infection, select lentivirus-infected cells by replacing medium with ESC medium supplemented with 2 mg/ml puromycin (Sigma; Cat. no. P8833) (Fig. 18.3). (Note: The optimal puromycin concentration needs to be predetermined for each cell type/line. We find that 2 mg/ml puromycin kills all CCE and NG4 mouse ESCs within 2–3 days.) 8. Replace medium with fresh puromycin-containing medium for the next 2 days. Ideally, all surviving cells should be GFP positive, which also indicates that every selected cells harbor the shRNA expression cassette.

4.5. Quantification of shRNA-mediated gene silencing efficiency It is important to determine the efficiency of the lentivirus-mediated mRNA knockdown in order to evaluate the physiological effect of silencing of your gene of interest. Several distinct methods can be applied to verify knockdown efficiency. These include real-time RT-PCR, Northern blotting, and Western blotting analyses. In order to exclude possible off-target effects of the designed shRNA, one may design two or more different shRNAs targeting a single target. Similar phenotypic changes by these individual shRNAs targeting the same gene, including similar changes in gene-expression profiles, reduce the likelihood of an off-target effect. The best way to exclude possible off-target effects is to reexpress an shRNAimmune version of the target gene which should rescue gene expression and Phase

GFP

Figure 18.3 ESCs infected with shRNA-expressing pLKO.pig lentiviruses. Mouse CCE ESCs were infected with pLKO.pig lentiviral particles and selected with puromycin (2 mg/ml). After puromycin selection, almost 100% of cells express GFP.

364

Christoph Schaniel et al.

the phenotypic changes (Ivanova et al., 2006). The straightforward rescue strategy of overexpressing the coding sequence (CDS) can be applied if the shRNA targets untranslated regions in the target mRNA. Should the shRNA however target the CDS, it will be necessary to generate a rescue clone containing silent mutations in the region of the CDS targeted by the shRNA.

5. Concluding Remarks The last few years have revealed many novel insights into the regulation of ESC self-renewal and pluripotency using high-throughput technologies, including global mRNA expression analysis, protein abundance measurements using mass spectrometry, genome-wide mapping of protein–DNA interaction and histone modifications. The wealth of this information provides us with a ‘‘parts list’’ or signature of pluripotent stem cells. With the discovery of RNAi and the continued development of improved RNAi systems, the functional analysis of genes has become relatively easy. RNAi also enables us to perform for the first time, unbiased screens targeting gene families or the entire genome in a short period of time. Finally, the use of siRNA or esiRNA, and shRNAs should be viewed as complementary approaches in the application of RNAi as a functional genetic tool in mammalian cells.

ACKNOWLEDGMENTS We would also like to acknowledge the NYU-RNAi Core facility (funded by the Kimmel Center for Stem Cell Biology and the NYU Cancer Institute) and a special thanks to Dr. Chi Yun and Shauna Katz for their valuable advice and technical assistance.

REFERENCES Birmingham, A., Selfors, L. M., Forster, T., Wrobel, D., Kennedy, C. J., Shanks, E., Santoyo-Lopez, J., Dunican, D. J., Long, A., Kelleher, D., Smith, Q., Beijersbergen, R. L., et al. (2009). Statistical methods for analysis of high-throughput RNA interference screens. Nat. Methods 6, 569–575. Brummelkamp, T. R., Bernards, R., and Agami, R. (2002). A system for stable expression of short interfering RNAs in mammalian cells. Science 296, 550–553. Buchholz, F., Kittler, R., Slabicki, M., and Theis, M. (2006). Enzymatically prepared RNAi libraries. Nat. Methods 3, 696–700. Conrad, C., and Gerlich, D. W. (2010). Automated microscopy for high-content RNAi screening. J. Cell Biol. 188, 453–461. Ding, L., Paszkowski-Rogacz, M., Nitzsche, A., Slabicki, M. M., Heninger, A. K., de Vries, I., Kittler, R., Junqueira, M., Shevchenko, A., Schulz, H., Hubner, N., Doss, M. X., et al. (2009). A genome-scale RNAi screen for Oct4 modulators defines a role of the Paf1 complex for embryonic stem cell identity. Cell Stem Cell 4, 403–415.

RNAi in ES Cells

365

Elbashir, S. M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K., and Tuschl, T. (2001). Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494–498. Fire, A., Xu, S., Montgomery, M. K., Kostas, S. A., Driver, S. E., and Mello, C. C. (1998). Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391, 806–811. Hannon, G. J., and Rossi, J. J. (2004). Unlocking the potential of the human genome with RNA interference. Nature 431, 371–378. Hu, G., Kim, J., Xu, Q., Leng, Y., Orkin, S. H., and Elledge, S. J. (2009). A genome-wide RNAi screen identifies a new transcriptional module required for self-renewal. Genes Dev. 23, 837–848. Ivanova, N., Dobrin, R., Lu, R., Kotenko, I., Levorse, J., DeCoste, C., Schafer, X., Lun, Y., and Lemischka, I. R. (2006). Dissecting self-renewal in stem cells with RNA interference. Nature 442, 533–538. McCaffrey, A. P., Meuse, L., Pham, T. T., Conklin, D. S., Hannon, G. J., and Kay, M. A. (2002). RNA interference in adult mice. Nature 418, 38–39. McManus, M. T., Petersen, C. P., Haines, B. B., Chen, J., and Sharp, P. A. (2002). Gene silencing using micro-RNA designed hairpins. RNA 8, 842–850. Meister, G., and Tuschl, T. (2004). Mechanisms of gene silencing by double-stranded RNA. Nature 431, 343–349. Moffat, J., Grueneberg, D. A., Yang, X., Kim, S. Y., Kloepfer, A. M., Hinkle, G., Piqani, B., Eisenhaure, T. M., Luo, B., Grenier, J. K., Carpenter, A. E., Foo, S. Y., et al. (2006). A lentiviral RNAi library for human and mouse genes applied to an arrayed viral high-content screen. Cell 124, 1283–1298. Paddison, P. J., Caudy, A. A., Bernstein, E., Hannon, G. J., and Conklin, D. S. (2002). Short hairpin RNAs (shRNAs) induce sequence-specific silencing in mammalian cells. Genes Dev. 16, 948–958. Paul, C. P., Good, P. D., Winer, I., and Engelke, D. R. (2002). Effective expression of small interfering RNA in human cells. Nat. Biotechnol. 20, 505–508. Schaniel, C., Li, F., Schafer, X. L., Moore, T., Lemischka, I. R., and Paddison, P. J. (2006). Delivery of short hairpin RNAs—triggers of gene silencing—into mouse embryonic stem cells. Nat. Methods 3, 397–400. Schaniel, C., Ang, Y. S., Ratnakumar, K., Cormier, C., James, T., Bernstein, E., Lemischka, I. R., and Paddison, P. J. (2009). Smarcc1/Baf155 couples self-renewal gene repression with changes in chromatin structure in mouse embryonic stem cells. Stem Cells 27, 2979–2991. Sui, G., Soohoo, C., Affar el, B., Gay, F., Shi, Y., and Forrester, W. C. (2002). A DNA vector-based RNAi technology to suppress gene expression in mammalian cells. Proc. Natl. Acad. Sci. USA 99, 5515–5520. Xia, H., Mao, Q., Paulson, H. L., and Davidson, B. L. (2002). siRNA-mediated gene silencing in vitro and in vivo. Nat. Biotechnol. 20, 1006–1010. Yang, D., Buchholz, F., Huang, Z., Goga, A., Chen, C. Y., Brodsky, F. M., and Bishop, J. M. (2002). Short RNA duplexes produced by hydrolysis with Escherichia coli RNase III mediate effective RNA interference in mammalian cells. Proc. Natl. Acad. Sci. USA 99, 9942–9947. Yu, J. Y., DeRuiter, S. L., and Turner, D. L. (2002). RNA interference by expression of short-interfering RNAs and hairpin RNAs in mammalian cells. Proc. Natl. Acad. Sci. USA 99, 6047–6052.

C H A P T E R

N I N E T E E N

Transgenic RNAi Applications in the Mouse Jost Seibler* and Frieder Schwenk*,† Contents 368 369 369 370 370 370 371 372 372 373 373 375 376 376 377 379 380 381 382 382 382

1. Introduction 2. Short Interfering RNAs 2.1. Selection of siRNA 2.2. Off-target effects by siRNAs 2.3. siRNAs in vivo 3. Short Hairpin RNA Expression Vectors 4. Transgenic shRNA Expression In Vivo 4.1. Oversaturation of the endogenous machinery 4.2. Recombinase-assisted integration of shRNA vectors 5. Conditional RNAi 5.1. Tissue-specific approaches 5.2. Inducible RNAi 6. Experimental Protocols 6.1. Design of shRNAs 6.2. Cloning of shRNA vectors 6.3. DNA preparation for RMCE 6.4. Transfection ES cells for RMCE 6.5. Doxycycline treatment of ES cells 6.6. Quantification of gene silencing 6.7. Generation and phenotyping of mice References

Abstract Within the past 10 years, RNA interference has emerged as a powerful experimental tool as it allows rapid gene function analysis. Unique features such as reversibility of gene silencing and simultaneous targeting of several genes characterize the approach. In this chapter, transgenic RNAi techniques in reverse mouse genetics are discussed and protocols are provided. * TaconicArtemis GmbH, Cologne, Germany University of Applied Science Gelsenkirchen, Recklinghausen, Germany

{

Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77019-1

#

2010 Elsevier Inc. All rights reserved.

367

368

Jost Seibler and Frieder Schwenk

1. Introduction The ability to manipulate the genome in embryonic stem (ES) cells and mice was developed in the late 1980s. Since then, gene targeting has been extensively used to study gene function in genetically modified mouse strains. As initially developed, the technique allows the disruption of target genes in the murine germ line by the insertion of a selectable marker. The vast majority of presently
4.000 existing ‘‘knockout’’ (KO) mouse models have been created following this design. Many of these KO strains have given valuable information on human physiology and disease processes. As ‘‘conventional’’ KO mice are usually homozygous for a null allele in the germ line, they provide an appropriate model for inherited diseases, leading to embryonic or early postnatal lethality in about 30% of cases. Beyond this application, germ line KO mice do not necessarily represent the best technical approach to study other aspects of gene function in vivo, in particular in adult mice. In many cases, the phenotype obtained in a conventional KO strain reflects a developmental defect rather than gene function in the mature organ. In addition, other gene products can compensate the activity of a continuously inactivated gene, thereby veiling the KO phenotype. A refined KO strategy, termed conditional gene targeting, has been developed, permitting target gene inactivation to be restricted to a particular organ and/or developmental stage. This is achieved by the expression of the site-specific DNA recombinase Cre in conjunction with the introduction of two recombinase recognition sequences (loxP) into the genomic locus of interest. These sites are usually placed into introns such that recombination results in gene inactivation through the deletion of the loxP-flanked exon(s), whereas the unrecombined allele is fully active. Since the initial demonstration of this technology in 1994 (Gu et al., 1994), an increasing number of conditional KO experiments has been published, most of which employed tissue-specific promoters to control Cre expression. Spatially regulated gene inactivation has proven to be powerful in revealing organ specific gene function. The method of choice for precise gene function analyses in adult mice, however, is the temporal control of gene inactivation as it can prevent impaired embryonic development until the time of induction. Furthermore, it permits the investigation of the effect of gene inactivation after the onset of a chronic or acute disease phenotype, simulating the activity of antagonistic drugs. This aspect is of particular interest for the validation of potential drug target genes in pharmaceutical drug development. Today, several strategies for inducible, recombinase-mediated gene targeting are established, including transcriptional control using tetracyclines (Schonig et al., 2002) and activation at the protein level by synthetic hormones (Seibler et al., 2003).

369

RNAi

Spatially and temporally regulated gene inactivation has facilitated detailed dissection of gene function in a physiological process, as demonstrated by studies concerning the role of insulin receptor signaling in glucose homeostasis and diabetes mellitus (Koch et al., 2008; Plum et al., 2005). An inherent feature of these recombinase-based approaches is a nonreversible gene switch that does not allow modulating gene expression in a given cell. In addition, the derivation of conditional mouse mutants is costly and time consuming due to extensive vector construction, ES cell manipulation, and breeding. These limitations have been tackled by RNA interference (RNAi), which will be delineated in the following sections.

2. Short Interfering RNAs RNAi is an endogenous cellular process by which messenger RNAs (mRNAs) are targeted for degradation by complementary pairing to short RNA molecules (Bernstein et al., 2001). Major components of the RNAi machinery involve intracellular trafficking, RNA processing and mRNA cleavage, and is at least in part identical with the endogenous micro RNA (miRNA) pathway (Nelson et al., 2003). In the late 1990s, RNAi has been recognized as a tool for experimental inhibition of gene expression by introducing double-stranded RNA (dsRNA) molecules into cells (Fire et al., 1998). Since long dsRNAs can provoke an interferon response in mammalian cells, the technology was initially restricted to organisms showing no interferon response, such as Caenorhabditis elegans and Drosophila melanogaster (Bass, 2001). The finding that short (<30 bp) interfering RNAs (siRNAs) circumvent the interferon response extended the application to mammalian cells (Elbashir et al., 2001). Since then, RNAi has become part of the standard repertoire for cell-based screens in basic research and drug development (Boutros and Ahringer, 2008).

2.1. Selection of siRNA Systematic comparisons revealed that siRNA oligonucleotides are at least as efficient as RNase H-dependent antisense strategies (Grunweller et al., 2003; Uprichard, 2005; Vickers et al., 2003). Unlike the current antisense approaches, however, the requirements in respect to the sequence and the site of the target mRNA appears to be less restricted for siRNAs (Vickers et al., 2003). The original design rules were published by the group of Thomas Tuschl (Elbashir et al., 2001), whereas the discovery of the asymmetric binding to the RNA-induced silencing complex (RISC) greatly improved the success rate of predicting effective siRNAs (Khvorova et al., 2003; Schwarz et al., 2003). Nevertheless the cross-silencing capability of

370

Jost Seibler and Frieder Schwenk

dsRNAs for other genes must be considered. Several algorithms are available to support the design of effective siRNAs (Henschel et al., 2004; Huesken et al., 2005; Reynolds et al., 2004; Tilesi et al., 2009).

2.2. Off-target effects by siRNAs Recently, ‘‘off-target effects’’ by siRNAs that cross-react with targets of limited sequence similarity have been reported (Jackson et al., 2003). It has been assumed that siRNAs bearing as few as seven complementary bases can exert gene silencing via the miRNA pathway (Jackson and Linsley, 2010; Lin et al., 2005). Birmingham et al. (2006) have identified stretches of 6–7 bp from position 2 onward of the siRNA matching with 30 UTR regions as being responsible for this mechanism. In addition, unspecific effects due to immune stimulation or saturation of the RNAi machinery are described (Grimm et al., 2006; Kleinman et al., 2008). Although this phenomenon has been observed in a minority of siRNA-based experiments so far, it clearly requires further attention. Off-target effects can be readily recognized by designing appropriate control experiments, such as transfection of different siRNAs directed against the target mRNA. Strategies to avoid potential off-target effects include the elimination of proinflammatory sequences as well as chemical modification and new delivery systems to guide siRNA distribution ( Jackson and Linsley, 2010).

2.3. siRNAs in vivo RNAi mediated inhibition of gene expression in vivo has been demonstrated by injecting purified siRNA into the tail vain of mice, resulting in transient gene silencing selected organs (Lewis et al., 2002). Numerous studies have documented siRNA efficacy in mammalian model organisms, and the first clinical trials of therapeutic applications are underway (Haussecker, 2008). Despite these early successes, the widespread use of RNAi therapeutics requires further development of clinically suitable, safe, and effective compounds (Grimm and Kay, 2007; Kleinman et al., 2008). New approaches to improve delivery and pharmacology of siRNAs may meet these issues (Rajendran et al., 2010; Whitehead et al., 2009). Nevertheless, direct administration of siRNA is still inappropriate for most ‘‘reverse genetics’’ gene function studies due to the short half-life and biased distribution in vivo.

3. Short Hairpin RNA Expression Vectors The basis of sustained gene silencing was provided by the finding that transgenic expression of short hairpin RNA (shRNA) results in the intracellular production of siRNA molecules by recruiting the endogenous

371

RNAi

processing machinery (Brummelkamp et al., 2002; Tuschl, 2002). As a specific algorithm for the design of shRNAs is still not established, testing of three to five sequences is required to assure effective gene silencing (Li et al., 2007). For the expression of shRNAs in transgenic cell lines, the RNA polymerase III (pol III)-dependent promoters U6 and H1 have been used (Brummelkamp et al., 2002; Devroe and Silver, 2002; Hemann et al., 2003; Jacque et al., 2002; Paddison et al., 2002; Paul et al., 2002; Sui et al., 2002; Xia et al., 2002; Yu et al., 2002). These promoters generate small transcripts lacking nonfunctional bases at the 50 and 30 ends. It has been shown that the presence of nonhomologous bases at the ends of the shRNA, as produced by the pol III transcription complex, lowers the efficiency of RNAi-mediated gene silencing (Xia et al., 2002). To achieve shRNA expression from RNA polymerase III (pol II)-dependent promoters anyhow, shRNAs have been embedded into miRNA transcripts (shRNA– mir), leading to a transcript that is processed by the endogenous miRNA machinery (Cullen, 2005; Zeng et al., 2002). This approach opens the large collection of well-established pol II promoters for RNAi applications, increasing the flexibility in transgenic shRNA expression significantly. In addition, polycistronic miRNA constructs allow the independent processing of several different shRNAs from a single transcript (Chung et al., 2006). In our experience however, shRNA-mediated gene silencing using miRNA-based vectors appeared to be less efficient as compared to shRNA constructs under the control of pol III promoters ( J. Seibler and F. Schwenk, unpublished observations), possibly due to a lower transcription level (McBride et al., 2008).

4. Transgenic shRNA Expression In Vivo The activity of shRNA in mice have been demonstrated through injection of shRNA expression vectors into the tail vain (Lewis et al., 2002; McCaffrey et al., 2002). These reports established for the first time that the shRNA processing machinery is active in different cell types of a mammal. Initial attempts to achieve sustained shRNA expression in mouse or rat applied random transgenesis by ES cell random transfection, pronucleus injection, or lentiviral transduction, resulting in efficient (up to 90%) gene knockdown in multiple tissues (Dann et al., 2006; Hasuwa et al., 2002; Kunath et al., 2003; Peng et al., 2006; Rubinson et al., 2003; Tiscornia et al., 2003). Since these experiments employed random transgenesis, each individual mouse line showed a unique, irreproducible shRNA expression pattern due to the inherent problem of position-dependent promoter activity. Thus, a thorough characterization of the resulting mouse lines was required to determine the level of RNAi in all relevant organs, which

372

Jost Seibler and Frieder Schwenk

is a time-consuming and laborious undertaking. Prescreening of ES cell clones for appropriate shRNA expression prior to blastocyst injection did not solve this problem as position effects are largely cell type specific (Carmell et al., 2003). Similarly, expression of lentiviral vectors is strongly influenced by the site of provirus integration (van den Brandt et al., 2004). Therefore, targeted transgenesis into a defined genomic locus represents the method of choice to facilitate appropriate expression in a given tissue ( Jasin et al., 1996). Integration of shRNA constructs into both the ubiquitously active rosa26 and hprt loci have been tested, resulting in body wide gene silencing with high efficiency, ranging between 80% and 95% knockdown in most organs (Christoph et al., 2008; Oberdoerffer et al., 2005; Seibler et al., 2005, 2007). The reason for the lower degree of RNAi observed in lymphocytes and testis remains unclear. It is unlikely that a limited activity of RNA pol III-dependent promoters accounts for these differences, as Oberdoerffer et al. (2005) did not find any correlation of shRNA expression level and RNAi efficiency. We rather suspect that the degree of RNAi is affected by other factors, such as the availability of RISC and/or dicer proteins. Anyhow, the pattern of shRNA expression in rosa26 knock-in mice is predictable and not affected by variable chromosomal position effects, as confirmed in
200 different RNAi projects performed in our laboratories so far ( J. Seibler and F. Schwenk, unpublished observations).

4.1. Oversaturation of the endogenous machinery A study of the long-term effects of shRNA expression in liver of adult mice sounds a note of caution, as several shRNAs turned out to be toxic when overexpressed in mice (Grimm et al., 2006; Xia et al., 2006). The toxicity seems to result from competition between shRNAs and endogenous miRNAs for binding to exportin-5, a factor involved in transporting RNAs out of the nucleus. Investigation of other shRNA-transgenic animal models did not reveal a similar effect (Herold et al., 2008; Kotnik et al., 2009; Sasaguri et al., 2009). Anyhow, the potential problem of miRNA pathway saturation as well as off-target effects due to shRNA overexpression by multicopy transgene insertions could also be circumvented by the targeted integration of a single construct, as suggested by Grimm et al. (2006). In fact, we never observed unspecific effects in any shRNA-transgenic mouse model generated by targeted integration of single-copy constructs into the rosa26 locus so far ( J. Seibler and F. Schwenk, unpublished observations).

4.2. Recombinase-assisted integration of shRNA vectors To further speed up the generation of shRNA-transgenic animals, recombinase-mediated cassette exchange (RMCE) has been applied for the integration of transgenes into a predefined locus using site-specific DNA recombination systems such as Flp/FRT or Cre/loxP (Baer and Bode,

373

RNAi

2001; Bode et al., 2000). The use of two incompatible recognition sequences, for example, FRT in combination with F3 (Schlake and Bode, 1994) as well as inverted recognition target sites (Feng et al., 1999) allows the insertion of DNA segments into a predefined chromosomal locus carrying target sequences in the same configuration. In contrast to approaches using a single recombination site, the ‘‘RMCE acceptor’’ locus carrying incompatible recognition sites is stable even under the permanent influence of the recombinase unless it is exposed to an exchange plasmid (Seibler and Bode, 1997). An improved RMCE-strategy has been devised that facilitates the insertion of shRNA transgenes into the rosa26 locus with high efficiency and accuracy (Seibler et al., 2007; SteuberBuchberger et al., 2008; Fig. 19.1). Given the simple cloning strategy for the insertion of shRNA transgenes into the RMCE exchange vector and the limited screening efforts, the RMCE approach largely facilitates the generation of shRNA-transgenic mice. Moreover, the RMCE-strategy should be applicable to other genetic model organisms such as rat, in which homologous recombination is not yet established for routine application.

5. Conditional RNAi Techniques for conditional gene silencing have been established to achieve tissue-specific and/or body-wide inducible shRNA expression. Experience over the past 10 years of technology development in recombinase-mediated gene targeting significantly guided this process. Spatially and temporally regulated RNAi not only allows a more detailed gene function analysis but also circumvents potential confounding developmental effects or compensatory mechanisms altering the phenotype.

5.1. Tissue-specific approaches Tissue-specific control of shRNA expression in mice has been demonstrated by inserting a loxP-flanked ‘‘stop cassette’’ close to the TATA-box of the pol III-promoter, blocking transcription. Derepression through Cremediated excision of the insert allowed gene silencing in selected tissues (Hitz et al., 2007; Oberdoerffer et al., 2005; Ventura et al., 2004). A second strategy employed a loxP-flanked ‘‘promoter inactivating element’’ (PIE) that completely blocks transcription when placed between the proximal (PSE) and distal sequence elements (DSE) of the U6 promoter. The DSE contains one or more binding sites of transcriptional activators such as Oct-1 and Staf; whereas the PSE recruits a stable protein complex containing subunits SNAPc and PTF (Myslinski et al., 2001). Since a single loxP site

374

Jost Seibler and Frieder Schwenk

RRS

RRS P

RMCE acceptor locus

RRS

RRS Selection marker

PolyA

shRNA Exchange vector

RRS

RRS P

Selection marker

shRNA

Exchanged RMCE locus

Figure 19.1 Introduction of shRNA expression cassettes using recombinase-mediated cassette exchange (RMCE). RMCE can be achieved by using different systems of sitespecific DNA recombination. To assure directed integration of the transgene, incompatible recombinase recognition sites (RRS) such as attB/attP of recombinase C31 (Belteki et al., 2003) or F3/FRT of Flp (Seibler et al., 1998) have been employed. In addition, placing RRS in inverted orientation to each other prevents deletion of the DNA segment after successful integration (Feng et al., 1999). The 50 truncated selection marker gene in the exchange vector lacks a functional promoter, facilitating expression through the promoter (P) in the acceptor locus only upon successful RMCE. The polyadenylation signal (poly A) in the backbone of the exchange vector prevents expression of the truncated selection marker gene after random integration downstream of an endogenous promoter (Seibler et al., 2005). By transfection of the exchange vector along with a recombinase expression construct, the exchanged RMCE locus will be generated. X, sites of cross over during recombination; shRNA, shRNA coding region under the control of an RNA polymerase III-dependent promoter.

in the same position does not interfere with transcription, the system allows tight control of RNAi through Cre-mediated recombination of PIE (Coumoul et al., 2005; Yu and McMahon, 2006). In a third method, the loxP-site flanked stop cassette was embedded in the loop of the shRNA (Hitz et al., 2007; Steuber-Buchberger et al., 2008). Although the remaining loxP site becomes part of the shRNA transcript upon Cre-mediated recombination, a loss of RNAi potency was not observed (Hitz et al., 2007). Moreover, this strategy has been applied for simultaneous regulation of two shRNA vectors combined into a single construct (Steuber-Buchberger et al., 2008). Finally, Cre-mediated inversion of an miRNA-based shRNA expression vector using two mutant loxP sites in opposite orientation has been applied for cell type-specific gene silencing (Stern et al., 2008). Since none of these approaches are tested on broad basis for multiple different mRNA targets so far, time will tell which of them are most suitable in terms of robustness and universal applicability (Fig. 19.2).

375

RNAi

A H1/U6

Stop shRNA loxP loxP

Cre

H1/U6

shRNA loxP

B H1/U6

s ––lo

op - as

Stop loxP

loxP

Cre

H1/U6

s – lo op -- as loxP s loxP as Loop

C TSP-Pol II

shRNA-mir shRNA-mir

Figure 19.2 Approaches for tissue-specific RNAi. (A) An RNA polymerase III-dependent promoter is inactivated through the insertion of a loxP-flanked DNA segment, blocking transcription (stop). Upon Cre-mediated deletion of stop, the promoter is converted to the activate state. The remaining loxP site should not disturb the promoter function. This is achieved either placing the stop sequence either between the PSE and DSE of the promoter (Myslinski et al., 2001) or downstream of the TATA-box (Oberdoerffer et al., 2005; Ventura et al., 2004). (B) A loxP-flanked stop sequence is placed into the loop coding region, preventing transcription of a proper shRNA. Cremediated excision of the stop signal activates the shRNA expression vector, whereas the remaining loxP site becomes part of the loop structure (Hitz et al., 2007; SteuberBuchberger et al., 2008). (C) The shRNA is embedded in an miRNA vector under the control of a tissue specific RNA polymerase II-dependent promoter (TSP-Pol II). Transcription leads to a product that is processed by the endogenous miRNA machinery (Cullen, 2005; Zeng et al., 2002). This configuration has not yet been tested in transgenic animals.

5.2. Inducible RNAi Analogous to the tissue-specific RNAi approaches described above, the Cre–loxP system has also been applied for temporal control of shRNA expression. By using a tamoxifen controllable Cre–ER fusion protein, inducible activation of an shRNA expression vector was achieved by deleting a loxPflanked stuffer sequence inserted between the PSE and DSE of the U6 promoter (Yu and McMahon, 2006). The major drawback of this approach, however, is its lack of reversibility reducing its usefulness for many applications. Reversible control of shRNA expression has been achieved by using engineered promoters containing a tetracycline operator (tetO) sequences (Ohkawa and Taira, 2000). The operator sequence itself has no effect on shRNA expression. In the presence of the tetracycline repressor (tetR), transcription is blocked through binding of the

376

Jost Seibler and Frieder Schwenk

repressor to tetO. Derepression is achieved by adding the inductor doxycycline, causing the release of the TetR protein from the tetO site and allowing transcription (Yao and Eriksson, 1999). Several attempts have been made to apply this strategy for the temporary control of antisense or shRNA expression in cultured cell lines (Czauderna et al., 2003; Matsukura et al., 2003; Ohkawa and Taira, 2000; van de Wetering et al., 2003). In these reports, the degree of doxycycline-inducible mRNA degradation was variable. In addition, background RNAi activity in the noninduced state was observed (van de Wetering et al., 2003), indicating a limiting level of tetR expression in these cell lines. We recently found that a codon-optimized repressor gene, such as the tetracycline repressor gene, suppresses the activity of shRNA genes under the control of an engineered H1 promoter containing the corresponding operator sequences in transgenic animals (Seibler et al., 2007) (Fig. 19.3A). This system has also been adapted for the rat using lentiviral infection and pronucleus injection strategies. Two recent publications illustrate the application of inducible gene knockdown in rat for generating disease models (Herold et al., 2008; Kotnik et al., 2009). Another approach the tetracycline (tet)-responsive system, initially developed for transgenic overexpression of proteins (Gossen and Bujard, 1992; Gossen et al., 1995), has been adapted for inducible expression of an miRNA-based shRNA vector (Dickins et al., 2007) (Fig. 19.3B).

6. Experimental Protocols In this section, we provide a collection of protocols for the generation of shRNA-transgenic mice using RMCE, as performed in our laboratories (Seibler et al., 2007). We usually generate two mouse strains in parallel, expressing different shRNA sequences directed against a given mRNA target to exclude unspecific phenotypes.

6.1. Design of shRNAs A given siRNA sequence can be converted into an shRNA coding region by inserting a loop sequence between the sense (bold) and the antisense strand (bold, underlined). The loop sequence of the hairpin structure is shown in italics and the overhangs for restriction endonucleases BbsI (TCCC) and MluI (CGCG) are underlined. siRNA: GGGAGAACGTACTTCGGAA (sense strand) shRNA: TCCC GGGAGAACGTACTTCGGAATTCAAGAGATTC CGAAGTACGTTCTCCC TTTTTACGCG

377

RNAi

A

tet R

+Dox

tet R Promoter itetR

B

H1-tetO

Pol III H1-tetO

shRNA

shRNA

VP16 rtet R VP16 +Dox

Promoter

rtTA

tetO7-PCMVmin

shRNA-mir

rtet R tetO7-PCMVmin

shRNA-mir

shRNA-mir

Figure 19.3 Schematic representation of inducible RNAi approaches. (A) Binding of the tetracycline resistance operon repressor (tetR) to the operator sequence (tetO) inhibits the expression of the shRNA through sterical hindrance. Administration of doxycycline (Dox) leads to derepression of the promoter through dissociation of tetR, allowing shRNA expression by the RNA polymerase III-dependent promoter (Seibler et al., 2007). (B) The rtTA consists of a reverse tetracycline resistance operon repressor fused to a VP16 domain (rtTA) for transactivation (Gossen and Bujard, 1992; Gossen et al., 1995). In the absence of Dox, a minimal CMV promoter carrying seven tetO sequences (tetO7PCMVmin) is silent. In the presence of Dox, the rtTA binds to the tet-operator sequence and thereby activates transcription of the shRNA–mir sequence (Dickins et al., 2007).

Oligonucleotides for simple vector construction: Sense strand TCCCGGGAGAACGTACTTCGGAATTCAAGAG ATTCCGAAGTACGTTCTCCCTT TTTA Antisense strand CGCGTAAAAAGGGAGAACGTACTTCGGAA TCTCTTGAATTCCGAAGTACGTTCTCCC

6.2. Cloning of shRNA vectors The exchange vector (pINV-7; Fig. 19.4) contains the following elements in 50 to 30 order: synthetic polyadenylation signal (poly A), F3-site, neomycin-resistance gene lacking the start ATG, poly A derived from the phosphoglycerate kinase gene (PGK), poly A derived from the human growth hormone gene (hGH), H1-tet-promoter (Seibler et al., 2007), multiple cloning site (MCS), CAGGS promoter (Okabe et al., 1997), the codonoptimized tet-repressor gene (Seibler et al., 2007), synthetic poly A (Seibler et al., 2005), and FRT-site. The shRNA oligonucleotides (see Section 6.1) can be directly cloned into the pINV-7 vector by the following protocols.

378

Jost Seibler and Frieder Schwenk

pA f1 (+) ori F3 5⬘ D neo pA Amp

pINV-7 8721 bp

pA hGH

H1-tetO promoter

FRT

Bb sI (2980)

pA htetR

Ml uI (3009)

CAGGS promoter

Synthetic intron

Figure 19.4 The Dox-inducible RMCE exchange vector pINV-7. The vector is designed for insertion of shRNA oligonucleotides, which can be directly ligated into the multiple cloning site (MCS) in a single step (Seibler et al., 2007). Poly A, synthetic polyadenylation signal; FRT/F3, recognition sequences for Flp recombinase; D neo, neomycin-resistance gene lacking promoter and start ATG; poly A, polyadenylation signal, poly A hGH, 50 fragment of the human growth hormone gene, including polyadenylation signal; H1-tetO, H1 promoter containing operator sequences (tetO) of the E. coli tetracycline resistance operon (Seibler et al., 2007); CAGGS, chicken actin enhancer/hCMV promoter (Okabe et al., 1997); htetR, codon-optimized tet-repressor gene (Seibler et al., 2007).

6.2.1. T4 Polynucleotide Kinase treatment and annealing of oligonucleotides Reagents     

Sense and antisense oligonucleotides (see Section 6.1) T4 kinase (Invitrogen) 10 mM ATP 5 M NaCl TE buffer Procedure

T4 Polynucleotide Kinase treatment: 1. Transfer 1 g oligonucleotides (1 g/l) into a clean tube. 2. Add 5 l 5 forward buffer,

RNAi

379

2.5 l 10 mM ATP, 1 l T4 kinase, and 15.5 l H2O. 3. Incubate for 30 min at 37  C. 4. Incubate for 10 min at 65  C. Annealing: 5. Add 14 l kinase treated sense oligonucleotides, 14 l kinase treated antisense oligonucleotides, and 4 l 5 M NaCl. 6. Incubate for 15 min at 65  C. 7. Cool down on room temperature (RT) slowly (outside the heatblock). 8. Add 68 l TE buffer. 9. Dilute annealed oligonucleotides 1:10 with ddH2O. 10. Use 4 l from dilution for ligation. 6.2.2. Ligation and transformation Reagents   

DNA fragments (vector, insert) T4 DNA ligase (5 U/l; Invitrogen, Cat. No. 15224-041) 5 ligase reaction buffer (from 20 l aliquots) Procedure

1. Add 4 l 5 ligase reaction buffer, 4 l annealed oligonucleotides, 50 ng BbsI/MluI restricted vector DNA, and 1–2 l ligase (5 U/l) into a clean tube. 2. Fill up to 20 l total volume using autoclaved ddH2O. 3. Incubate for 15 min at RT. 4. Use 5 l of ligation mixture for heat shock transformation of competent Escherichia coli cells. 5. Plate heat shock transformed cells on LB-agar with 100 g/ml ampicillin. Due to the high rate of correctly ligated fragments, we usually do not perform restriction digests to screen for correct clones prior to sequencing of the insert.

6.3. DNA preparation for RMCE Reagents 

RMCE exchange vector pINV7

380     

Jost Seibler and Frieder Schwenk

PureLinkTM HiPure Plasmid Filter Midiprep Kit (Cat. No. K2100-14) 3 M sodium acetate (NaOAc), pH 5.3 100% ethanol 70% ethanol Embryo tested H2O (sterile) Procedure

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

Transfer 20–50 g RMCE vector into a clean tube. Fill up to 200 l total volume using autoclaved ddH2O. Add 1/10 volume 3 M NaOAc, pH 5.3 (20 l). Add 2 volumes 100% ethanol (440 l). Shake the tube. Incubate at least 30 min at 20  C. Centrifuge for 15 min at 4  C (maximal speed). Remove supernatant. Add 500 l 70% ethanol. Centrifuge for 10 min at RT (maximal speed). Remove supernatant. Add 500 l 70% ethanol. Centrifuge for 10 min at RT (maximal speed). Remove supernatant with a sterile pipette tip under sterile conditions (hood). Air-dry the pellet (usually a few minutes). Dissolve pellet in embryo tested H2O (20 g in /30 l). Determine the concentration (photometric). Adjust DNA solution to a final concentration of 0.5 g/l using H2O (embryo tested).

6.4. Transfection ES cells for RMCE ES cell culture is carried out as described in Hogan et al. (1994). ES cells carrying the RMCE docking sites (Fig. 19.1) are transfected with the exchange vector preparation (see Section 6.3). Reagents FugeneÒ6 transfection reagent (Roche), 6 l/well 100 l/well liquid OptiMEMÒ Reduced-Serum Medium (Cat. No. 11058021)  ES cell culture medium consisting of ○ 472 ml DMEM (high glucose) ○ 90 ml FCS (US certified and high quality) ○ 12 ml 200 mM glutamin ○ 6 ml 100 nonessential amino acids ○ 6 ml 100 mM sodium pyruvate  

RNAi

381

12 ml 1 M Hepes buffer 1.2 ml 50 mM b-mercaptoethanol ○ 90 l 107 U/ml LIF ART4.12 ES cells ([C57BL/6  129S6/SvEvTac] F1) Mitomycin c (MMC)-treated feeder cells on 6-well plates Exchange vector preparation (0.5 g/l; see Section 6.3) Flpe expression vector, CAGGs-Flp (Schaft et al., 2001) Genecitin (Invitrogen; Cat. No. 10131-019) Haemocytometer (Neubauer) ○ ○

     

Procedure 1. One day before transfection, seed 1.5  105 ES cells/well into a 6-well plate containing a MMC treated feeder layer. 2. Replenish cell culture medium 3–4 h before transfection. 3. Preparation 1 (for each well): transfer 1 g circular exchange vector, 1 g Flpe expression vector and 10 l of OptiMEM into a sterile tube. 4. Preparation 2 (for each well): transfer 100 l OptiMEM and, subsequently, 6 l FugeneÒ6 transfection reagent without touching the plastic surface of the tube; vortex briefly. 5. Incubate for 5 min at RT. 6. For each well, transfer 100 l of preparation 2 into preparation 1. 7. Incubate transfection mixture for 30–60 min at RT. 8. Add 100 l of the transfection mixture dropwise into 1-well of the 6-well plate; swirl the plate to ensure equal distribution. 9. Replenish cell culture medium 24 h after transfection. 10. For selection of transgenic ES cell clones, add Genecitin to the cell culture medium 1–2 days after transfection. 11. Single clones can be picked approximately 7 days after transfection.

6.5. Doxycycline treatment of ES cells Reagents    

Prewarmed ES-medium (37  C) 15 ml conical tubes 10 cm cell culture dishes or 6-well plates Filter sterilized doxycycline (Sigma D-9891; 10 mg/ml stock solution in ddH2O; 1 ml aliquots stored at  20  C) Procedure

1. Thaw ES cell aliquot and plate on gelatinized 6-well plate with ES cell medium.

382

Jost Seibler and Frieder Schwenk

2. Replate cells on gelatinized plate to remove residual feeder cells and start with doxycycline (Dox) treatment 1 day later: a. Thaw stock solution of 10 mg/ml Dox and prepare 15 ml aliquots; store at 20  C. b. Add 10 ml of Dox stock solution to 990 ml of prewarmed cell culture medium containing Genecitin (final concentration: 1 mg/ml Dox). c. Incubate cells with freshly repaired medium containing Dox and Genecitin for 3 days; change medium daily and split cells, if required. 3. Harvest cells, wash once with PBS and freeze dry cell pellet at 80  C or perform lyses using appropriate buffer for RNA or protein extraction.

6.6. Quantification of gene silencing Levels of mRNA can be measured using the 7900 HT Fast Real-Time PCR System (Applied Biosystems, 96-well format) according to the manufacturer’s instructions. Preconfigured-20 Primer-Probe Gene Expression Assay (Applied Biosystems) is used in a total reaction volume of 20 l. Two microliters of cDNA solution are used for each well. Each sample is measured in triplicates. Results are normalized by using the GAPDH gene expression assay (Applied Biosystems).

6.7. Generation and phenotyping of mice For generation and phenotypic analysis of mice, we refer to Hogan et al. (1994). Additional protocols for phenotyping are provided by the contribution of Shin et al. in this volume (Chapter 21).

REFERENCES Baer, A., and Bode, J. (2001). Coping with kinetic and thermodynamic barriers: RMCE, an efficient strategy for the targeted integration of transgenes. Curr. Opin. Biotechnol. 12, 473–480. Bass, B. L. (2001). RNA interference. The short answer. Nature 411, 428–429. Belteki, G., et al. (2003). Site-specific cassette exchange and germline transmission with mouse ES cells expressing phiC31 integrase. Nat. Biotechnol. 21, 321–324. Bernstein, E., et al. (2001). Role for a bidentate ribonuclease in the initiation step of RNA interference. Nature 409, 363–366. Birmingham, A., et al. (2006). 30 UTR seed matches, but not overall identity, are associated with RNAi off-targets. Nat. Methods 3, 199–204. Bode, J., et al. (2000). The transgeneticist’s toolbox: Novel methods for the targeted modification of eukaryotic genomes. Biol. Chem. 381, 801–813. Boutros, M., and Ahringer, J. (2008). The art and design of genetic screens: RNA interference. Nat. Rev. Genet. 9, 554–566. Brummelkamp, T. R., et al. (2002). A system for stable expression of short interfering RNAs in mammalian cells. Science 296, 550–553.

RNAi

383

Carmell, M. A., et al. (2003). Germline transmission of RNAi in mice. Nat. Struct. Biol. 10, 91–92. Christoph, T., et al. (2008). Investigation of TRPV1 loss-of-function phenotypes in transgenic shRNA expressing and knockout mice. Mol. Cell Neurosci. 37, 579–589. Chung, K. H., et al. (2006). Polycistronic RNA polymerase II expression vectors for RNA interference based on BIC/miR-155. Nucleic Acids Res. 34, e53. Coumoul, X., et al. (2005). Conditional knockdown of Fgfr2 in mice using Cre-LoxP induced RNA interference. Nucleic Acids Res. 33, e102. Cullen, B. R. (2005). RNAi the natural way. Nat. Genet. 37, 1163–1165. Czauderna, F., et al. (2003). Inducible shRNA expression for application in a prostate cancer mouse model. Nucleic Acids Res. 31, e127. Dann, C. T., et al. (2006). Heritable and stable gene knockdown in rats. Proc. Natl. Acad. Sci. USA 103, 11246–11251. Devroe, E., and Silver, P. A. (2002). Retrovirus-delivered siRNA. BMC Biotechnol. 2, 15. Dickins, R. A., et al. (2007). Tissue-specific and reversible RNA interference in transgenic mice. Nat. Genet. 39, 914–921. Elbashir, S. M., et al. (2001). Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494–498. Feng, Y. Q., et al. (1999). Site-specific chromosomal integration in mammalian cells: Highly efficient CRE recombinase-mediated cassette exchange. J. Mol. Biol. 292, 779–785. Fire, A., et al. (1998). Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391, 806–811. Gossen, M., and Bujard, H. (1992). Tight control of gene expression in mammalian cells by tetracycline-responsive promoters. Proc. Natl. Acad. Sci. USA 89, 5547–5551. Gossen, M., et al. (1995). Transcriptional activation by tetracyclines in mammalian cells. Science 268, 1766–1769. Grimm, D., and Kay, M. A. (2007). Therapeutic application of RNAi: Is mRNA targeting finally ready for prime time? J. Clin. Invest. 117, 3633–3641. Grimm, D., et al. (2006). Fatality in mice due to oversaturation of cellular microRNA/short hairpin RNA pathways. Nature 441, 537–541. Grunweller, A., et al. (2003). Comparison of different antisense strategies in mammalian cells using locked nucleic acids, 20 -O-methyl RNA, phosphorothioates and small interfering RNA. Nucleic Acids Res. 31, 3185–3193. Gu, H., et al. (1994). Deletion of a DNA polymerase beta gene segment in T cells using cell type-specific gene targeting. Science 265, 103–106. Hasuwa, H., et al. (2002). Small interfering RNA and gene silencing in transgenic mice and rats. FEBS Lett. 532, 227–230. Haussecker, D. (2008). The business of RNAi therapeutics. Hum. Gene Ther. 19, 451–462. Hemann, M. T., et al. (2003). An epi-allelic series of p53 hypomorphs created by stable RNAi produces distinct tumor phenotypes in vivo. Nat. Genet. 33, 396–400. Henschel, A., et al. (2004). DEQOR: A web-based tool for the design and quality control of siRNAs. Nucleic Acids Res. 32, W113–W120. Herold, M. J., et al. (2008). Inducible and reversible gene silencing by stable integration of an shRNA-encoding lentivirus in transgenic rats. Proc. Natl. Acad. Sci. USA 105, 18507–18512. Hitz, C., et al. (2007). Conditional brain-specific knockdown of MAPK using Cre/loxP regulated RNA interference. Nucleic Acids Res. 35, e90. Hogan, C., et al. (1994). Manipulating the Mouse Embryo: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Huesken, D., et al. (2005). Design of a genome-wide siRNA library using an artificial neural network. Nat. Biotechnol. 23, 995–1001. Jackson, A. L., and Linsley, P. S. (2010). Recognizing and avoiding siRNA off-target effects for target identification and therapeutic application. Nat. Rev. Drug Discov. 9, 57–67.

384

Jost Seibler and Frieder Schwenk

Jackson, A. L., et al. (2003). Expression profiling reveals off-target gene regulation by RNAi. Nat. Biotechnol. 21, 635–637. Jacque, J. M., et al. (2002). Modulation of HIV-1 replication by RNA interference. Nature 418, 435–438. Jasin, M., et al. (1996). Targeted transgenesis. Proc. Natl. Acad. Sci. USA 93, 8804–8808. Khvorova, A., et al. (2003). Functional siRNAs and miRNAs exhibit strand bias. Cell 115, 209–216. Kleinman, M. E., et al. (2008). Sequence- and target-independent angiogenesis suppression by siRNA via TLR3. Nature 452, 591–597. Koch, L., et al. (2008). Central insulin action regulates peripheral glucose and fat metabolism in mice. J. Clin. Invest. 118, 2132–2147. Kotnik, K., et al. (2009). Inducible transgenic rat model for diabetes mellitus based on shRNA-mediated gene knockdown. PLoS One 4, e5124. Kunath, T., et al. (2003). Transgenic RNA interference in ES cell-derived embryos recapitulates a genetic null phenotype. Nat. Biotechnol. 21, 559–561. Lewis, D. L., et al. (2002). Efficient delivery of siRNA for inhibition of gene expression in postnatal mice. Nat. Genet. 32, 107–108. Li, L., et al. (2007). Defining the optimal parameters for hairpin-based knockdown constructs. RNA 13, 1765–1774. Lin, X., et al. (2005). siRNA-mediated off-target gene silencing triggered by a 7 nt complementation. Nucleic Acids Res. 33, 4527–4535. Matsukura, S., et al. (2003). Establishment of conditional vectors for hairpin siRNA knockdowns. Nucleic Acids Res. 31, e77. McBride, J. L., et al. (2008). Artificial miRNAs mitigate shRNA-mediated toxicity in the brain: Implications for the therapeutic development of RNAi. Proc. Natl. Acad. Sci. USA 105, 5868–5873. McCaffrey, A. P., et al. (2002). RNA interference in adult mice. Nature 418, 38–39. Myslinski, E., et al. (2001). An unusually compact external promoter for RNA polymerase III transcription of the human H1RNA gene. Nucleic Acids Res. 29, 2502–2509. Nelson, P., et al. (2003). The microRNA world: Small is mighty. Trends Biochem. Sci. 28, 534–540. Oberdoerffer, P., et al. (2005). Efficiency of RNA interference in the mouse hematopoietic system varies between cell types and developmental stages. Mol. Cell. Biol. 25, 3896–3905. Ohkawa, J., and Taira, K. (2000). Control of the functional activity of an antisense RNA by a tetracycline-responsive derivative of the human U6 snRNA promoter. Hum. Gene Ther. 11, 577–585. Okabe, M., et al. (1997). ‘‘Green mice’’ as a source of ubiquitous green cells. FEBS Lett. 407, 313–319. Paddison, P. J., et al. (2002). Short hairpin RNAs (shRNAs) induce sequence-specific silencing in mammalian cells. Genes Dev. 16, 948–958. Paul, C. P., et al. (2002). Effective expression of small interfering RNA in human cells. Nat. Biotechnol. 20, 505–508. Peng, S., et al. (2006). A transgenic approach for RNA interference-based genetic screening in mice. Proc. Natl. Acad. Sci. USA 103, 2252–2256. Plum, L., et al. (2005). Transgenic and knockout mice in diabetes research: Novel insights into pathophysiology, limitations, and perspectives. Physiology (Bethesda) 20, 152–161. Rajendran, L., et al. (2010). Subcellular targeting strategies for drug design and delivery. Nat. Rev. Drug Discov. 9, 29–42. Reynolds, A., et al. (2004). Rational siRNA design for RNA interference. Nat. Biotechnol. 22, 326–330.

RNAi

385

Rubinson, D. A., et al. (2003). A lentivirus-based system to functionally silence genes in primary mammalian cells, stem cells and transgenic mice by RNA interference. Nat. Genet. 33, 401–406. Sasaguri, H., et al. (2009). Silencing efficiency differs among tissues and endogenous microRNA pathway is preserved in short hairpin RNA transgenic mice. FEBS Lett. 583, 213–218. Schaft, J., et al. (2001). Efficient FLP recombination in mouse ES cells and oocytes. Genesis 31, 6–10. Schlake, T., and Bode, J. (1994). Use of mutated FLP recognition target (FRT) sites for the exchange of expression cassettes at defined chromosomal loci. Biochemistry 33, 12746–12751. Schonig, K., et al. (2002). Stringent doxycycline dependent control of CRE recombinase in vivo. Nucleic Acids Res. 30, e134. Schwarz, D. S., et al. (2003). Asymmetry in the assembly of the RNAi enzyme complex. Cell 115, 199–208. Seibler, J., and Bode, J. (1997). Double-reciprocal crossover mediated by FLP-recombinase: A concept and an assay. Biochemistry 36, 1740–1747. Seibler, J., et al. (1998). DNA cassette exchange in ES cells mediated by Flp recombinase: An efficient strategy for repeated modification of tagged loci by marker-free constructs. Biochemistry 37, 6229–6234. Seibler, J., et al. (2003). Rapid generation of inducible mouse mutants. Nucleic Acids Res. 31, e12. Seibler, J., et al. (2005). Single copy shRNA configuration for ubiquitous gene knockdown in mice. Nucleic Acids Res. 33, e67. Seibler, J., et al. (2007). Reversible gene knockdown in mice using a tight, inducible shRNA expression system. Nucleic Acids Res. 35, e54. Stern, P., et al. (2008). A system for Cre-regulated RNA interference in vivo. Proc. Natl. Acad. Sci. USA 105, 13895–13900. Steuber-Buchberger, P., et al. (2008). Simultaneous Cre-mediated conditional knockdown of two genes in mice. Genesis 46, 144–151. Sui, G., et al. (2002). A DNA vector-based RNAi technology to suppress gene expression in mammalian cells. Proc. Natl. Acad. Sci. USA 99, 5515–5520. Tilesi, F., et al. (2009). Design and validation of siRNAs and shRNAs. Curr. Opin. Mol. Ther. 11, 156–164. Tiscornia, G., et al. (2003). A general method for gene knockdown in mice by using lentiviral vectors expressing small interfering RNA. Proc. Natl. Acad. Sci. USA 100, 1844–1848. Tuschl, T. (2002). Expanding small RNA interference. Nat. Biotechnol. 20, 446–448. Uprichard, S. L. (2005). The therapeutic potential of RNA interference. FEBS Lett. 579, 5996–6007. van den Brandt, J., et al. (2004). Lentivirally generated eGFP-transgenic rats allow efficient cell tracking in vivo. Genesis 39, 94–99. van de Wetering, M., et al. (2003). Specific inhibition of gene expression using a stably integrated, inducible small-interfering-RNA vector. EMBO Rep. 4, 609–615. Ventura, A., et al. (2004). Cre-lox-regulated conditional RNA interference from transgenes. Proc. Natl. Acad. Sci. USA 101, 10380–10385. Vickers, T. A., et al. (2003). Efficient reduction of target RNAs by small interfering RNA and RNase H-dependent antisense agents. A comparative analysis. J. Biol. Chem. 278, 7108–7118. Whitehead, K. A., et al. (2009). Knocking down barriers: Advances in siRNA delivery. Nat. Rev. Drug Discov. 8, 129–138.

386

Jost Seibler and Frieder Schwenk

Xia, H., et al. (2002). siRNA-mediated gene silencing in vitro and in vivo. Nat. Biotechnol. 20, 1006–1010. Xia, X. G., et al. (2006). Pol II-expressed shRNA knocks down Sod2 gene expression and causes phenotypes of the gene knockout in mice. PLoS Genet. 2, e10. Yao, F., and Eriksson, E. (1999). A novel tetracycline-inducible viral replication switch. Hum. Gene Ther. 10, 419–427. Yu, J., and McMahon, A. P. (2006). Reproducible and inducible knockdown of gene expression in mice. Genesis 44, 252–261. Yu, J. Y., et al. (2002). RNA interference by expression of short-interfering RNAs and hairpin RNAs in mammalian cells. Proc. Natl. Acad. Sci. USA 99, 6047–6052. Zeng, Y., et al. (2002). Both natural and designed micro RNAs can inhibit the expression of cognate mRNAs when expressed in human cells. Mol. Cell 9, 1327–1333.

C H A P T E R

T W E N T Y

Gene Knockdown in the Mouse Through RNAi ¨hn† Aljoscha Kleinhammer,* Wolfgang Wurst,†,‡ and Ralf Ku Contents 1. Introduction 2. Principle of the Approach 2.1. shRNA design and cloning 2.2. Genomic integration of vectors and generation of knockdown mice 2.3. Timescale and principle of shRNA validation 2.4. Principle of constitutive and conditional shRNA expression 3. Materials 3.1. Plasmids 3.2. Cell lines 3.3. Primers 3.4. Reagents 4. Methods 4.1. Constitutive knockdown 4.2. Conditional knockdown 5. Concluding Remarks Acknowledgments References

388 389 390 392 394 395 398 398 398 398 398 399 399 405 412 413 413

Abstract RNA interference (RNAi)-mediated gene knockdown has developed into a routine method to assess gene function in cultured mammalian cells in a fast and easy manner. For the use of RNAi in mice, short hairpin (sh) RNAs expressed from transgenic vectors are a fast alternative to conventional knockout approaches. We describe our strategy to elicit body-wide, cell type-specific, or inducible gene silencing in mice by control of shRNA expression through Cre

* Institute for Developmental Genetics, Helmholtz Center Munich, German Research Center for Environmental Health, Neuherberg, Munich, Germany Technical University Munich, Munich, Germany { Max-Planck-Institute of Psychiatry, Molecular Neurogenetics, Munich, Germany {

Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77020-8

#

2010 Elsevier Inc. All rights reserved.

387

388

Aljoscha Kleinhammer et al.

recombinase or doxycycline. For reproducible expression of shRNAs, vectors are placed into the Rosa26 locus of ES cells by recombinase-mediated cassette exchange and transmitted through the germ line of chimeric mice. The site specific insertion of single copy shRNA vectors allows to expedite and reproducible production of knockdown mice and provides a simple approach to assess gene function in vivo.

1. Introduction Silencing of gene expression by RNA interference (RNAi) is used as standard tool for functional genomics in cultured mammalian cells. RNAi is a sequence specific gene silencing process that occurs at the messenger RNA (mRNA) level. In invertebrate cells, long double stranded RNAs (dsRNA), which are processed into short interfering RNAs (siRNA) by the ribonuclease Dicer, induce efficient and specific gene silencing (Elbashir et al., 2001). In this process the siRNA antisense strand serves as a template for the RNA-induced silencing complex (RISC). RISC recognizes and cleaves the complementary mRNA, which is then rapidly degraded. In mammalian cells, long dsRNAs (>30 bp) elicit an interferon response resulting in the global inhibition of protein synthesis and nonspecific mRNA degradation. However, short dsRNAs (<30 bp) trigger the specific knockdown of mRNAs in mammalian cells without interferon activation (Elbashir et al., 2001). Such synthetic siRNAs can be introduced into cultured cells and induce a transient knockdown of target genes. Alternatively, RNA polymerase III (pol III) driven expression vectors enable the permanent production of small dsRNAs in mammalian cells (Brummelkamp et al., 2002; Lee et al., 2002; Paddison et al., 2002). Such vector derived transcripts are designed to contain an antisense region, complementary to a selected mRNA sequence, and the corresponding mRNA sense region. These transcripts fold into a stem-loop structure and form short hairpin RNAs (shRNAs) that are processed by Dicer in a similar way as siRNAs. Expression vectors for shRNAs can be stably integrated into the genome and allow permanent gene silencing in cell lines and organisms. Like in cultured cells, gene silencing can be transiently induced in mice by delivery of siRNA into somatic tissues. Upon intravenous administration, gene silencing occurs mainly in the liver. Other organs like brain or eyes can be targeted by the local infusion of siRNA. Direct siRNA delivery is frequently used in experimental settings that aim for a short-term test of gene function, for example, to study viral replication in the liver (Kuhn et al., 2007). In clinical studies, siRNAs have great potential as drugs to knockdown disease-related genes (Dykxhoorn and Lieberman, 2006).

Gene Knockdown in the Mouse Through RNAi

389

Alternatively, permanent gene silencing can be achieved in mice or rats by the integration of shRNA transgenes into the genome. The efficiency of gene silencing in transgenic mice can achieve levels of 90% knockdown or more. The phenotypes of knockdown and corresponding knockout mice were compared in several instances and found to be identical or very similar. Vector-based transgenic RNAi provides a tool to achieve efficient gene silencing during embryonic development as well as in organs of adult mice. As compared to somatic siRNA delivery, for which the uptake rate is critical, transgenic animals harbor an shRNA expression vector in all cells and provide a well-defined experimental setup. Vector shRNA transgenic mice have been produced by pronucleus injection, viral embryo infection or by targeted transgenesis (knock-in) in embryonic stem (ES) cells (Hasuwa et al., 2002; Hitz et al., 2007; Kunath et al., 2003; Rubinson et al., 2003). In this chapter, we summarize our approach for the streamlined production of shRNA transgenic mice by targeted transgenesis in ES cells (Delic et al., 2008; Hitz et al., 2007; Seibler et al., 2005; Steuber-Buchberger et al., 2008). This approach is based on the insertion of shRNA vectors into the Rosa26 locus of ES cells by recombinase-mediated cassette exchange (RMCE) and the subsequent generation of germ line chimeras. The targeted insertion of shRNAs into the Rosa26 locus ensures reproducible transgene expression and avoids ectopic expression often associated with randomly integrated transgenes. A single shRNA vector copy within the Rosa26 locus is sufficient to elicit constitutive, body-wide gene silencing (Seibler et al., 2005). For the knockdown of gene expression in specific tissues conditional shRNA vectors, that are activated by Cre/loxP recombination, can be used (Hitz et al., 2007). Alternatively, doxycycline (Dox)regulated shRNA vectors allow inducible and reversible expression of shRNAs and enable the temporal control of gene silencing (Seibler et al., 2007). Using the same set of tools a target gene can be silenced in vivo in three different ways by constructing constitutive knockdown mice (Section 4.1); Cre-inducible, conditional knockdown mice (Sections 2.4.1 and 4.2.1); or Dox-inducible, conditional knockdown mice (Sections 2.4.2 and 4.2.2).

2. Principle of the Approach In the following section, we describe step by step how the different types of knockdown mouse are built, starting from the sequence of the target gene. Detailed protocols for the construction of constitutive and inducible shRNA vectors are given in Sections 4.1 and 4.2.

390

Aljoscha Kleinhammer et al.

2.1. shRNA design and cloning First, the mRNA of the target gene is downloaded from the NCBI or Ensembl database (www.ncbi.nlm.nih.gov, www.ensembl.org). As RNAi is working on the mRNA level, it is crucial to employ only the coding sequence (CDS) format. Next, the coding region is screened for sequences that match closest with the criteria of an ideal siRNA sequence. Therefore, the CDS is uploaded to an siRNA prediction program to identify siRNA sequences that fit optimal to the known design rules. Free siRNA prediction programs are available at the web pages of Ambion, Invitrogen, or Qiagen and the University of Tokyo (http://genomics.jp/sidirect/index. php?type¼fc), University of Vienna (http://rna.tbi.univie.ac.at/cgi-bin/ RNAxs), and the Whitehead Institute (http://jura.wi.mit.edu/bioc/siRNAext/home.php). However, only a part of such in silico determined target sites shows efficient gene silencing in vivo. This discrepancy is likely due to the imprecise prediction of mRNA secondary structures that determine the accessibility of siRNA target sequences. Hence, we initially select five predicted siRNA sequences for experimental validation (Fig. 20.3) by combining the best hits of various prediction programs and published siRNA sequences, if available. Even if efficient gene silencing is described for a specific target site, it can be worthwhile to screen the gene sequence again for siRNA sequences as prediction algorithms might have improved in the meantime. Once an siRNA sequence has been selected, a vector has to be chosen to express this sequence as a short dsRNA in mammalian cells. In this configuration, it will be recognized and processed by Dicer and efficiently exert gene silencing. We routinely use shRNA expression vectors to produce shRNAs in mouse cells, as described below. Basis for the expression of shRNAs is a 19–29-bp siRNA sequence (‘‘sense’’) followed by its complementary sequence (‘‘antisense’’) as depicted in Fig. 20.1. Both elements are separated by an 8-bp loop region and linked with an RNA pol III driven promoter from the human U6 or H1 gene and with a termination signal (poly T). When this sequence is transcribed to an 60 nucleotide RNA molecule, sense and antisense sequences are connected by the loop region which allows bending and hybridization of the complementary regions. This reaction results in the dsRNA configuration required for recognition by Dicer. To clone a selected siRNA sequence into an shRNA expression vector, two complementary oligonucleotides, shA and shB, are designed for each siRNA sequence (Fig. 20.1B). The hybridization of the shA and shB oligonucleotides results in an 60 bp DNA fragment in which the shRNA sequence is linked to single-strand overhangs. These overhangs permit ligation of the DNA fragment into corresponding restriction sites of the selected shRNA expression vector.

sh1

sh2

sh3

sh4

sh5

CDS 5¢ -

- 3¢

Sense

shA 5¢ -

shB 5¢ -

Loop

Antisense

GAAGCTTG

Termination Sense

Sense

TTTTTT

Loop

Antisense

CAAGCTTC

AAAAAA

shA 5¢ shB 3¢ -

Termination

Loop

Antisense

Termination TTTTTT AAAAAA

GAAGCTTG CTTCGAAC

- 3¢

1. siRNA prediction

2. Oligo design

- 3¢

- 3¢ - 5¢

3. Oligo hybridization

Restriction enzyme specific overhangs

shA + B

4. Cloning of annealed oligos into shA + B expression vectors

shA + B

Constitutive vector

loxP vector

tetO/R vector

Constitutive KD

Conditional KD, Cre-inducible; Section 4.2.1

Conditional KD, Dox-inducible; Section 4.2.2

Section 4.1

5¢ -

Sense

Loop GAAGCTTG

Antisense

Termination TTTTTT

- 3¢

5. Expression of shRNA

shRNA

Figure 20.1 Design of shRNAs and approaches for their expression. For the coding sequence (CDS) of a target gene five siRNA sequences are predicted using free prediction programs. A pair of self-complementary oligonucleotides shA and shB is designed for each siRNA sequence. The siRNA sequence (‘‘sense,’’ white arrow) is followed by its complementary sequence (‘‘antisense,’’ black arrow). Both elements are separated by an 8-bp loop region. Oligos are annealed to a cloneable DNA fragment with restriction enzyme specific overhangs. We present here approaches for subsequent cloning of annealed oligos in expression vectors of shRNAs (detailed protocols in Sections 4.1–4.2.2) for constitutive and conditional knockdown (KD). In all approaches a pol III promoter drives transcription of the shRNA sequence to an RNA fragment where sense and antisense sequences are linked by the loop region. The latter allows for bending and thus annealing of the complementary regions, hence establishing the shRNA conformation.

392

Aljoscha Kleinhammer et al.

Depending on the knockdown approach, we ligate the hybridized oligos shA/B into one of three shRNA expression vectors. These vectors allow for: (i) constitutive knockdown (see Section 4.1), (ii) irreversible induction of knockdown by Cre-mediated recombination in a regional and temporal controlled manner (see Section 4.2.1), or (iii) reversible induction of knockdown by doxycycline in a temporal controlled manner (see Section 4.2.2). In each of these cases, gene silencing results from the expression of shRNAs driven by an RNA pol III promoter (Fig. 20.1E).

2.2. Genomic integration of vectors and generation of knockdown mice To generate a mouse line that expresses a chosen shRNA sequence, the corresponding expression vector must be stably integrated into the mouse genome. Since homologous recombination is laborious, we use RMCE for the insertion of a single shRNA vector copy into a modified Rosa26 allele of ES cells (Hitz et al., 2007) that serves as genomic acceptor site. This Rosa26 allele provides a pgk promoter driven hygromycin resistance gene (hygro) flanked by a pair of phiC31 Integrase (C31Int) attP recognition sites (Fig. 20.2). Integrase C31Int originates from the bacteriophage phiC31 and recombines one attB and one attP recognition site in a unidirectional manner. Any gene cassette flanked by two attB sites represents an RMCE donor allele and will be exchanged by C31Int against the attP flanked hygro gene within the modified Rosa26 allele. For the approach described here, the RMCE donor vector consists of the shRNA expression construct ligated with a linked promoterless neomycin resistance gene (neo). RMCE therefore results in the genomic integration of a single copy shRNA gene cassette into Rosa26 of ES cells. Since the Rosa26 locus is ubiquitously active, integrated shRNA constructs are produced in corresponding mouse lines in all mouse tissues and induce body-wide gene silencing (Seibler et al., 2005). An shRNA expressing mouse line is generated upon the selection of RMCE ES cell clones. For this purpose, the RMCE donor vector is electroporated into an ES cell line that harbors the modified Rosa26 allele as RMCE acceptor site. In the same step, a C31Int expression plasmid is coelectroporated. Once the C31 Integrase is expressed, the enzyme mediates the integration of the shRNA gene cassette into the Rosa26 RMCE allele. Upon successful cassette exchange, ES cells express the neomycin resistance gene driven by the pgk promoter inserted into Rosa26. By selection for neomycin-resistant ES cell clones, RMCE positive cell lines are isolated and subsequently used for injection into blastocysts. The resulting chimeric mice are then further mated to establish an shRNA transgenic mouse line.

393

Gene Knockdown in the Mouse Through RNAi

A

attP

attP pgk

Hygromycin

IDG3.2 Rosa26.10 ES cells

RMCE acceptor allele

Electroporation

B attB

attB neo IDG3.2 Rosa26.10 ES cells

U6 shRNA

RMCE donor allele

C31 integrase attP

attP pgk

Hygromycin

RMCE acceptor allele

C attR

attP pgk

neo

Rosa26.10 constitutive shRNA allele

U6 shRNA

IDG3.2 Rosa26.10 shRNA ES cells

shRNA

Testing of KD by WB, RT

D IDG3.2 Rosa26.10 shRNA ES cells

Injection of transgenic cells into wildtype blastocyst Wild-type blastocyst

E

Constitutive knockdown mouse

Characterization of KD by WB, RT, in situ hybridization

Figure 20.2 Genomic integration of shRNA vectors by RMCE into Rosa26 of ES cells and subsequent generation of constitutive knockdown mice. (A) IDG3.2 Rosa26.10 cells contain a modified Rosa26 locus providing a pgk promoter driven hygromycin resistance gene (hygro) flanked by a pair of C31 Integrase attP recognition sites. This configuration represents the acceptor allele for RMCE (recombinase-mediated cassette exchange). (B) After electroporation of the donor allele carrying the shRNA expression cassette, C31 Integrase catalyzes the recombination of the attB and attP sites leading to RMCE of hygro by the Neo-U6-shRNA segment. (C) ES cells that underwent correct recombination (IDG3.2 Rosa26.10 shRNA ES cells) can be selected via the neomycin gene expressed from the pgk promoter. These cells are already expressing shRNAs and

394

Aljoscha Kleinhammer et al.

2.3. Timescale and principle of shRNA validation Since the gene silencing efficiency of an siRNA target site cannot be predicted by currently existing programs various siRNA target sites have to be compared to assess their knockdown capacities in vivo. In our experience, five predicted siRNA sequences should be compared to identify one or two sequences that result in 90% knockdown of the target mRNA. To avoid the production of knockdown mice for all five shRNAs, we pretest their efficacy in ES cells (Fig. 20.3). In a first step, we transiently transfect all five vectors into ES cells and assess their gene silencing efficiency by Western blot or real-time PCR. If the target gene is not expressed in ES cells a cDNA expression vector can be cotransfected with the shRNA plasmids. For transient transfection, shRNA vectors are electroporated into wild-type ES cells. After 48 h, the cells are harvested to analyze the knockdown efficiency of a given vector. The two most efficient vectors, that induce at least 70% knockdown, are used for the next step. In transient transfections, the vector copy number and thus the number of produced shRNAs varies among the transfected cells, such that stable transfection of a single vector copy is important for a reliable validation of its gene silencing Day 0 1 day

Selection of 5 siRNA sequence candidates

1 week

Cloning of 5 shRNA vectors

2 weeks

Transient transfection of 5 vectors in ES cells

Analysis by WB, RT-PCR

6 weeks

Stable transfection of 2 best vectors in ES cells

Analysis by WB, RT-PCR

6 months Generation of 2 knockdown mouse lines

Analysis by WB, RT-PCR, in situ

6 months

Figure 20.3 Timescale and paradigm of shRNA validation. Initially five siRNA target sites are selected based on the sequence of the target gene. Including cloning and validation of the shRNA vectors, knockdown mouse lines can be obtained within 6 months. WB, Western blot; RT-PCR, real-time PCR; in situ, in situ hybridization.

can be used for validation of knockdown (KD) efficiency by Western blot (WB) or qRT-PCR (RT) analysis. (D) IDG3.2 Rosa26.10 shRNA ES cells are injected into wild-type blastocysts for generation of chimera and subsequent generation of transgenic knockdown mice. (E) Constitutive knockdown mice are characterized for knockdown efficiency by WB, RT-PCR, and in situ hybridization.

Gene Knockdown in the Mouse Through RNAi

395

efficiency. For this purpose, the two most efficient shRNA vectors are integrated into the Rosa26 RMCE allele to generate shRNA expressing ES cell lines. After cultivation for 48 h, the cells are harvested and the efficiency gene silencing is determined. Note that, if conditional shRNA vectors are used, it is necessary to perform this assay under inducing conditions. Upon the confirmation of gene silencing in ES cells, the shRNA ES cell lines are used for the production of corresponding shRNA transgenic mouse lines via blastocyst injection. The production and analysis of two lines of knockdown mice transgenic for different shRNA sequences is recommended because each shRNA may induce the silencing of nonintentional target genes as off-targets. This activity depends on the specific shRNA sequence, such that a phenotype common to a pair of shRNA lines is very likely resulting from the knockdown of the intended target. Phenotypes observed only in a single shRNA mouse line can be challenged to be related to off-target effects. shRNA mouse lines are first validated for the knockdown efficacy by Western blotting, RT-PCR or in situ hybridization and then analyzed for biological phenotypes. The generation of a pair of shRNA mouse lines usually takes about 6 months starting from the target gene sequence.

2.4. Principle of constitutive and conditional shRNA expression Depending on the scientific question and the specifics of a target gene, distinct expression systems for shRNA are required. Constitutive shRNA expression vectors produce body-wide gene silencing at all time points in development and adulthood since RNA pol III and the RNAi machinery are ubiquitously present. The constitutive knockdown approach is limited by the embryonic lethality of many mutants and the lack of tissue-specificity thus narrowing the mechanistic study of gene function. These limitations can be overcome by conditional approaches for shRNA expression that are presented in the following section. 2.4.1. Cre recombinase-inducible system for shRNA expression Starting point is an shRNA expression cassette that is structured as described in Section 2 and thus provides a HindIII restriction site in the loop region (Fig. 20.4). The insertion of a loxP-flanked transcriptional stop cassette into the HindIII site prevents the production of a complete shRNA. Upon the expression of Cre recombinase in cells, the stop cassette is excised from the shRNA construct, leaving a single loxP site within the shRNA loop region. This loxP site serves as loop region that links the sense and antisense sequence of the shRNA and allows the transcription of a functional shRNA. Considering the diversity of existing Cre and Cre-ER transgenic

396

Aljoscha Kleinhammer et al.

HindIII

A U6

Antisense

Sense

TTTTTT

Sense Antisense

+loxP flanked stop cassette

B SacIII

EcoRI U6

Sense

TTTTTT

TTTTTT

Antisense

TTTTTT

loxP

loxP

+Cre recombinase

C U6

Sense

Antisense

TTTTTT

loxP

Sense Antisense

Figure 20.4 Conditional shRNA vectors controlled by the Cre/loxP system. (A) A constitutive shRNA expression cassette is opened with HindIII for insertion of a loxPflanked stop cassette from plasmid pNEB-lox-stop-lox. (B) The stop cassette impairs completely shRNA expression. (C) Removal of the stop cassette by Cre recombinase enables shRNA expression leaving back a single loxP site that serves as a new loop region linking sense and antisense region of the shRNA transcript. EcoRI and SacII restriction sites used to test the orientation of the stop cassette.

mouse lines (see Nagy et al., 2009 for overview of existing Cre-lines), the Cre/loxP-directed shRNA approach enables sophisticated studies of gene function by tissue- and time-specific gene silencing. 2.4.2. Dox-inducible system for shRNA expression Based on the Escherichia coli tetracycline operon, the tet repressor/tet operator system represents a tool for the reversible and systemic induction of shRNA expression and gene silencing (Seibler et al., 2007; van de Wetering et al., 2003). In the absence of an inducer the tetracycline repressor (tetR) protein binds tightly to the tetracycline operator (tetO) DNA sequence

397

Gene Knockdown in the Mouse Through RNAi

A

tetR

tetR H1 tetO

−Dox

Sense

Loop Antisense TTTTTT

CAG tetR

Sense

Loop Antisense TTTTTT

CAG tetR

B tetR

+Dox

H1 tetO

Sense Antisense

C

IDG26.10

+

IDG26.10-sh

+

−

Dox GSK3a GSK3b b-actin

Figure 20.5 Conditional shRNA vectors controlled by the tetR/O system. (A) In the absence of doxycycline, the tetracycline repressor (tetR) binds to the tetracycline operator (tetO) thus impairing shRNA expression. (B) Upon addition of the inducer doxycycline, tetR dissociates from tetO resulting in transcription of shRNA from the H1 promoter. The tetR protein is constitutively produced under the control of a CAG promoter. (C) Dox-induced knockdown of GSK3b in ES cells detected by Western blot. A tetR/O system controlled expression vector for shRNAs directed against GSK3b was genomically integrated by RMCE in IDG26.10 cells. Cells lacking the construct (IDG26.10) or cells with the construct (IDG26.10-sh) but cultivated without inducer exhibit normal levels of GSK3b. Cultivation of IDG26.10-sh ES cells for 2 days with 1 g/ml doxycycline results in the knockdown of GSK3b without affecting GSK3a levels. Loading control: b-Actin.

resulting in inhibition of transcription by steric hindrance (Fig. 20.5A). Upon addition of tetracycline or its derivative doxycycline tetR dissociates from tetO allowing transcription to proceed. We use a tetO-modified H1 promoter (van de Wetering et al., 2003) that will therefore only drive expression of shRNAs in the presence of doxycycline (Fig. 20.5B). Thereby the target gene is silenced upon addition of doxycycline but is expressed at wild-type levels in the absence of inducer. See Fig. 20.5C for an example of efficient gene silencing by Dox-induced expression of shRNAs targeting GSK3b.

398

Aljoscha Kleinhammer et al.

3. Materials 3.1. Plasmids pbs-U6 pNeb-lox-stop-lox pEx-H1tetO-CAG-tetR pRMCE-II pCAG-C31Int pRosa26-50 -probe Plasmid sequence information can be found at: http://www.rnai.ngfn. de/index_4.htm.

3.2. Cell lines ES cell line IDG26.10-3 ES cell line IDG3.2 Feeder cells: G418-resistant murine embryonic fibroblasts Plasmids and cell lines are available from the authors.

3.3. Primers Neo (50 -GTT GTG CCC AGT CAT AGC CGA ATA G-30 ) Pgk (50 -CAC GCT TCA AAA GCG CAC GTC TG-30 ) Hyg-1 (50 -GAA GAA TCT CGT GCT TTC AGC TTC GAT G-30 ) Hyg-2 (50 -AAT GAC CGC TGT TAT GCG GCC ATT G-30 ) Rosa-50 (50 -CGT GTT CGT GCA AGT TGA GT-30 ) Rosa-30 (50 -ACT CCC GCC CAT CTT CTA G-30 )

3.4. Reagents 

Doxycycline: Doxycycline hyclate (Sigma). Dissolve in water (1 mg/ml) and store in single-use aliquots at 20  C.  F1 ES cell medium: 500 ml Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen), supplemented with 75 ml ES-qualified FCS (PAN Biotech), 9 ml 200 mM L-glutamine (Invitrogen), 12 ml 1 M HEPES (Invitrogen), 6 ml 100 MEM nonessential amino acids (Invitrogen), 1.2 ml 50 mM beta-mercaptoethanol (Invitrogen), 90 ml LIF (107 U/ml, Chemicon). Store at 4  C (stable for 4 weeks).  G418: Geneticin solution with an active concentration of 50 mg/ml (Invitrogen). Store in single-use aliquots at 20  C.

Gene Knockdown in the Mouse Through RNAi

     

399

Gelatine: 0.1% gelatine in water, autoclaved. PBS: 1 PBS (Invitrogen) for cell culture, sterile. Store at room temperature. Trypsin: 0.05% trypsin in Tris saline/EDTA buffer (Invitrogen). Store at 4  C. QIAprep Spin Miniprep Kit, QIAGEN Plasmid Maxi Kit (all Qiagen). QIAGEN Gel Extraction Kit (Qiagen). WizardÒ Genomic DNA Purification Kit (Promega).

4. Methods 4.1. Constitutive knockdown This protocol describes the design and cloning of shRNAs into the U6 promoter vector pbs-U6 that is derived from pShag (Paddison et al., 2002). 4.1.1. Cloning of constitutive active shRNA vectors 1. Analyze the CDS of your target gene with siRNA prediction tools (Section 2.1) and select five siRNA target sequences with a length of 20–25 nucleotides. For constitutive and Cre/loxP-regulated shRNA expression, the U6 promoter is used that required a G as first nucleotide of the siRNA/shRNA. Design a pair of complementary oligonucleotides shA and shB for each siRNA sequence as shown in Fig. 20.6A, providing vector compatible ends. We design shRNAs such that the sense region is transcribed first and the antisense region folds back via the 8 nt loop resulting in the order: U6promoter > target sense > loop > target antisense > termination (Fig. 20.1). For cloning of one shRNA, design two DNA oligonucleotides as shown in Fig. 20.6A. For oligonucleotide shA, start 50 with the selected sense sequence omitting the G at position þ1, add 8 nt loop sequence (GAAGCTTG), add the complete sense sequence in reverse orientation, and add 11 nt termination sequence (TTTTTT GGAAA). For oligonucleotide shB, start 50 with the 15 nt sequence containing the termination signal in reverse orientation (GATCTTTCCAAAAAA), add the complete target sense sequence, add the 8 nt loop sequence in reverse orientation (CAAGCTTC), add the target sense sequence in reverse orientation, and CG as final nucleotides. In the templates of the oligonucleotides shA and shB (Fig. 20.6A), the sense and antisense regions are filled by ‘‘N’’ and have to be complementary to each other. No phosphorylation or other modifications are needed, but HPLC purification enhances the sequence accuracy of your product. 2. The two oligonucleotides are designed such that they hybridize with each other and form a double-stranded DNA fragment with

400

Aljoscha Kleinhammer et al.

A

Sense +2 Loop Antisense Termination - 3′ shA 5′ - NNNNNNNNNNNNNNNNNNNNGAAGCTTGNNNNNNNNNNNNNNNNNNNNCTTTTTTGGAAA shB 3′ - GCNNNNNNNNNNNNNNNNNNNNCTTCGAACNNNNNNNNNNNNNNNNNNNNGAAAAAAGCTTTCTAG - 5′

BseRI overhang

B

M13rev

HindIII

U6-promoter

BamHI overhang

M13rev

+1

AsiSI

U6-promoter

BseRI BamHI

shRNA sense

HindIII 0.35 kb

pbs-U6-shRNA

pbs-U6

shRNA antisense

SfiI

Open with BseRI + BamHI Ligate annealed oligos shA+B Redigest ligation with BamHI

shRNA vector

C attB

Neo

AsiSI

attB

Neo

SfiI attB U6-promoter

pRMCEU6-shRNA

pRMCE-II Open with AsiSI + SfiI

shRNA sense shRNA antisense attB

Ligate with 0.35 kb AsiSI-SfiI fragment from pbs-U6-shRNA

Figure 20.6 Generation of constitutive shRNA vectors and RMCE vectors. (A) ShRNA vectors are designed such that the U6 promoter drives the expression of a short RNA that includes a sense and antisense (19–29 nucleotide) sequence of the targeted gene. A short termination signal (TTTTTT) determines the end of the transcript. The sense and antisense target sequences are connected by an invariable loop region that enables the formation of a short hairpin (sh) RNA through selfhybridization. The shRNA sequence is cloned downstream of the U6 promoter of the plasmid pbs-U6 vector using a pair of self-complementary oligonucleotides with BseRI and BamHI compatible overhangs. N sequences (NNN. . .) must be replaced by the sense and antisense target gene sequence. The first base of oligonucleotide shA represents the second transcribed nucleotide (þ2) of the shRNA. The invariable G at position þ1 is contained within the vector. The shRNA loop region contains a unique HindIII site. (B) Cloning of an shRNA vector (pbs-U6-shRNA) through ligation of the annealed oligonucleotides shA and shB into pbs-U6, opened with BseRI and BamHI. Since shA/shB are not phosphorylated and the vector ends are not dephosphorylated, the background of recircularized vector is suppressed by BamHI digestion of the ligation reaction prior to transformation. In correct cloning products the vector BamHI site is not reconstituted. An adjacent M13rev primer binding site enables the sequence confirmation of the cloned products; the 0.35 kb U6 expression cassette can

Gene Knockdown in the Mouse Through RNAi

3.

4.

5. 6.

7. 8. 9.

10.

401

single-stranded overhangs compatible to a BseRI and BamHI site (Fig. 20.6A). Using these overhangs the synthetic DNA fragment is cloned into the vector pbs-U6, opened with BseRI and BamHI. The vector ends remain phosphorylated and can be ligated to the nonphosphorylated overhangs, provided the ligation reaction is redigested with BamHI that cuts only religated empty vector. This procedure avoids the purchase of costly phosphorylated oligonucleotides. Anneal complementary oligonucleotides shA/B (1 g/l) by dilution of 1 l shA and 1 l shB in 98 l TE (10 mM Tris-HCl pH 7.5, 1 mM EDTA) and incubate at 100  C for 5 min, followed by slow cooling over 20 min to room temperature. For preparation of the vector backbone digest 5 g pbs-U6 plasmid DNA (Fig. 20.6B) with 20 units BseRI and 20 units BamHI in a total volume of 70 l and incubate at 37  C for 2 h. Load the restriction digest on a 0.8% agarose gel, isolate the 3.2 kb vector fragment with the QIAGEN Gel Extraction Kit, and determine the concentration of the vector DNA by photometry or by estimation on an agarose gel. Ligate annealed shA/B with the opened vector by incubation of 2 l hybridization reaction (step 3) with 100 ng vector (step 4) in a 10-ml standard ligation reaction at 16  C overnight. To eliminate self-ligated empty vector, redigest the ligation with BamHI by adding 2 l 10 buffer for BamHI, 7 l H2O and 1 l BamHI (10 U/l). Incubate the reaction for 30 min at 37  C and then inactivate the enzyme for 10 min at 65  C. Use 1 l of the ligation reaction for transformation of competent E. coli cells (we use DH5a), plate and incubate the cells overnight at 37  C. Pick 12 colonies and inoculate cultures of 2 ml LB amp medium, shake overnight at 37  C. Isolate plasmid DNA by using the QIAprep Spin Miniprep Kit and digest 10 l of each Miniprep DNA for 1 h at 37  C with PstI, that cuts upstream of the U6 promoter, and HindIII that cuts within the loop region. Analyze digestion products on a 1% agarose gel: correctly ligated clones show two bands of 310 and 2.9 kb, plasmids without insert show only the 2.9 kb vector band. Check one positive clone for the correct sequence by sequencing with the M13 forward and M13 reverse sequencing primers.

be isolated by digestion with AsiSI and SfiI. (C) Cloning of constitutive shRNA vectors for RMCE in IDG26.10-3 ES cells. Plasmid pRMCE-II, containing a promoterless neo resistance gene and a pair of attB recognition sites, is opened with AsiSI and SfiI. This vector is ligated with the 0.35 kb U6 expression cassette from pbs-U6-shRNA (B), for production of constitutive knockdown mice.

402

Aljoscha Kleinhammer et al.

11. Inoculate a confirmed clone in 200 ml LB amp medium, shake overnight at 37  C and isolate plasmid DNA by using the QIAGEN Plasmid Maxi Kit. Use the five generated plasmids for the next testing steps. 4.1.2. Pretest of shRNA vectors by transient transfection into ES cells 1. To sterilize the DNA, mix 80 g of plasmid (Section 4.1.1) with 1/10 volume of 3 M sodium acetate (pH 5.3) and 2.5 volumes of 100% ethanol. After incubation for 2 h at  20  C, pellet the DNA by centrifugation. Wash the pellet with 70% ethanol and dissolve the DNA in 80 l sterile water under a laminar flow hood. Add PBS to a final volume of 800 l. Use the five sterilized plasmids for transient transfection into IDG3.2 ES cells. 2. Cultivate IDG3.2 ES cells on gelatine for at least 2 days. 3. Trypsinize two to three 10-cm plates of IDG3.2 ES cells; one dish usually yields 1–2  107 cells. For one transfection, 5  106 cells are pelleted and washed with PBS. Discard the supernatant and resuspend the cells in 800 l PBS containing the plasmid DNA (step 1). All five plasmids should be tested in parallel. 4. Transfer the cell suspension into an electroporation cuvette (Biorad 4 mm), electroporate at 330 V (3 ms time constant) and let the cells rest for 5 min. Resuspend the cells in 10 ml medium and plate on a gelatine-coated 10-cm plate. Replace medium after 24 h. 5. After 48 h, harvest the cells by trypsinization. Pellet the cells, wash with PBS, and use the cell pellet directly for RNA or protein extraction or store at  80  C. 6. Analyze gene silencing efficiency for each of the five plasmids by determining expression level of your target gene by Western blotting or real-time PCR. 7. Select the two best performing shRNA vectors yielding at least 70% knockdown and use them for stable transfection of ES cells by RMCE. 4.1.3. Stable transfection of shRNA vectors into ES cells by RMCE 4.1.3.1. Construction of RMCE vectors For RMCE into Rosa26, the shRNA unit is ligated into the pRMCE-II vector that provides elements (neomycin resistance, attB sites) required for RMCE. 1. For ligation of a U6-shRNA expression cassette into the exchange vector pRMCE-II, we use the restriction sites AsiSI and SfiI. Digest 4 mg pRMCE-II and 4 g of your pbs-U6-shRNA plasmid with 40 U AsiSI for 1 h at 37  C. Add 40 U SfiI to each reaction and incubate for 1 h at 50  C.

Gene Knockdown in the Mouse Through RNAi

403

2. Load the restriction digests on a 1% agarose gel and isolate the 4.3 kb pRMCE-II vector backbone and the 350 bp AsiSI–SfiI U6-shRNA expression cassette (pbs-U6 backbone: 3 kb) by gel extraction. Determine the concentration of the insert DNA by photometry or gel electrophoresis and set up a ligation reaction. 3. Transform the ligation reaction into competent E. coli cells, plate cells on agar plates containing ampicillin, and incubate overnight at 37  C. 4. Pick 12 colonies and inoculate cultures of 2 ml LB amp medium, shake overnight at 37  C. 5. Isolate plasmid DNA by using the QIAprep Spin Miniprep Kit and digest 10 l of each Miniprep DNA for 1 h at 37  C with SalI and HindIII. Analyze digestion products on a 1% agarose gel: correctly ligated clones show two bands of 1.6 and 3.0 kb, plasmids without insert show only the 4.3 kb vector band. 6. Sequence one positive clone of your pRMCE-II-U6-shRNA with M13 forward, M13 reverse, and bpA-for primers and compare the obtained sequences to the predesigned plasmid file. 7. Inoculate a confirmed clone in 200 ml LB amp medium, shake overnight at 37  C and isolate plasmid DNA by using the QIAGEN Plasmid Maxi Kit. Use the two generated plasmids for RMCE in ES cells. 4.1.3.2. RMCE into Rosa26 of ES cells

1. Sterilize 25 g of each of the RMCE shRNA vectors selected in the previous section and 50 g of C31 Integrase expression vector as described in Section 4.1.2. Dissolve DNA in 1 l sterile water per used microgram DNA and add 25 l of C31Int plasmid to each RMCE vector. Add PBS to a final volume of 800 l and use the two plasmid preparations for transfection. 2. Cultivate IDG26.10-3 ES cells on feeders for 2 days at least. IDG26.10-3 ES cells contain a modified Rosa26 locus serving as acceptor allele for RMCE. 3. After trypsinizing the cells, pellet 5  106 cells per transfection sample and wash cells once with PBS. Discard the supernatant and resuspend the pellet in 800 l PBS containing the plasmid DNAs (step 1). Both plasmid combinations should be transfected in parallel. 4. Transfer the suspension into an electroporation cuvette, electroporate at 300 V (2 ms time constant), and let the cells rest for 5 min. Resuspend the cells in 10 ml medium and plate them on a 10-cm plate containing neomycin-resistant feeder cells. Replace medium 1 day after transfection. 5. At days 2–8 after transfection, select for neomycin-resistant clones by providing fresh medium with G418 (140 g/ml) every day. 6. Day 9: Pick 12 (or more) colonies per transfected shRNA–RMCE vector. Before colony picking, wash the plate with 5 ml PBS and add

404

Aljoscha Kleinhammer et al.

8 ml PBS. Transfer each colony in a well of a 96-well plate containing 50 l trypsin solution. 7. After 10 min at room temperature, add 50 l medium to the cells and transfer each colony into a well of a 96-well feeder plate containing 100 l medium. 8. Expand cells on feeders for freezing and on gelatine for DNA isolation. 9. Extract genomic DNA and identify positive clones by PCR or Southern blot as described in the following steps.

4.1.3.3. Identification of RMCE positive ES cells clones It is possible to identify positive clones via PCR or Southern blotting. Both methods are shown in the following.

1. First, extract genomic DNA of ES cell clones from the gelatine-coated 24-well plate using, for example, the WizardÒ Genomic DNA Purification Kit. 2. To identify positive clones by PCR, use primers located in the pgk promoter and in the neomycin resistance gene (primers Neo and Pgk, Section 3; annealing temperature: 65  C; product size 280 bp; if you use feeder cells harboring a pgk–neo resistance gene, an additional band of 160 bp appears); ES cell clones that did not undergo a complete recombination event (partial recombination, random integration, mixed clones) contain the hygro gene and can be identified by a second, hygro-specific PCR (primers Hyg-1 and Hyg-2, Section 3, annealing temperature: 65  C; product size: 550 bp; if you use feeder cells also harboring a hygro resistance gene, this PCR cannot be used for screening). Thus, for the typing of ES cell clones, two PCR reactions (neo, hygro) should be run, and for the typing of mice from tail DNA, the pgk–neo PCR is sufficient. The Rosa26 wild-type allele can be identified with the primers Rosa-50 and Rosa-30 (Section 3, annealing temperature: 57  C; product size: 536 bp). 3. To verify correct clones, it is important to perform Southern blot analysis. Only by this method, partially recombined clones or chromosomal rearrangements can be recognized. For Southern blotting, we follow standard protocols. To distinguish the wild-type and RMCE Rosa26 alleles, digest genomic DNA with EcoRV and use the 450 bp EcoRI fragment from plasmid pRosa-50 -probe as genomic 50 -probe. The Rosa26 locus gives rise to a wild-type band of 11.5 kb. Positive RMCE clones show an additional band of 13.7 kb due to the integration of U6-shRNA and neo gene cassette. Partially recombined clones show a band of 7.4 kb, and the parental IDG26.10-3 ES cells show a band of 4.5 kb derived from the modified Rosa26 allele before RMCE. Usually, 40% of the neomycin-resistant clones undergo complete RMCE;

Gene Knockdown in the Mouse Through RNAi

405

 10% contain a partial recombination event and 50% harbor integrations at unrelated genomic sites. 4. Select positive clones and expand clones until freezing aliquots in liquid nitrogen. 5. Use the RMCE shRNA ES cell clones to confirm the efficiency of gene silencing in ES cells (in case the target gene is expressed in ES cells, Section 4.1.2) and to generate chimeric mice through blastocyst injection. 4.1.4. Generation of shRNA transgenic mice The generation of chimeric mice by blastocyst injection is beyond the scope of this chapter, protocols are described elsewhere (Reid and Tessarollo, 2009). Here, we give only the specifics of using IDG26.10-3 ES cells. This ES cell clone is derived from the ES cell line IDG3.2 that was established from a male (C57BL/6J  129SvEvTac/S6)F1 hybrid blastocyst (Hitz et al., 2007). Hence, clones derived from the IDG3.2 ES line are heterozygous (Aw/a) at the agouti locus and can be injected into C57BL/6 blastocysts. Chimeric mice are recognized by the presence of agouti-colored fur on the background of the black host. We usually mate only male chimeras to C57BL6/J females to identify males that give rise to agouti-colored offspring. All agouti and black-colored pups derived from such males are then genotyped for the presence of the shRNA/RMCE allele by PCR (or Southern blotting, Section 4.1.3.3) because half of the ES cell derived sperm inherit the a allele, leading to black fur color.

4.2. Conditional knockdown 4.2.1. Cre-inducible conditional knockdown This section describes the conversion of constitutively active shRNA vectors (Section 4.1) into conditional vectors that can be activated in vitro or in vivo by Cre recombinase (Fig. 20.4). First a loxP-flanked stop element is inserted into the unique HindIII restriction site within the shRNA loop region; for control the stop element is deleted by Cre recombination in vitro to generate the activated vector. The functionality of the conditional and the activated version of the vectors can be confirmed by transient transfection into ES cells. 4.2.1.1. Insertion of the loxP-flanked stop element

1. For preparation of the vector fragment, digest 5 g pbs-U6-shRNA plasmid DNA (from Section 4.1) with 10 units HindIII in a total volume of 20 l and incubate at 37  C for 2 h. Add 1 l dNTPs (10 M) and 1 l Klenow fragment, incubate at room temperature for 20 min. Dephosphorylate the 50 ends of the digested plasmid by incubation with 1 unit

406

A

Aljoscha Kleinhammer et al.

shRNA sense

HindIII

U6-promoter

shRNA antisense

shRNA sense

SfiI

U6-promoter

lox-stop-lox

AsiSI

AsiSI

shRNA antisense

SfiI 1.2 kb

pbs-U6-lox/lox-sh RNA

pbs-U6-shRNA Open with HindIII, fill ends Ligate with 863 bp fragment from pNEB-lox-stop-lox

Conditional shRNA vector (loxP)

B

attB

Neo

AsiSI

AsiSI

SfiI attB

U6-promoter

shRNA sense

Neo

lox-stop-lox

attB

shRNA antisense

pRMCE-U6lox/lox-shRNA

pRMCE-II

SfiI attB

Open with AsiSI + SfiI Ligate with 1.2 kb AsiSI-SfiI fragment from pbs-U6-lox/lox-shRNA

Conditional shRNA vector (loxP) for RMCE

Figure 20.7 Generation of conditional Cre/loxP shRNA vectors and RMCE vectors. (A) ShRNA vector pbs-U6-shRNA (Fig. 20.6) can be turned into a conditional shRNA vector by opening at the HindIII site within the shRNA loop region, end filling, and ligation with an 863-bp MlyI fragment from plasmid pNEB-lox-stop-lox that act as removable stop element between the shRNA sense and antisense regions. A single loxP site remains within the shRNA loop of the activated construct, but does not affect RNAi efficiency. The 1.2-kb conditional U6 expression cassette can be isolated from pbs-U6-lox/lox-shRNA by digestion with AsiSI and SfiI. B Cloning of conditional shRNA vectors for RMCE in IDG26.10-3 ES cells. Plasmid pRMCE-II, containing a promoterless neo resistance gene and a pair of C31 Integrase attB recognition sites, is opened with AsiSI and SfiI. This vector is ligated with the 1.2-kb AsiSI–SfiI conditional shRNA cassette from pbs-U6-lox/lox-shRNA (A) for production of conditional, Cre/ loxP-regulated knockdown mice.

shrimp alkaline phosphatase (SAP) at 37  C for 1 h. Inactivate all enzymes by incubation at 65  C for 10 min (Fig. 20.7A). 2. For preparation of the insert containing the stop cassette, digest 10 g of the plasmid pNEB-lox-stop-lox with 10 units MlyI in a total volume of 40 l and incubate at 37  C for 2 h. Inactivate the enzyme by incubation at 65  C for 10 min. 3. Load both restriction digests on a 1% agarose gel, isolate the fragments with the QIAGEN Gel Extraction Kit. The vector fragment is 3.3 kb in size and the loxP-stop-loxP insert fragment is 863 bp in size (other

Gene Knockdown in the Mouse Through RNAi

4.

5. 6. 7.

8. 9.

407

fragments of this digest: 263, 486, 502, and 1194 bp). Determine the concentration of the DNA fragments by photometry or gel electrophoresis. Set up a 10-ml ligation reaction with 50 ng purified pbs-U6-shRNA vector fragment and 35 ng purified loxP-stop-loxP insert fragment. Upon ligation, self-ligated empty vector can be eliminated by redigestion of the ligation with HindIII by adding 2 l 10 buffer NEB2, 7 l H2O, and 1 l HindIII (10 U/l). Incubate the reaction for 30 min at 37  C and inactivate the enzyme again for 10 min at 65  C. Use 5 l of the ligation reaction for transformation of 50 l competent E. coli cells, like DH5a, and plate cells on agar plates containing ampicillin (100 g/ml). Incubate the plates overnight at 37  C. Pick 12 colonies and inoculate cultures of 2 ml LB amp medium, shake overnight at 37  C. Isolate plasmid DNA by using the QIAprep Spin Miniprep Kit and digest 10 l of each Miniprep DNA for 1 h at 37  C with EcoRI and SacII. Analyze digestion products on a 1% agarose gel: correctly ligated clones show two bands of 850 bp and 3.4 kb, plasmids with insert in the false orientation show two bands of 150 bp and 4 kb, and plasmids without insert show only the 3.3 kb vector band. Sequence one positive clone with M13 forward, M13 reverse primers, and compare the obtained sequences to the predesigned plasmid file. Inoculate a confirmed clone in 200 ml LB amp medium, shake overnight at 37  C, and isolate plasmid DNA by using the QIAGEN Plasmid Maxi Kit.

4.2.1.2. In vitro deletion of the loxP-flanked stop cassette

1. To delete the stop cassette and to activate shRNA expression in vitro, incubate 1 g DNA of the conditional shRNA plasmid with 4 units Cre recombinase in a total volume of 50 l 1 Cre buffer at 37  C for 30 min. Inactivate the recombinase for 10 min at 70  C. 2. To eliminate nonrecombined plasmids, add 1 l EcoRI (10 U/l) together with 3.3 l 1 M NaCl and 0.5 l BSA (100, NEB) and incubate the reaction for 30 min at 37  C. This restriction digest will linearize all molecules, which still harbor the stop cassette. Heat inactivate the enzyme at 65  C for 20 min. 3. Use 2 l of the recombination reaction for transformation of 50 l competent E. coli cells, like DH5aTM, and plate cells on agar plates containing ampicillin (100 g/ml). Incubate the plates overnight at 37  C. 4. Pick six colonies grown on the agar plates and inoculate cultures of 2 ml LB amp medium, shake overnight at 37  C. Isolate plasmid DNA from the culture using the QIAprep Spin Miniprep kit. Use 10 l of each

408

5. 6. 7.

8.

Aljoscha Kleinhammer et al.

Miniprep plasmid DNA to set up a 30-l digestion reaction with 5 units XbaI, incubate for 1 h at 37  C, and analyze the digestion products on a 1% agarose gel. Correctly recombined clones show two bands of 460 bp and 2.9 kb and plasmids, which have not been recombined by Cre recombinase show two bands of 1.3 and 2.9 kb. Sequence one positive clone with M13 forward and M13 reverse and compare the obtained sequences to the predesigned plasmid file. Inoculate a confirmed clone in 200 ml LB amp medium with ampicillin (50 g/ml) and shake it overnight at 37  C. Isolate the plasmid DNA using the QIAGEN Plasmid Maxi Kit. Evaluate the efficiency of the conditional shRNA constructs with the stop cassette and after recombination of the stop cassette by transient transfection into ES cells and analyze gene silencing by Western blotting or real-time PCR (Section 4.1.2). Choose the best constructs in the conditional state for transfer into ES cells by RMCE.

4.2.1.3. Construction of RMCE vectors and RMCE in ES cells The construction of RMCE vectors with conditional shRNA units and RMCE into the Rosa26 locus of IDG26.10-3 ES cells is largely the same as described for constitutive shRNA units as described in Section 4.1.3. Here we point out only those steps that differ to the protocol in the various sections as follows: 

Section 4.1.3.1, step 2

Load the restriction digests on a 1% agarose gel and isolate the 4.3 kb pRMCE-II vector backbone and the 1.2 kb AsiSI–SfiI U6-shRNA expression cassette (pbs-U6 backbone: 3 kb) by gel extraction. Determine the concentration of the insert DNA by photometry or gel electrophoresis and set up ligation reaction (Fig. 20.7B). 

Section 4.1.3.1, step 5

Isolate plasmid DNA by using the QIAprep Spin Miniprep Kit and digest 10 l of each Miniprep DNA for 1 h at 37  C with SalI and EcoRI. Analyze digestion products on a 1% agarose gel: correctly ligated clones show two bands of 1.7 and 3.0 kb, plasmids without insert show only the 4.3 kb vector band. 

Section 4.1.3.3, step 3

To distinguish the wild-type and RMCE Rosa26 alleles, digest genomic DNA with EcoRV and use the 450 bp EcoRI fragment from plasmid pRosa50 -probe as genomic 50 -probe. The Rosa26 locus gives rise to a wild-type band of 11.5 kb. Positive RMCE clones show an additional band of 15 kb

Gene Knockdown in the Mouse Through RNAi

409

due to the integration of U6-shRNA and neo gene cassette. Partially recombined clones show a band of 10 kb and the parental IDG26.10-3 ES cells show a band of 4.5 kb derived from the modified Rosa26 allele before RMCE. 4.2.1.4. Generation of shRNA transgenic mice See also the text in Section 4.1.4. Mice harboring a conditional shRNA cassette must be crossed with a Cre transgenic mouse line to activate shRNA expression through deletion of the loxP-flanked stop cassette. The specificity and efficiency of Cre-mediated vector activation can be controlled by Southern blot analysis of EcoRV-digested genomic DNA isolated from mouse tissues. Using the Rosa26 50 -probe for hybridization, the wild-type Rosa26 allele gives a band of 11.5 kb, the nonrecombined shRNA allele gives a band of 15 kb, while Cre-mediated deletion of the 800 bp loxP-flanked stop cassette reduces the band of the Rosa26 shRNA allele to 14.2 kb.

4.2.2. Dox-inducible conditional knockdown In the following section, we describe how to generate vectors for reversible control of shRNA expression via the tetR/O system (Section 2.4.2). The shRNA sequence is introduced into the vector pEx-H1tetO-CAG-tetR containing a tetO-modified H1 promoter and a codon-optimized tetR gene under the control of the constitutive CAG promoter. As the vector provides in addition a promoterless neomycin resistance gene and a pair of attB sites, the generated cloning product can be used without further cloning for RMCE in IDG26.10-3 ES cells (Fig. 20.7B and Fig. 20.8). 4.2.2.1. Cloning of conditional tetR/O shRNA vectors

1. Upload the CDS of your target gene to the mentioned web resources (Section 2.1) and select five siRNA sequences. In contrast to the U6 promoter used in the constitutive and the cre/loxP approach, any nucleotide can be used at position þ1 to be transcribed by the H1 promoter. 2. Design a pair of complementary oligonucleotides shA and shB for each siRNA sequence as shown in Fig. 20.8A, providing BsaBI and AscI compatible ends after annealing. 3. Anneal complementary oligonucleotides shA/B (c ¼ 1 g/l) by incubating 1 l shA and 1 l shB in 98 l TE at 100  C for 5 min and subsequent slow cooling down to room temperature. 4. Open the plasmid pEx-H1tetO-CAG-tetR by digestion with AscI at 37  C for 1 h (Fig. 20.8B). 5. Digest reaction with BsaBI at 60  C for 1 h. 6. Load on 0.9% agarose gel and isolate 7 kb vector band.

410

A

Aljoscha Kleinhammer et al.

+1 Sense Loop Antisense Termination - 3′ shA 5′ -AGATGCCNNNNNNNNNNNNNNNNNNNNGAAGCTTGNNNNNNNNNNNNNNNNNNNNTTTTTT shB 3′ -TCTACGGNNNNNNNNNNNNNNNNNNNNCTTCGAACNNNNNNNNNNNNNNNNNNNNAAAAAAGCGC- 5′ AscI overhang

HindIII

B

M13for attB

H1tetO

M13for attB

BsaBI

H1tetO shRNA

AscI CAG

CAG htetR

pEx-H1tetOCAG-tetR

pEx-H1tetOshRNA-CAG-tetR

htetR

Open with BsaBI + AscI Neo attB

Ligate annealed oligos shA + B Redigest ligation with BsaBI + AscI

Neo attB

Dox-inducible shRNA vector for RMCE

Figure 20.8 Generation of conditional tetR/O shRNA vectors. (A) A pair of selfcomplementary oligonucleotides (shA/B) containing the shRNA sequence is designed and annealed resulting in a cloneable DNA fragment with an AscI compatible overhang at the 30 end. The 50 end is BsaBI compatible as this enzyme produces blunt ends as a substrate. White and black boxed sequences (NNN. . .) represent the target sequence in sense and antisense orientation, respectively. The loop region contains a HindIII site and ‘‘TTTTTT’’ serves as termination signal for pol III. (B) Cloning of a conditional, Dox-inducible shRNA vector (pEx-tetO-shRNA-CAG-tetR) through ligation of the annealed oligos shA and shB into the vector pEx-H1-tetO-CAG-tetR, opened with BsaBI and AscI. Redigestion of the reaction with BsaBI and AscI is recommended to cut religated plasmid. As the vector provides attB sites and a promoterless neomycin resistance gene, it can be directly used for recombinase-mediated cassette exchange (RMCE).

7. Extract DNA by gel extraction and determine concentration of opened vector. 8. Ligate annealed shA/B into the opened vector by incubation of 2 l hybridization reaction (step 3) with 100 ng vector (step 7) in a typical ligation reaction at 16  C overnight (Fig. 20.8B). 9. For removal of religated plasmid, digest first with AscI for 10 min at 37  C and second with BsaBI for 10 min at 60  C. Heat inactivate restriction enzymes by final incubation for 15 min at 75  C. 10. Use 1 l of the ligation reaction for transformation of competent E. coli cells, plate and incubate cells overnight at 37  C. 11. Pick colonies and inoculate 12 cultures of 2 ml LB amp medium, shake overnight at 37  C. 12. Isolate plasmid DNA by using the QIAprep Spin Miniprep Kit and digest 10 l of each Miniprep with PstI and HindIII for 1 h at 37  C. Analyze digestion reactions on a 0.9% agarose gel: Correctly ligated

Gene Knockdown in the Mouse Through RNAi

411

clones show bands of 0.2, 1.0, 2.8, and 3.3 kb. Empty plasmids show bands of 0.2, 1.0, and 6.1 kb. 13. Check one positive clone for the correct sequence by sequencing with M13 forward and M13 reverse. 14. Inoculate a confirmed clone in 200 ml LB amp medium, shake overnight at 37  C and isolate plasmid DNA by using the QIAGEN Plasmid Maxi Kit. Use the five generated plasmids of pEx-H1tetOshRNA-CAG-tetR for further testing steps. 4.2.2.2. Pretesting of tetR/O shRNA vectors by transient transfection in ES cells

1. For sterilization, mix 80 g of plasmid DNA (from Section 4.2.2.1) with 1/10 volume of 3 M sodium acetate (pH 5.3) and 2.5 volumes of 100% ethanol. After incubation for 2 h to overnight at 20  C pellet the DNA by centrifugation. Wash with 70% ethanol and dissolve the DNA pellet in 80 l sterile water while working under a sterile flow. Add PBS to a final volume of 800 l. Use the five sterilized plasmids for transient transfection in IDG3.2 ES cells. 2. Cultivate IDG3.2 ES cells on gelatine for 2 days at least. 3. Trypsinize two to three 10-cm plates with IDG3.2 ES cells on gelatine, one dish usually yielding 1–2  107 cells. For one transfection sample, 5  106 cells are pelleted and washed with PBS. Discard the supernatant and resuspend the pellet in 800 l PBS containing the plasmid DNA (step 1). All five plasmids should be tested in parallel. 4. Transfer the suspension into an electroporation cuvette, electroporate at 330 V for 3 ms, and let the cells rest for 5 min. Resuspend the cells in 10 ml medium containing 1 g/ml doxycycline and plate them on a gelatine-coated 10 cm plate. 5. After 24 h change medium, new medium must contain 1 g/ml doxycycline. 6. After 48 h harvest cells by trypsinization, pellet the cells, wash with PBS, and discard the supernatant. Use the cell pellet directly for analysis or store at  20  C. 7. Analyze gene silencing efficiency for each of the five plasmids by determining expression level of your target gene, for example, by Western blot or real-time PCR. 8. Select the two best performing shRNA vectors yielding at least 70% knockdown and use these for stable transfection by RMCE. 4.2.2.3. Stable transfection of tetR/O shRNA vectors by RMCE into ES cells The stable transfection of tetR/O vectors follows the same protocol as used for constitutive shRNA vectors, see Section 4.1.3.2. For the

412

Aljoscha Kleinhammer et al.

Southern blot analysis of tetR/O shRNA alleles, digest genomic DNA with EcoRV. Using the Rosa26 50 -probe RMCE alleles exhibit a band of 17.7 kb, as compared to the 11.5 kb wild-type fragment. 4.2.2.4. Testing of stably integrated tetR/O shRNA vectors in ES cells

1. Expand two positive clones per shRNA vector on feeders up to 10 cm plates. Freeze aliquots and transfer the ES cells to gelatine-coated dishes. 2. Cultivate the ES cells on gelatine for 2 days at least. 3. Start induction of shRNA expression by adding 1 g/ml doxycycline to the medium. 4. After 24 h, discard the medium and add fresh medium containing doxycycline (1 g/ml). 5. After 48 h, harvest the cells and analyze the four clones for gene silencing efficiency as described in Section 4.2.2.2, steps 6 and 7. 6. For each of the two shRNA vectors, choose the most efficient gene silencing clone and use this ES cell line for generation of the corresponding two shRNA mouse lines using blastocyst injection of tetraploid aggregation. 4.2.2.5. Generation of shRNA transgenic mice See also the text in Section 4.1.4. Mice harboring a Dox-inducible shRNA cassette must be supplied with doxycycline in the drinking water to activate shRNA expression. The level of shRNA production (and thereby the extend of gene silencing) can be regulated by the doxycycline concentration. For full vector activation, supply the drinking water with 2 mg/ml doxycycline and 10% sucrose (to cover the aversive taste of doxycycline). Protect the water bottles from light and replace every second day to prevent microbial contamination (Seibler et al., 2007).

5. Concluding Remarks The single copy shRNA vector approach described in this chapter enables to elicit body-wide or cell type-specific gene silencing in most organs of adult mice. In the brain of shRNA vector transgenic mice, the levels of a target gene mRNA and protein can be reduced by 90% or more, comparable to the utility of RNAi in cultured cell lines. However, to achieve a satisfying level of gene silencing in vivo, it is instrumental to identify well working RNAi target sequences through initial in vitro screening. As compared to the generation of knockout mice by gene targeting, the shRNA vector approach is technically less demanding, since vector construction and vector insertion into ES cells are simple and reproducible. The

Gene Knockdown in the Mouse Through RNAi

413

insertion of shRNA vectors into the Rosa26 locus as universal docking site enables the production of transgenic mice with reproducible expression properties. Upon the initial characterization of the basic parameters of this approach, as summarized in the previous section, any newly generated strain can be faithfully expected to represent a functioning knockdown mouse strain. In addition, the generation of knockdown mice is relatively fast—a first group of adult mutants can be generated within a period of 11 months. In a specific configuration of the described technique, shRNA expression can be controlled by doxycycline as inducing compound. This system, that enables the reversible and body-wide expression of shRNAs in mice (Seibler et al., 2007), offers the opportunity to reverse the induced gene knockdown at a given time. This property of the shRNA system offers unique applications to study gene function in mice that cannot be achieved with knockout technologies.

ACKNOWLEDGMENTS This work has been funded by the Volkswagen Foundation and the Federal Ministry of Education and Research (BMBF) in the framework of the National Genome Research Network (FKZ:01GR0404).

REFERENCES Brummelkamp, T. R., et al. (2002). A system for stable expression of short interfering RNAs in mammalian cells. Science 296, 550–553. Delic, S., et al. (2008). Genetic mouse models for behavioral analysis through transgenic RNAi technology. Genes Brain Behav. 7, 821–830. Dykxhoorn, D. M., and Lieberman, J. (2006). Knocking down disease with siRNAs. Cell 126, 231–235. Elbashir, S. M., et al. (2001). Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494–498. Hasuwa, H., et al. (2002). Small interfering RNA and gene silencing in transgenic mice and rats. FEBS Lett. 532, 227–230. Hitz, C., et al. (2007). Conditional brain-specific knockdown of MAPK using Cre/loxP regulated RNA interference. Nucleic Acids Res. 35, e90. Kuhn, R., et al. (2007). RNA interference in mice. Handb. Exp. Pharmacol. 178, 149–176. Kunath, T., et al. (2003). Transgenic RNA interference in ES cell-derived embryos recapitulates a genetic null phenotype. Nat. Biotechnol. 21, 559–561. Lee, N. S., et al. (2002). Expression of small interfering RNAs targeted against HIV-1 rev transcripts in human cells. Nat. Biotechnol. 20, 500–505. Nagy, A., et al. (2009). Creation and use of a Cre recombinase transgenic database. Methods Mol. Biol. 530, 365–378. Paddison, P. J., et al. (2002). Short hairpin RNAs (shRNAs) induce sequence-specific silencing in mammalian cells. Genes Dev. 16, 948–958. Reid, S. W., and Tessarollo, L. (2009). Isolation, microinjection and transfer of mouse blastocysts. Methods Mol. Biol. 530, 269–285.

414

Aljoscha Kleinhammer et al.

Rubinson, D. A., et al. (2003). A lentivirus-based system to functionally silence genes in primary mammalian cells, stem cells and transgenic mice by RNA interference. Nat. Genet. 33, 401–406. Seibler, J., et al. (2005). Single copy shRNA configuration for ubiquitous gene knockdown in mice. Nucleic Acids Res. 33, e67. Seibler, J., et al. (2007). Reversible gene knockdown in mice using a tight, inducible shRNA expression system. Nucleic Acids Res. 35, e54. Steuber-Buchberger, P., et al. (2008). Simultaneous Cre-mediated conditional knockdown of two genes in mice. Genesis 46, 144–151. van de Wetering, M., et al. (2003). Specific inhibition of gene expression using a stably integrated, inducible small-interfering-RNA vector. EMBO Rep. 4, 609–615.

C H A P T E R

T W E N T Y- O N E

In Vivo Analysis of Gene Knockdown in Tetracycline-Inducible shRNA Mice Christopher S. Raymond, Lei Zhu, Thomas F. Vogt, and Myung K. Shin Contents 1. Introduction 2. Methods 2.1. In vivo doxcycline induction shRNA 2.2. Harvesting and storage of mouse tissues following doxycycline-induced shRNA expression 2.3. Tissue homogenization and RNA purification from adult mouse tissues 2.4. cDNA generation from mouse tissues 2.5. Real-time PCR expression analysis of RNA knockdown in mouse tissues 2.6. Determining induced shRNA expression in mouse tissues 3. Notes References

416 419 419 420 420 422 422 423 426 426

Abstract Expression of small hairpin RNA (shRNA) in mammalian cells can trigger potent RNAi-mediated gene silencing. The dominant-acting RNAi can often result in phenotypes similar to that of a null allele. Moreover, the generation of shRNA knockdown mice and subsequent phenotypic analysis can be achieved in a condensed timeline compared to that of conventional gene-targeting knockout strategies. Here, we discuss methods for the in vivo analysis of gene function in adult mouse tissues following tetracycline-induced RNA knockdown in a single-copy inducible polymerase III promoter-driven shRNA system.

Merck & Co., Inc., Rahway, New Jersey, USA Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77021-X

#

2010 Elsevier Inc. All rights reserved.

415

416

Christopher S. Raymond et al.

1. Introduction Small interfering RNAs (siRNAs) are short (<30 bp), doublestranded RNA molecules that can be used to elicit gene-specific silencing in diverse species and cultured cells (Hannon, 2002). This can either be achieved in vitro by transfecting cells with commercially synthesized siRNAs or through the use of vectors that express short hairpin RNAs (shRNAs) (Elbashir et al., 2001; Paddison et al., 2002; Siolas et al., 2005). The shRNA expression vectors can also be stably integrated into the mouse genome and used to generate transgenic mouse stains that exhibit a constitutive knockdown phenotype, which can often be similar to that obtained with a conventional knockout mouse model (Coumoul et al., 2005; Kunath et al., 2003; Lickert et al., 2004; Oberdoerffer et al., 2005; Seibler et al., 2005). Most commonly, shRNA expression vectors are designed to consist of an RNA polymerase III (pol III) promoter, such as U6 or H1, which is used to drive high levels of shRNA transcription in almost all cell types. The pol III promoters are preferred over pol II promoters since they generate RNA transcripts that lack a 50 cap and polyadenylation tail. Due to the dominant acting nature of RNA interference (RNAi), the constitutive expressing shRNA vectors are not suitable for silencing genes that are essential for mouse development, or when it is simply desired to examine only the adult role of a gene’s function. Therefore, the use of an inducible knockdown (iKD) models would be advantageous in these instances. The generation of iKD models can be accomplished using a vector system that features a tetracycline (Tet)-inducible system (Gossen and Bujard, 1992). Briefly, this platform consists of a hybrid H1 promoter with the inclusion of a Tet operon sequence that is placed just prior to the transcriptional start site (van de Wetering et al., 2003). With the inclusion of a constitutively expressing minimal Tet repressor (TetR) into the vector, shRNA transcription is prevented and thus allowing for normal gene function prior to gene silencing. The shRNA expression can then be induced upon the addition of the tetracycline analog, doxycycline, which binds to the TetR protein, causing a conformational change and preventing it from binding to the Tet operator region (Fig. 21.1). In order to increase the predictability and reproducibility of the iKD phenotypes from animal to animal, a strategy for incorporating a single-copy inducible shRNA expression cassette into the ROSA26 genomic locus in mouse embryonic stem (ES) cells has been described (Seibler and Schwenk, 2010; Seibler et al., 2007). The ROSA26 locus has been used to embed and broadly express transgenes in mice (Soriano, 1999). Furthermore, the integration of transgenes into this locus is considered neutral because no phenotypic consequences have been reported in these lines. By incorporating strategies for a targeted single-copy inducible shRNA construct together with laser-assisted microinjection of mouse ES cells

417

In Vivo Inducible RNAi Analysis

A

−Dox

tetR

Off

H1-tetO

shRNA

CAGGs

tetR

B tetR On +Dox

H1-tetO

shRNA

CAGGs

tetR

Figure 21.1 (A) A single-copy tetracycline-inducible shRNA vector is targeted to the ROSA26 locus via homologous recombination in ES cells. In the absence of doxycycline, constitutive expression of the Tet Repressor (TetR) protein driven from a CAGGs promoter binds to the Tet Operon (TetO) sequence just prior to the transcriptional start site, thus preventing expression of the shRNA. (B) With the addition of doxycline, TetR is now bound causing a conformational change that prevents the molecule from binding the TetO sequence, thus allowing transcription of the shRNA driven by the pol III H1 promoter.

into eight cell embryos, can lead to the efficient generation of fully ES-derived gene-targeted mice that are directly available for phenotypic analysis in the G0 generation (Poueymirou et al., 2007). The single-copy iKD platform offers the ability to analyze gene function in a greatly compressed time period compared to the traditional conditional knockout (cKO) mice to interrogate gene function in adults. The iKDs are hypermorphs with varying levels of target gene KD across different tissues that could achieve >75% KD across many tissues with a potent shRNA (Seibler et al., 2007; Fig. 21.2). This level of mRNA modulation is equal to or better than tamoxifen-induced cKO. As an example, Takeda et al. (2007) demonstrated that a 7-day tamoxifen-induction of a ROSA26CreERT2 allele only lead to a 10–85% deletion efficiency across several tissues in a conditional floxed-Phd2 allele. Similarly, a 5–70% deletion efficiency of a conditional floxed-VHL allele was observed upon tamoxifen-induction with a chicken beta-actin-Cre-ER mouse line (Minamishima et al., 2008). In addition to the tissue-to-tissue variability, conditional gene inactivation in adult mice with a tamoxifen-inducible Cre can also lead to a large animal-toanimal variability compared to a iKD line. Gruber et al. (2007) showed that tamoxifen-induced Cre deletion of a floxed-Hif-2a allele varied significantly from animal to animal, ultimately leading to phenotypic variability in these groups. Finally, the cKO results in tissue mosaicism where a cell with a KO genotype is adjacent to a wild-type (WT) cell, whereas it is likely that iKD lines would have same level of KD for all cells within the same lineage. While using iKD models may solve many of the shortcomings associated with cKOs, there still remain some areas of concern to keep in mind with

418

Christopher S. Raymond et al.

A 100.0 80.0 60.0 40.0 20.0 0.0 Brain (Ct = 23)

Heart (Ct = 23)

Muscle (Ct = 23)

Lung (Ct = 24)

WT + Dox

WT − Dox

B

Kidney (Ct = 25)

Colon (Ct = 25)

Liver (Ct = 27)

Dox

− −

− +

Pancreas (Ct = 27)

shRNA + Dox

shRNA − Dox

Kidney

Liver shRNA

Spleen (Ct = 27)

+ −

+ +

− −

− +

+ −

+ +

Figure 21.2 (A) Example of RNA knockdown in 10 adult mouse tissues from a singlecopy tetracycline-inducible shRNA vector targeted to the ROSA26 locus. Note the leakiness of shRNA expression in the tissue samples from brain and kidney in the absence of doxycycline treatment. Ct values represent WT gene levels as determined by qPCR. (B) Western blot displaying corresponding decreased protein expression in liver and kidney tissues as a result of the tetracycline-inducible shRNA expression.

iKD models. The primary concern when generating a knockdown mouse model is the possible unintended effects on gene expression mediated by RNAi which are referred to as ‘‘off-target effects.’’ This is caused when the shRNA used contains partial complementary sequence to mRNAs in addition to that of the intended target gene. This can lead to the manifestation of phenotypes in the mouse model that are not the consequence of knocking down the gene of interest. Therefore, careful attention must be paid to the design of the shRNA sequence using the most advanced algorithms in order to minimize the chance of creating ‘‘off-target effects’’ in the knockdown model, and confirmation of the phenotype with two different shRNAs may be needed. Another area of concern with the iKD model lies in the potential toxicity associated with the administration of doxycycline to the mice. Although a recent study has shown very few transcriptome changes in the livers of a ‘‘tet-on’’ mouse model upon doxycycline treatment (Reboredo et al., 2008), investigators should be vigilant about potential doxycycline effects in their phenotypic assays. As more iKD models are created and analyzed, the full utility and limitations of this platform will be further defined. Here, we describe in detail, protocols for assessing in vivo modulation of target genes in iKD mice following doxycycline-induction of shRNA.

In Vivo Inducible RNAi Analysis

419

2. Methods 2.1. In vivo doxcycline induction shRNA Doxycycline-inducible germ line RNAi mice are generated as previously described (Seibler et al., 2007). Cohorts of 6–8 adults (8–12-week-old) that are heterozygous for the inducible shRNA allele are set up with one group receiving doxycycline treatment while the other group receives no treatment and serves as a control. Similar cohorts of WT mice are also set up to receive either doxycycline treatment or no treatment. Doxycycline is a water-soluble tetracycline derivative that in regard to this platform’s design serves to relieve the repression of shRNA transcription. Doxycycline can be administered to mice through several routes including by oral gavage, in jelly, but most commonly either by the addition of doxycycline to the drinking water or mouse diet (Cawthorne et al., 2007). The following protocol for a doxycline-containing drinking water solution generally yields robust RNA knockdown in most adult mouse tissues. Doxycycline drinking water solution (2 mg/ml): 1 g doxycycline (D9891, Sigma) 5 mM saccharin (S4067, Sigma) Sterile tap water to 500 ml It should be noted that doxycyline is a light-sensitive material and therefore it is recommended that precautions should be taken, such as the use of brown or opaque water bottles, to protect it from prolonged exposure to light. Furthermore, saccharin is included in the drinking water solution to mask the bitter taste of doxycycline (Ste Marie et al., 2005). The doxycycline drinking water solution must be prepared fresh every 2 days. A 10-day induction with doxycycline is usually sufficient to drive maximal RNA knockdown in most tissues; however, the exact length of doxycycline treatment required to reveal the maximum phenotypic manifestations of RNA knockdown will need to be determined empirically by the investigator. Alternatively, mouse chow with added doxycycline is also available from most rodent diet vendors (Teklad, TD.98186; Bio-Serv S3888). These diets are commonly supplied at a doxycycline concentration of 200 mg/kg, but diets with custom doxycycline concentrations are also available. The benefit of using doxycycline-containing diet as opposed to water delivery is the increased stability due to the low degradation rate of doxycycline owing to the reduced exposure to light in the diet. Therefore, a replacement of the diet is only required every 1–2 weeks. The doxycycline chow can be stored for up to 6 months at 4  C, and is available in an irradiated form for use in mouse barrier facilities.

420

Christopher S. Raymond et al.

2.2. Harvesting and storage of mouse tissues following doxycycline-induced shRNA expression For an accurate determination of RNA levels in mouse tissues, it is important to isolate a high yield of intact RNA devoid of contaminating genomic DNA. The use of RNAlaterÒ (Ambion) allows harvested adult mouse tissues to be stored for long periods of time without significant degradation to the RNA quality. The following method provides for storage of highquality tissues until further processing. 1. Dissect and harvest tissues of interest for RNA knockdown analysis from adult animals following doxycycline shRNA induction. 2. Cut tissue samples into small pieces (tissue pieces not to exceed a thickness of 0.5 cm) before immersion into RNAlaterÒ Solution. 3. Place the fresh tissue in 5–10 volumes of RNAlaterÒ Solution on ice (see Note 1). 4. Store tissues in RNAlaterÒ at 4  C overnight in order to allow the solution to thoroughly penetrate the tissue. 5. Tissue samples can now be used immediately for RNA isolation or can be stored up to 1 month at 4  C without significant RNA degradation. 6. If longer term storage of tissues is desired, place samples at  20 or 80  C indefinitely.

2.3. Tissue homogenization and RNA purification from adult mouse tissues Several methods for tissue homogenization may be employed depending on the number of samples to be processed. We have found the following methods work well for low-, medium-, and high-throughput processing of tissue samples. 2.3.1. Low-throughput tissue homogenization and isolation of RNA 1. Remove  100 mg of tissue from RNAlaterÒ and place into 2 ml of TRIzolÒ (Invitrogen) in 14 ml polypropylene tubes. 2. Clean the probe of a Polytron electric homogenizer (Kinematica) by submerging for 10–15 s in 0.1 M NaOH, then DEPC-treated water, and finally 95% ethanol. Allow homogenizer probe to air dry. 3. Homogenize the tissue sample in the 14-ml tube on ice for 15 s. 4. Transfer the homogenized tissue slurry to a 2-ml microtube tube and centrifuge samples for 10 min at 4  C at 12,000g to remove insoluble material. 5. Following centrifugation, remove 1 ml of the TRIzolÒ supernatant to a fresh microtube (see Note 2).

In Vivo Inducible RNAi Analysis

421

6. Add 200 l of chloroform and shake vigorously for 15 s. 7. Incubate at room temperature for 2–3 min. 8. Centrifuge at 12,000–16,000g for 15 min to separate the aqueous and organic phases. 9. Remove the aqueous (upper) phase to a fresh tube and add an equal volume of 70% ethanol to the aqueous layer, mix well, and apply the sample to an RNeasy Mini spin column (Qiagen) placed in a 2-ml collection tube. 10. To complete the total RNA isolation, follow the RNeasy Mini protocol for the purification of total RNA from animal cells according to the manufacturer’s directions (Qiagen) (see Notes 3 and 4). 2.3.2. Medium-throughput homogenization and isolation of RNA For larger scale tissue processing the FastPrepÒ-24 (MP Biomedicals) system is an attractive alternative to single tissue processing by homogenization. This system uses special matrix lyzing beads to quickly disrupt up to 24 tissues samples in parallel. 1. Set up 2 ml FastRNA Pro Green tubes (MP Biomedicals) with 800 l QIAzol (Qiagen) and in general 20–100 mg of the tissue to be homogenized. 2. Press ‘‘Set’’ on the FastPrep-24 to the desired speed and duration of the homogenization choosing either between a factory default setting, or creating a custom setting. 3. Press ‘‘Run’’ to initiate the homogenization program. 4. Following completion of the program, allow the samples to sit at room temperature for 5 min. 5. Add 160 l of chloroform and shake the tubes vigorously for 15 s. 6. Allow homogenized samples to sit at room temperature for an additional 2–3 min. 7. Proceed to Step 8 as described in Section 2.3.1. 2.3.3. High-throughput homogenization and isolation of RNA For high-throughput homogenizing applications, we have found the TissueLyser II (Qiagen) to be an excellent platform that allows up to 192 samples to be homogenized simultaneously. 1. Place a 5-mm stainless steel bead into each tube of 96-well collection microtube plate (Qiagen) and keep on ice. 2. Place 10–30 mg of tissue in each well of the 96-well collection microtube plate. 3. Add 750 l QIAzol (Qiagen) into collection microtubes and cap wells. 4. Assemble the collection microtube plate into the TissueLyser adaptor sets and homogenize samples for 2–5 min at 25 Hz.

422

Christopher S. Raymond et al.

5. Remove and reverse the adaptor sets, and homogenize an additional 2–5 min at 25 Hz. 6. Let the plate stand at room temperature for 5 min. 7. Briefly, centrifuge the collection microplate at 3000g for 1 min at 4  C to remove the solution from the caps. 8. Add 150 l of chloroform and mix vigorously for 15 s. 9. Allow samples to incubate at room temperature for 2–3 min. 10. Proceed to Step 8 as described in Section 2.3.1.

2.4. cDNA generation from mouse tissues 1. Measure and determine RNA concentration from the OD260 reading in a UV spectrophotometer. The OD260/280 ratio should have a value >1.8 for highly purified RNA. To ensure the purified total RNA is devoid of degradation, RNA quality is also tested using a 2100 Bioanalyzer (Agilent Technologies) with the RNA 6000 Nano Chip kit (Agilent Technologies) or using the QIAxcel System (Qiagen) according to the manufacturer’s protocol. 2. Add equivalent amounts of RNA from each sample (usually 1 g RNA per reaction) to 1-well of a 96-well PCR-compatible microplate. 3. Bring total volume of each well up to 25 l with RNase-, DNase-free water. 4. Prepare a 2 RT Master Mix (5 l 10 reverse transcription buffer, 2 l 25 dNTPs, 5 l 10 random primers, 2.5 l MultiScribeTM Reverse Transcriptase, 50 U/ml, 11.5 nuclease-free water per each cDNA reaction) according to manufacturer’s protocol (High-Capacity cDNA Archive kit, Applied Biosystems). 5. Add 25 l of the 2 RT Master Mix to each well containing the RNA, bringing the total volume to 50 l. 6. Perform cDNA generation by loading the plate into a thermal cycler and running with cycling conditions of 25  C for 10 min followed by 37  C for 120 min.

2.5. Real-time PCR expression analysis of RNA knockdown in mouse tissues 1. Following the RT step, combine 2 l of RT reaction with 0.5 l of a 20 TaqMan Gene Expression Assay (which consists of a forward primer, reverse primer, and probe) directed to the gene that the shRNA is directed to 5 l of TaqMan Universal PCR Master Mix (Applied Biosystems), and 2.5 l of nuclease-free water for a 10-l final volume per reaction. Each TaqMan Assay is run in triplicate. In addition to setting up TaqMan reactions to the gene that the shRNA

In Vivo Inducible RNAi Analysis

423

is directed to, it is also necessary to set up separate reactions using a TaqMan endogenous control such as ACTB (43523341E, Applied Biosystems) or GAPD (4352339E, Applied Biosystems) assay reagents. 2. Place an ABIPRISMÒ Optical Adhesive cover (Applied Biosystems) on the microplate and centrifuge briefly to remove any air bubbles. 3. Perform real-time PCR using an Applied Biosystems 7900HT Fast Real-Time PCR System in 9600 emulation Run mode with cycling conditions of 95  C for 10 min (followed by 95  C for 15 s and 60  C for 60 s) for a total of 40 cycles. 4. Use the DDCt study method to calculate RNA knockdown (an example of such results can be seen in Fig. 21.2A). Further information and the mathematics that underlay this approach can be found in Applied Biosystems 7900HT Fast Real-Time PCR System Relative Quantitation Using Comparative CT Getting Started Guide (Applied Biosystems, PN 4364016). The resultant decrease in protein expression from RNA knockdown can also be examined by Western blot analysis if a suitable antibody is available (Fig. 21.2B).

2.6. Determining induced shRNA expression in mouse tissues To test whether the observed RNA knockdown in tissues correlates with shRNA expression levels, either of the two following procedures can be performed. 2.6.1. Northern blot protocol 1. Run 5 g of RNA from each adult tissue sample with an equal volume of 2 TBE-Urea sample buffer on a 15% Novex TBE-Urea gel (Invitrogen). Also include 5 l RNA ladder plus (Invitrogen) plus 5 l 2 RNA sample buffer for each lane of marker. Heat samples and RNA ladder at 70  C for 2 min before loading on the gel. Run gel at 180 V for 1 h in 1 TBE running buffer until the dye front reaches approximately 2/3 down the gel. 2. Stain the gel with 0.5 g/ml ethidium bromide for 5 min and place on a UV gel box and record image. 3. Transfer RNA onto Hybond-XL membrane (Amersham) using an electroblot apparatus (Bio-Rad) in 0.5 TBE transfer buffer at 80 V for 1 h. 4. UV-cross-link RNA onto the membrane with the optimal program. At this point the membrane can be stored in plastic wrap and kept at 4  C until ready for hybridization. 5. Design a custom 21-mer single-stranded DNA oligo probe to the shRNA sequence according to the following format: 50 -AA-19mer sense target shRNA sequence-30 .

424

Christopher S. Raymond et al.

6. Label the shRNA DNA oligo probe as follows (mirVana Probe&Marker kit, Ambion): 1 l DNA oligo probe (30 pmol), 5 l 10 PNK buffer, 5 l T4-PNK, 5 l gamma-ATP (Amersham), 34 l dH2O. Incubate at 37  C for 1 h. 7. Prespin a G25 Sephadex column (GE Healthcare Life Sciences) for 1 min at 736g. Purify DNA oligo probe from unincorporated radioisotope by adding 100 l of dH2O to probe labeling reaction and applying the sample to the G25 column. Centrifuge for 2 min at 736g. Collect and save flow-through fraction and determine radioactivity level of probe using a liquid scintillation counter. 8. Prehybridize membrane with 5 ml of Rapid-Hyb hybridization buffer (GE Healthcare Life Sciences) at 42  C for 1 h. 9. Heat the DNA oligo probe at 95  C for 5 min before adding 5  105 cpm per each milliliter of hybridization buffer. Hybridize the membrane at 50  C for 2–3 h. 10. Wash with 5 SSC, 0.1% SDS at room temperature for 5 min. Repeat the wash step two additional times. 11. Wash with 1 SSC, 0.1% SDS at 50  C for 15 min. 12. Expose to X-ray film in cassettes with intensifier screens at 80  C. 13. After obtaining a suitable exposure, strip the membrane by boiling in 0.1% SDS for 10 min. 14. Label and reprobe the membrane as outlined above with a 30-mer U6 shRNA oligo probe (50 -GCAGGGGCCATGCTAATCTTCTCTGTATCG-30 ) as an internal control, and expose to X-ray film. An example of the Northern blot results can be seen in Fig. 21.3A.

2.6.2. Real-time PCR protocol for shRNA expression 1. Design small RNA assay to the antisense shRNA sequence using Applied Biosystems Custom TaqMan Small RNA Assay Design tool. 2. Homogenize tissues in 800 l QIAzol, add 160 l of chloroform, vigorously shake for 15 s, centrifuge for 15 min for 12,000g at 4  C. 3. Transfer the supernatant into 2 ml tubes and prepare total RNA including small RNA with miRNeasy Mini kit (Qiagen), and measure RNA concentration. 4. The reverse transcription reaction is carried out by diluting RNA samples to 2 ng/l, and using a total of 10 ng RNA per each 15 l RT reaction. Each sequence requires two RT reactions, one for the target gene (custom designed small RNA assay) and one for an endogenous control such as snoRNA202 (ABI). 5. Prepare 7 l RT Master Mix per reaction (0.15 l 100 mM dNTP with dTTP, 1 l MultiScribe Reverse Transcriptase 50 U/l, 1.5 l 10 RT buffer, 0.19 l RNase inhibitor 20 U/l, 4.16 l nuclease-free water)

425

In Vivo Inducible RNAi Analysis

A

Kidney

Liver shRNA

−

+

+

−

+

+

Dox

−

−

+

−

−

+

U6 control

Relative shRNA expression

B 100 80 60 40 20 0 K

K

K

id

id

ne

ne

y+

y+

sh

sh

sh

R

R

R

N

N

A

A

+

−

−

D

D

D

ox

ox

ox

ox

ox

ox

D

D

D

A

+

−

−

N

A

A

A

N

N

N

R

R

R

sh

sh

sh

y−

r+

r+

r−

ne

ve

ve

ve

id

Li

Li

Li

Figure 21.3 (A) Example of Northern blot displaying doxycycline-induced shRNA expression in adult liver and kidney tissues. The blot is probed with a U6 probe as an endogenous control. (B) Example of data from the real-time PCR protocol for shRNA expression displaying doxycycline-induced shRNA expression in the corresponding adult liver and kidney tissues from (A).

6. 7. 8. 9.

(TaqMan MicroRNA Reverse Transcription kit, Applied Biosystems). Mix gently, place on ice until RT reaction. Add 5 ml total RNA and 3 ml RT primer to each RT reaction, and mix well by pipetting. Incubate at 16  C for 30 min, 42  C for 30 min, and 85  C for 5 min in a thermal cycler. Dilute cDNA reaction with 30 l of RNase-, DNase-free water, and mix well. Following the RT step, combine 2 l of diluted RT reaction with 0.5 l of the Custom 20 TaqMan Small RNA Assay (20; forward primer, reverse primer, and probe), 5 l of TaqMan Universal PCR Master Mix No AmpErase UNG (Applied Biosystems), and 2.5 l of nuclease-free water for a 10-l final volume. Each TaqMan Assay is run in triplicate.

426

Christopher S. Raymond et al.

10. Perform real-time PCR using an Applied Biosystems 7900HT Fast Real-Time PCR System in 9600 emulation Run mode with cycling conditions of 95  C for 10 min (followed by 95  C for 15 s and 60  C for 60 s) for a total of 40 cycles. An example of the resulting data from the Real-Time PCR Protocol for shRNA expression can be seen in Fig. 21.3A.

3. Notes 1. Only use RNAlaterÒ Solution with fresh tissue only; do not freeze tissues before or immediately after immersion in the RNAlaterÒ Solution. 2. The remaining 1 ml of TRIzolÒ supernatant can be transferred to a separate fresh microtube and stored at 80  C, and if needed, can be processed at a later date. 3. Always do the optional DNase step according to the manufacturer’s directions. 4. For processing a large number of RNA isolation needs, the RNeasy protocol can be scaled-up for medium- and high-throughput applications using automated platforms such as the QiaCube and QiaRobot.

REFERENCES Cawthorne, C., Swindell, R., Stratford, I. J., Dive, C., and Welman, A. (2007). Comparison of doxycycline delivery methods for Tet-inducible gene expression in a subcutaneous xenograft model. J. Biomol. Tech. 18, 120–123. Coumoul, X., Shukla, V., Li, C., Wang, R. H., and Deng, C. X. (2005). Conditional knockdown of Fgfr2 in mice using Cre-LoxP induced RNA interference. Nucleic Acids Res. 33, e102. Elbashir, S. M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K., and Tuschl, T. (2001). Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494–498. Gossen, M., and Bujard, H. (1992). Tight control of gene expression in mammalian cells by tetracycline-responsive promoters. Proc. Natl. Acad. Sci. USA 89, 5547–5551. Gruber, M., Hu, C. J., Johnson, R. S., Brown, E. J., Keith, B., and Simon, M. C. (2007). Acute postnatal ablation of Hif-2alpha results in anemia. Proc. Natl. Acad. Sci. USA 104, 2301–2306. Hannon, G. J. (2002). RNA interference. Nature 418, 244–251. Kunath, T., Gish, G., Lickert, H., Jones, N., Pawson, T., and Rossant, J. (2003). Transgenic RNA interference in ES cell-derived embryos recapitulates a genetic null phenotype. Nat. Biotechnol. 21, 559–561. Lickert, H., Takeuchi, J. K., Von Both, I., Walls, J. R., McAuliffe, F., Adamson, S. L., Henkelman, R. M., Wrana, J. L., Rossant, J., and Bruneau, B. G. (2004). Baf60c is

In Vivo Inducible RNAi Analysis

427

essential for function of BAF chromatin remodelling complexes in heart development. Nature 432, 107–112. Minamishima, Y. A., Moslehi, J., Bardeesy, N., Cullen, D., Bronson, R. T., and Kaelin, W. G., Jr. (2008). Somatic inactivation of the PHD2 prolyl hydroxylase causes polycythemia and congestive heart failure. Blood 111, 3236–3244. Oberdoerffer, P., Kanellopoulou, C., Heissmeyer, V., Paeper, C., Borowski, C., Aifantis, I., Rao, A., and Rajewsky, K. (2005). Efficiency of RNA interference in the mouse hematopoietic system varies between cell types and developmental stages. Mol. Cell. Biol. 25, 3896–3905. Paddison, P. J., Caudy, A. A., Bernstein, E., Hannon, G. J., and Conklin, D. S. (2002). Short hairpin RNAs (shRNAs) induce sequence-specific silencing in mammalian cells. Genes Dev. 16, 948–958. Poueymirou, W. T., Auerbach, W., Frendewey, D., Hickey, J. F., Escaravage, J. M., Esau, L., Dore, A. T., Stevens, S., Adams, N. C., Dominguez, M. G., Gale, N. W., Yancopoulos, G. D., et al. (2007). F0 generation mice fully derived from gene-targeted embryonic stem cells allowing immediate phenotypic analyses. Nat. Biotechnol. 25, 91–99. Reboredo, M., Kramer, M. G., Smerdou, C., Prieto, J., and De Las Rivas, J. (2008). Transcriptomic effects of Tet-on and mifepristone-inducible systems in mouse liver. Hum. Gene Ther. 19, 1233–1247. Seibler, J., and Schwenk, F. (2010). Transgenic RNAi applications in mice. Methods Enzymol. 477, 367–386. Seibler, J., Kuter-Luks, B., Kern, H., Streu, S., Plum, L., Mauer, J., Kuhn, R., Bruning, J. C., and Schwenk, F. (2005). Single copy shRNA configuration for ubiquitous gene knockdown in mice. Nucleic Acids Res. 33, e67. Seibler, J., Kleinridders, A., Kuter-Luks, B., Niehaves, S., Bruning, J. C., and Schwenk, F. (2007). Reversible gene knockdown in mice using a tight, inducible shRNA expression system. Nucleic Acids Res. 35, e54. Siolas, D., Lerner, C., Burchard, J., Ge, W., Linsley, P. S., Paddison, P. J., Hannon, G. J., and Cleary, M. A. (2005). Synthetic shRNAs as potent RNAi triggers. Nat. Biotechnol. 23, 227–231. Soriano, P. (1999). Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat. Genet. 21, 70–71. Ste Marie, L., Luquet, S., Cole, T. B., and Palmiter, R. D. (2005). Modulation of neuropeptide Y expression in adult mice does not affect feeding. Proc. Natl. Acad. Sci. USA 102, 18632–18637. Takeda, K., Cowan, A., and Fong, G. H. (2007). Essential role for prolyl hydroxylase domain protein 2 in oxygen homeostasis of the adult vascular system. Circulation 116, 774–781. van de Wetering, M., Oving, I., Muncan, V., Pon Fong, M. T., Brantjes, H., van Leenen, D., Holstege, F. C., Brummelkamp, T. R., Agami, R., and Clevers, H. (2003). Specific inhibition of gene expression using a stably integrated, inducible small-interfering-RNA vector. EMBO Rep. 4, 609–615.

C H A P T E R

T W E N T Y- T W O

The Power of Reversibility: Regulating Gene Activities via Tetracycline-Controlled Transcription ¨nig,* Hermann Bujard,† and Manfred Gossen‡,§ Kai Scho Contents 1. Introduction 2. The Tet Regulatory Systems 2.1. Tet-Off 2.2. Tet-On 3. Tet-Transgenic Mice Available from Repositories 4. Tet-Controlled Transgenes to Study Learning and Memory 5. Tet-Transgenic Animals in Cancer Research 6. Secondary iPSC Technology 7. Transgenic Rats 8. Future Perspectives 8.1. Conditional knockdown via RNAi 8.2. Integrating the binary Tet-control system into genomes References

430 431 432 432 434 439 439 443 444 444 445 446 447

Abstract Tetracycline-controlled transcriptional activation systems are widely used to control gene expression in transgenic animals in a truly conditional manner. By this we refer to the capability of these expression systems to control gene activities not only in a tissue specific and temporal defined but also reversible manner. This versatility has made the Tet regulatory systems to a preeminent tool in reverse mouse genetics. The development of the technology in the past 15 years will be reviewed and guidelines will be given for its implementation in

* Zentralinstitut fu¨r Seelische Gesundheit, Mannheim, Germany Zentrum fu¨r Molekulare Biologie Heidelberg (ZMBH), Im Neuenheimer Feld, Heidelberg, Germany Berlin-Brandenburg Center for Regenerative Therapies (BCRT), Berlin, Germany } Max Delbru¨ck Center for Molecular Medicine (MDC), Berlin, Germany { {

Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77022-1

#

2010 Elsevier Inc. All rights reserved.

429

430

¨nig et al. Kai Scho

creating transgenic rodents. Finally, we highlight some recent exciting applications of the Tet technology as well as its foreseeable combination with other emerging technologies in mouse transgenesis.

1. Introduction Classical genetics laid the foundation of understanding gene function. Traditionally, the mutations analyzed were either spontaneous or introduced by what we would call today ‘‘genotoxic stress.’’ The careful analysis of resulting phenotypes, their patterns of inheritance, and conclusions from combining different genotypes by breeding or other ‘‘conventional’’ methods of reshuffling genomes allowed deep insights into many biological processes, even without the possibility to determine the molecular nature of the mutation. Most mutants were either plain loss-of-function or gainof-function, resulting from point mutations, deletions, inversions, or translocations. However, some of these mutations resulted in hypomorphs, that is, alleles characterized by a partial loss-of-function. Some of them were conditional in the sense that the expressivity of the mutant allele could be controlled by external cues. Well-known examples are ts alleles, in which a mutation renders the functionality of a corresponding protein temperature sensitive. The possibility to control the phenotype of a given genotype at will has greatly contributed to the understanding of physiological and developmental processes, in particular, in microbial and in Drosophila genetics. Mouse genetics could not provide a comparable depth of insight in gene function until the advent of genetic engineering. However, the possibility to generate transgenic mice by random integration of transgenes, and thereafter the development of embryonic stem cell (ES cell) technology with its potential to introduce targeted mutations in the genome, made the mouse amenable for genetic approaches at the molecular level, and thus to a most relevant mammalian model system. In many model systems of ‘‘classical’’ genetics, the possibility to create conditional mutants has greatly broadened the spectrum of questions that can be asked, particularly when introducing potentially detrimental mutations. In the mouse, conditionality can be achieved by spatial and temporal control over recombinases (Anastassiadis et al., 2010), generally resulting in an irreversible genetic alteration. Given the proper targeting vectors, this approach can ensure a complete null genotype. By contrast, true conditionality, that is, reversible alteration of a phenotype, can be achieved by expression systems controlling gene activity on the transcriptional level. These are mostly heterologous, binary expression systems, consisting of an engineered transcription factor and a corresponding response unit encompassing the transgene of interest (Gossen et al., 1993). Several such systems where gene activities may be

The Power of Reversibility

431

controlled at the transcriptional level have been developed, which function in cultured cells of various origins. However, only few of them are widely applied. The diversity of transcriptionally controlled expression systems with established functionality in mice is even more limited and almost exclusively restricted to the Tet-On and the Tet-Off system.

2. The Tet Regulatory Systems The key elements of the Tet regulatory systems, the Tet repressor (TetR) and the tet operator (tetO), are derived from the Tn10 tetracycline resistance operon, evolutionarily adapted to the prokaryote Escherichia coli. They turned out to be ideally suited for functioning in mammalian and other eukaryotic cells in a highly selective mode. Most importantly, the effector molecules that interact with TetR, the tetracyclines, permeate tissues with relative ease, are well tolerated and their pharmacology has been thoroughly analyzed due to their medical use in humans. Moreover, the interaction of the TetR with tetO is of exquisite specificity. This is especially important, when the protein/DNA interaction has to occur in the context of a mammalian genome with its vast excess of unspecific binding sites. For more information on these topics, we refer to reviews presenting the relevant issues in more detail (Berens and Hillen, 2003; Gossen et al., 1993). Based on the tetracycline-sensitive interaction between tetO and the DNA binding domain of TetR, this prokaryote-derived regulatory principle has been adapted for the use in eukaryotes. Shortly, TetR has been fused to a eukaryotic transcription activation domain, converting the E. coli TetR into a Tet transactivator, a highly specific transcription factor for mammalian cells, which would only interact with a properly reengineered eukaryotic promoter. To this end, eukaryotic minimal promoter sequences were generated which are inactive in the absence of bound Tet transactivator, but retain the capacity to unambiguously define a transcription start site upon binding of these activators in their proximity. For the Tet regulatory systems, this is guaranteed by fusing multimerized tetO sequences upstream to the minimal promoter. This operator array and the adjacent minimal promoter constitute a functional unit and will be referred to as Tet-responsive promoter, or Ptet. In the presence of Tet transactivators, Ptet activity can be controlled by tetracycline. Again, for a comprehensive description of this technology, we refer to the review literature (Gossen and Bujard, 2002; Scho¨nig and Bujard, 2003). Only the distinction between the two basic versions of the Tet regulatory systems as shown in Fig. 22.1, Tet-Off and Tet-On, will be introduced here in some detail, as this is particularly important when applications in transgenic animals are considered.

432

¨nig et al. Kai Scho

A

B

Ptsp tTA

−Dox

GOI (A)n−

−Dox

GOI Ptet

(A)n−

rtTA-M2

+Dox

Ptet

+Dox

GOI (A)n−

(A)n−

Ptet

GOI (A)n− Ptet

100 Gene activation (%)

100 Gene activation (%)

Ptsp

(A)n−

50

0

50

0 0 0.1

1

10

Doxycycline (ng/ml)

0

1

10 100 Doxycycline (ng/ml)

Figure 22.1 Schematic presentation of the Tet regulatory system. (A) The principles of the Tet-Off system, with a tissue-specific promoter (Ptsp) driving expression of tTA. The effects of Dox administration on binding of tTA to the corresponding tetracyclineresponsive promoter (Ptet) and on the expression of the gene of interest (GOI) are indicated. The lower panel shows an approximation of the dose response of the Tet-Off system on target gene expression. (B) The principles of the Tet-On system in an analogous manner, specified for the rtTA-M2 transactivator.

2.1. Tet-Off This is the originally introduced Tet system (Gossen and Bujard, 1992). The TetR as found in E. coli was fused to a transcription activation domain of a viral transcription factor active in mammals (derived from herpes simplex virus viral protein 16). The resulting protein tTA (tetracycline-controlled transactivator) resembles a typical eukaryotic transcription factor, also in its domain modularity. It retained the binding specificity to tetO sequences, which can be abrogated by tetracycline or some of its derivatives such as doxycycline (Dox). The presence of these substances prevents the binding of tTA to tetO, thereby inactivating the expression system. This type of response led ultimately to the coining of the term ‘‘Tet-Off.’’

2.2. Tet-On The Tet-On system utilizes the same Ptet, but makes use of an altered transcription factor containing a few amino acid exchanges in the TetR moiety (Gossen et al., 1995). These reverse its DNA binding properties in

The Power of Reversibility

433

response to Dox, therefore its name reverse tetracycline-controlled transactivator or rtTA. Thus, it binds to tetO sequences only in the presence of Dox. In analogy to Tet-Off, the regulation system is mostly referred to as ‘‘Tet-On’’ system. Both the Tet-On and the Tet-Off system are routinely applied in a wide variety of transgenic mice. In the majority of applications this is achieved by breeding of a responder line containing the gene of interest under control of a Ptet with a transactivator line expressing tTA or rtTA. The transactivators are mostly under the control of tissue-specific promoters, depending on the scientific question asked. The choice between tTA and rtTA is often—and for good reasons— made according to the experimental scheme employed. Whenever the focus of a transgenic study relates to the phenotypic consequences of turning off a transgene under Ptet control, ideally a Tet-Off type transactivator line will be used. Here, the transgene will be active during embryonic development and adulthood until the inactivation of the transgene expression appears appropriate. This is achieved by feeding (or injecting) the double transgenic animals with Dox. The uptake of Dox is efficient and can be readily controlled (Kistner et al., 1996). If the same type of response is to be achieved in a Tet-On type transactivator line, this would not only involve the feeding of mothers with Dox during pregnancy and during lactation, but also the removal of Dox from the animal. In contrast to the situation in tissue cultures, this process cannot be influenced experimentally and is comparatively slow as it is governed by the biological and chemical half-life time of the effector Dox. Conversely, if the focus of a transgenic study relates primarily to the phenotypic consequences of turning on a transgene under Ptet control, ideally a Tet-On approach is indicated. However, the respective rtTA lines require higher concentrations of Dox for induction than what is needed for the shutdown of gene expression via tTA. This has not been a problem for rtTA-controlled expression in tissues other than the brain, where, due to the blood–brain barrier, the Dox concentrations required for full rtTA activation are more difficult to attain. Accordingly, for studying brain functions the Tet-On system has been much less frequently used. The problem has been partly overcome by the introduction of rtTA-M2 (Michalon et al., 2005; Urlinger et al., 2000), an rtTA allele which responds to 10-fold lower concentrations of Dox. (Note: rtTA-M2 is often referred to as Tet-On-Advanced.) The recurrent problems associated with brainspecific expression of transgenes under Tet-On control may be further ameliorated by recently identified new Tet-On type transactivator alleles (Das et al., 2004; Zhou et al., 2006). Some of these transactivators respond to even lower concentrations of Dox than rtTA-M2. In tissue culture the novel transactivators are fully induced at around 30 ng/ml of Dox as compared to 100 and 1000 ng/ml for rtTA-M2 and rtTA, respectively.

434

¨nig et al. Kai Scho

Some reports described difficulties in inducing expression from Ptet promoters in certain cells of the brain, when the system was kept in the ‘‘OFF’’ state during embryonic development (Bejar et al., 2002; Krestel et al., 2004; Zhu et al., 2007). The phenomenon appeared to be independent of Dox disposability and might be a specific feature of some subtypes of neurons (Zhu et al., 2007). In the following, we will give a short survey of available Tet-transgenic mouse lines, and of examples for exciting new applications of the technology, also in the context of newly emerging concepts for genetic modifications in rodents. The selection of topics is subjective and far from comprehensive. But, it should convey an idea about the broad applicability of the technology, which hopefully will also stimulate and encourage researchers to further explore novel applications.

3. Tet-Transgenic Mice Available from Repositories Establishing a conditional transgenic mouse model involves a considerable experimental effort, which is both expensive and time consuming. The general strategy for Tet-transgenic mice relies on breeding of a Tet transactivator line with a responder line, both of which are established independently. Transactivator lines harbor a tTA or rtTA gene under the control of mostly tissue or cell-type-specific promoters. Proper transactivator expression is verified by either RNA analysis or upregulation of gene expression when crossed to a Tet-responsive indicator line. For a first identification of transgenic lines harboring a functional Tet-responsive gene of interest, primary ear fibroblast cultures are a valuable aid, as these cells can be transfected with tTA/rtTA-expressing constructs and gene induction can be monitored in dependence of Dox (Scho¨nig and Bujard, 2003). The safest but also most time consuming and costly alternative is the direct cross of the founders with functionally characterized transactivator lines, which will allow more thorough analyses like tightness of control, position effect variegation, etc. The latter approach is facilitated by the use of bidirectional Tet promoters, where reporter genes like lacZ, green fluorescent protein, or luciferase are coexpressed with the gene of interest (Baron et al., 1995). As with any other research tool, once published any wellcharacterized transgenic line should be available to other researchers. In reality, however, sharing of these valuable research tools is often hampered by problems like shipping, unclear hygiene status, etc. Moreover, for many laboratories maintaining colonies or even collections of frozen embryos is

435

The Power of Reversibility

costly and requires proper administration. As a result, many valuable lines are actually lost to the scientific community. The major problems could be overcome by transferring respective lines to repositories from where they can be distributed. For Tet-controlled gene expression in the mouse, collections of well-characterized transactivator lines are particularly valuable, as they can be reutilized for breeding with novel Tet-responsive lines. Along the lines of this concept of a ‘‘Tet Zoo’’ (Scho¨nig and Bujard, 2003), a single transactivator line has the capacity to address numerous questions, which are related only through the cell type, in which the effect of conditional expression of a gene of interest is studied. In Table 22.1, we list the repositories we are aware of, which distribute Tet-transgenic mice. For the available Tet regulator lines, we distinguish between tTA and rtTA lines and also include Tet transsilencer lines (see below). For the Tet-responsive mouse lines, we would like to particularly point out available reporter lines, which are most valuable for setting up the system in a defined manner. In Table 22.2, we specifically indicate a selection of transactivator lines showing well-characterized tissue-specific expression patterns. When these match the requirements for new questions to be addressed, such animals can readily be brought to use again.

Table 22.1 Repositories distributing Tet-transgenic mice Number of transregulator lines

Repository

JAX EMMA RIKEN MMRRC

Link

www.jax.org www.emmanet.org/ mutant_types.php www.brc.riken.go. jp/lab/animal/en www.mmrrc.org

tTA

18 (2) 6

rtTA a

tTS

Number of responder lines

Total

With reporter gene

19 (1) 4 1

31 (7) 6 11 3

5

2

20

13

4

4

10

1

We are aware of four repositories from which Tet-transgenic mice are available. A distinction is made between transregulator and responder lines. The number of different mouse lines available is indicated (current as of May 2010). Several other organizations maintain databases where some of these animals are listed: Mugen: www.bioit.fleming.gr/mugen/mde.jsp IMSR (International Mouse Strain Resource): www.findmice.org CMMR (Canadian Mouse Mutant Repository): www.cmmr.ca Other helpful webtools for identifying Tet-transgenic mice outside of repositories are TET Systems: www.tetsystems.com/support/transgenic-mouse-lines University of Mainz: www.zmg.uni-mainz.de/tetmouse/index.htm a Numbers in parentheses indicate mouse lines under development.

Table 22.2 Mouse lines expressing tTA or rtTA genes (selection from table on TET Systems’ web site) Promoter (origin)

Axin2

Bovine keratin 5 (K5) VE-cadherin 5 CaMKIIa CaMKIIa CD34 Clara cell 10 kDa protein (CC10) Clara cell secretory protein (CCSP) Intronic IgH enhancer, minimal promoter Immunoglobulin heavy chain enhancer and SR a promoter Insulin Insulin Keratin 5 (K5) Keratin 6 (K6) Keratin 14 (K14)

Tissue specificity

tTA

Reference 2

Developing kidney (Wolfian duct and ureteric bud epithelia) Epithelial cells Endothelial cells Brain Brain Early bone marrow progenitor and stem cells Lung airway (parenchyma)

rtTA

þ þ

Respiratory epithelial cells

rtTA S-M2

Shakya et al. (2005), Yu et al. (2007)

þ

Vitale-Cross et al. (2004) Ye et al. (2008) Mayford et al. (1996) Mansuy et al. (1998a) Radomska et al. (2002)

þ þ þ

Mehrad et al. (2002)

þ

Tichelaar et al. (2000)

Thymus, bone marrow

þ

Hess et al. (2001)

Hematopoietic system

þ

Felsher and Bishop (1999)

Pancreatic b-cells Pancreatic b cells Epidermis, hair follicle Keratinocytes Epidermis, squamous epithelia

þ þ þ

þ þ þ

Efrat et al. (1995) Thomas et al. (2001) Diamond et al. (2000) Guo et al. (1999) Xie et al. (1999)

Keratin 14 (K14) Keratin 18 (K18) Krt12-knock-in (mouse) Lck Liver-enriched activator protein (LAP) Liver-enriched activator protein (LAP) a-Myosin heavy chain (MHC a) a-Myosin heavy chain (MHC a) MMTV Neuron-specific enolase (NSE) P0 P2 ROSA26

Stem cell leukemia (SCL) gene 30 enhancer SM22 (mouse) SM22 a SM22 a Surfactant protein-C (SP/C)

Mammary gland Trachea, upper bronchi, submucosal glands Corneal epithelium T cell lineage Liver

þ

þ þ

Liver Heart muscle

Schwann cells Olfactory sensory neurons Various tissues depending on CRE expression Hematopoietic cells Smooth muscle cells Smooth muscle cells Smooth muscle cells Respiratory epithelial cells

Chikama et al. (2005) Leenders et al. (2000) Kistner et al. (1996)

þ

Scho¨nig et al. (2002)

þ þ

þ þ

Dunbar et al. (2001) Ye et al. (2001)

IRES-rtTA

þ

Heart muscle Mammary gland Brain

þ

þ þ þ loxP-STOPloxP rtTA2S-M2 þ þ þ

Passman and Fishman (1994) Valencik and McDonald (2001) Gunther et al. (2002) Chen et al. (1998) Pot et al. (2002) Gogos et al. (2000) Yu et al. (2005)

Koschmieder et al. (2005) West et al. (2004) Ju et al. (2001) Proctor et al. (2008) Tichelaar et al. (2000) (continued)

Table 22.2 (continued) Promoter (origin)

Tissue specificity

tTA

rtTA

Reference 2

Thy1 Thyroglobulin promoter Tie

Retinal ganglion cells Thyroid gland Endothelial cells

Tie2

Vascular endothelium

þ

Tyrosinase Tyrosinase LCR Vav Villin

Melanocytes Retina Hemotopoietic system Intestinal epithelium

þ þ

þ

þ

rtTA S-M2 þ

rtTA2S-M2

Kerrison et al. (2005) Knostman et al. (2007) Sarao and Dumont (1998) Deutsch et al. (2008), Doring et al. (2007), Teng et al. (2002) Chin et al. (1999) Gimenez et al. (2004) Wiesner et al. (2005) Roth et al. (2009)

Tissue specificity of Tet transactivator expressing mice. Shown is an extensive but incomplete overview about tTA and rtTA transgenic activator mouse lines. The target tissue for expression and the primary reference for each line are provided.

The Power of Reversibility

439

4. Tet-Controlled Transgenes to Study Learning and Memory In this chapter we would like to give an intriguing example for the success of such ‘‘recycling strategies’’ of a particular mouse line. Learning and memory formation involves multiple phases regulated by distinct gene activities (Dudai, 2002). In this context, it is of considerable interest to find out whether a gene plays a direct role within a specific phase of memory establishment or whether it has a more indirect effect on, for example, the development of the neuronal circuits involved. By placing a respective gene under Tet control, the effects of its induction can be monitored in response to its spatially and temporally defined expression. Such experiments could greatly contribute to answering questions as formulated above. Indeed, in 1996, the group of Kandel generated a transgenic mouse line expressing tTA under the control of the calcium-calmodulindependent kinase IIa (CamKIIa) promoter (Mayford et al., 1996) and demonstrated specific expression of the transactivator in defined regions of the brain, predominantly in the forebrain including hippocampus, cerebral cortex, and striatum. As exemplified first by the Kandel laboratory with the conditional expression of an activated calcium-independent form of CamKIIa, numerous researchers combined the ‘‘Mayford mouse’’ with Tet-responder mouse lines carrying their gene of interest under Ptet control. Seminal insights into the mechanisms of synaptic plasticity and learning and memory in mammals were thus gained in a relatively efficient and comparable way. This very CamK-tTA mouse line (available through The Jackson Laboratory/ JAXÒ Mice) is still being used and meanwhile one of, if not the single most ‘‘recycled’’ Tet transactivator mouse line. A rather arbitrary collection of publications arising from the use of this mouse line with a spectrum of Tet-responder lines is provided in Table 22.3. With over 600 citations of the original publication by Mayford et al. (1996), this table, while far from complete, illustrates the power of properly setup binary systems, which will eventually result in pools of mouse lines which significantly facilitate the generation novel double transgenic animals.

5. Tet-Transgenic Animals in Cancer Research Another scientific area where the Tet technology had and still has significant impact is cancer research. Numerous mouse lines with oncogenes or tumor suppressor genes have been established, taking advantage not only of the tissue specificity and the temporal control over gene expression intrinsic to the Tet regulatory system but also of its reversibility.

440

¨nig et al. Kai Scho

Table 22.3 Transgenic mouse lines used in crosses to CaMKIIa-tTA mice (Mayford et al., 1996) Gene of interest

Reference

Mouse model for

VEGF (vascular endothelial growth factor A) CK1d (casein kinase 1d) CRH (corticotropinreleasing hormone) AlstR (Drosophila allatostatin neuropeptide receptor) MYC

Licht et al. (2010)

Olfactory bulb adult neurogenesis

Zhou et al. (2010) Kolber et al. (2010)

Regulation of dopamine signaling Anxiety and depression

Wehr et al. (2009)

Central neural circuits

Lee et al. (2009)

CREB (cAMP response element binding protein) APPNLI, tauP301L TeTX (tetanus light chain) a-syn (human a-synuclein) Human mutant TauRD MeCP2 (methylCpG-binding protein 2) Ht31 (inhibitor of PKA anchoring) hTau40DK280 hTau40DK280/pp Homer 1a

Viosca et al. (2009)

Neurodegenerative diseases, for example, Alzheimer Fear memory

Paulson et al. (2008) Nakashiba et al. (2008)

Alzheimer disease Learning and memory

Nuber et al. (2008)

Alzheimer disease

Mocanu et al. (2008)

Tauopathies

Alvarez-Saavedra et al. (2007)

Rett syndrome

Nie et al. (2007)

Memory storage

Eckermann et al. (2007)

Tauopathies

Celikel et al. (2007)

Synaptic plasticity and spatial memory Mossy fiber synaptic plasticity Tauopathies Tauopathies Parkinson’s disease

Gai2 (a subunit of G protein) tauP301L tauP301L GDNF (rat Glial cell line-derived neurotrophic factor)

Nicholls et al. (2006) Ramsden et al. (2005) Santacruz et al. (2005) Kholodilov et al. (2004)

441

The Power of Reversibility

Table 22.3 (continued) Gene of interest

Reference

Mouse model for

hIkBa-AA

Fridmacher et al. (2003)

5-HT1AR1 (serotonin1A receptor) VP16-CREB (cAMP response element binding fusion protein) GluR-A (glutamate receptor A subunit) polyQ-htt NR-1 (NMDR subunit) DCaM-A1 (truncated form of calcineurin)

Gross et al. (2002)

Neurodegenerative disease Anxiety behavior

Barco et al. (2002)

Memory

Mack et al. (2001)

Synaptic plasticity

Yamamoto et al. (2000) Shimizu et al. (2000)

Huntington’s disease Memory

Mansuy et al. (1998b)

Memory

Recycling of the CaMKIIa-tTA mice (Mayford et al., 1996). A rather arbitrary excerpt from a literature survey about the use of one single Tet transactivator line demonstrates the utility of setting up conditional mouse models by independently creating transactivator and response lines.

The issue of ‘‘oncogene addiction’’ as analyzed by such conditional models is only mentioned here without giving further details—we rather refer to an excellent review on this topic by Felsher (2008). Here, we like to focus on recent work by Podsypanina, Varmus, and colleagues on the mechanisms of cancer dissemination via metastatic invasion of remote tissues (Podsypanina et al., 2008). It has been a long held view that this process is a late stage event in cancer progression, resulting from an accumulation of mutations ‘‘enabling’’ the conversion to full malignancy including the potential to colonize other organs. However, the careful analysis of mouse models, patient data as well as theoretical considerations about the sequence of events in metastasis has already cast some doubts on the general validity of this model (Klein, 2008; Weinberg, 2008). Podsypanina and colleagues established now two transgenic mouse models with oncogenes under Tet control. They demonstrated the capacity of untransformed mammary cells to invade, reside, and proliferate in nonmammary tissues. To this end, a mouse line expressing rtTA under the control of the mammary-specific MMTV promoter was bred with animals carrying the MYC and activated Kras oncogene under the control of Ptet. The resulting tri-transgenic mice were phenotypically normal unless exposed to Dox.

442

¨nig et al. Kai Scho

When induced, however, these animals rapidly developed tumors, proving the expected functionality of all the critical elements in this model. These animals could now be used to bypass the step that has been widely considered to be crucial for metastasis: the escape of transformed cells into the vasculature as one bottleneck for dissemination. Podsypanina and colleagues rather isolated untransformed cells from rtTA-MYC/KrasD12 transgenic mice never exposed to Dox and injected them directly into the bloodstream of recipient ‘‘wild-type’’ mice. When exposed to Dox—and only then—the animals developed ectopic tumors. The result that untransformed cells can colonize organs like the lung was further confirmed in a second conditional model, using the same rtTA line but in combination with a different oncogene (polyoma middle T). Again, after the cell transfer, tumors formed exclusively after Dox exposure. Monitoring oncogene activation at different times after cell transfer showed that untransformed cells can reside at ectopic sites for prolonged periods of time (see Fig. 22.2). During this time window, they could accumulate mutations resulting in their conversion to a fully cancerous state. These

10 d postinjection, no doxycycline 20000 15000

107

10000

106

24 d postinjection, 14 d on doxycycline 3

5

10

104

2

⫻106

Relative luminescence units

108

1

1

2

6

8

9 10 17 18

Weeks post-injection

Figure 22.2 Tet-control over ectopic tumor formation. Mammary cells isolated from double transgenic mice harboring a Ptet–polyoma middle T oncogene construct controlled by an MMTV-driven rtTA gene form ectopic tumors after transfer into transgene negative recipient mice. Tumor formation occurs even after delayed oncogene induction upon Dox administration. The left panel shows the schedule for the induction experiment, where oncogene activation can trigger tumor growth as much as 17 weeks postcell transplantation. Coexpression of luciferase with the oncogene allowed the quantification of tumor outgrowth by bioluminescence imaging. The right panel shows the localization of the tumors in life mice and also illustrates the tightness of Tet control in this experimental system. The figure is taken from Podsypanina et al. (2008) (Fig. 2A in the original manuscript; with permission).

The Power of Reversibility

443

findings, if applicable to the human case, will without any doubt have profound implications for the diagnosis and treatment of human cancer.

6. Secondary iPSC Technology Recently, a new ‘‘category’’ of Tet-transgenic mice emerged, which is not anticipated to be directly used for the analysis of the mice themselves. At the cellular level, the Tet technology has frequently been an integral part in the reprogramming of somatic cells to reach a pluripotent stage. The resulting iPSCs (induced pluripotent stem cells) hold great promise as powerful tools in developmental biology but also in regenerative medicine (Kiskinis and Eggan, 2010). The most common experimental approach involves the introduction of an ES cell-specific set of transcription factors into somatic cells (Takahashi and Yamanaka, 2006). Early on in the development of this technology, these so-called reprogramming factors have been placed under the control of Ptet. Tailor-made gene transfer vectors for the purpose of Tet control over these genes include adenoviruses (Stadtfeld et al., 2008), transposons (Woltjen et al., 2009), and most notably lentiviruses (Brambrink et al., 2008). When the latter were introduced in mouse embryo fibroblasts expressing rtTA-M2 under control of the ubiquitously active ROSA26 promoter, clonal ‘‘primary’’ iPSCs could be recovered with relatively high frequency (Wernig et al., 2008). Dox control enabled the shutdown of expression of the reprogramming factors, upon which the iPSCs were injected into blastocysts. These experiments yielded chimeras, from which genetically homogenous ‘‘secondary’’ somatic cells could be isolated that contained all the genes essential for regaining iPSC status under Tet-On control. Activating these genes by administering Dox induces the formation of genetically homogenous ‘‘secondary’’ iPSCs. In continuation of this work the Jaenisch group bred chimeric mice from primary iPSC injection into blastocytes with wild-type littermates, thus segregating the independently integrated proviral copies with the reprogramming factors under Tet control (Markoulaki et al., 2009). This work resulted in a collection of different transgenic lines containing defined subsets of reprogramming factor genes. These animals will be a valuable source for cell lines that can be used for screening substances that, in the presence of Dox, can substitute for the missing reprogramming factors. Any such findings would be a considerable step in bringing iPSC technology closer to the clinic, as it would facilitate the generation of human iPSC without genetic intervention. This is one of the rare examples where Tet-transgenic mice were not generated for the primary sake of creating conditional mouse models, but rather as a resource for a ‘‘mouse-derived product.’’ Nevertheless, this approach will undoubtedly find widespread applications.

444

¨nig et al. Kai Scho

7. Transgenic Rats Rats became the first mammalian species to be domesticated for scientific purposes and are therefore well-characterized animals, valuable in biological and medical research. Moreover, in comparison to mice, their physiology and larger size make them distinctly more suitable models for medical research including cardiovascular, renal, and pulmonary diseases as well as transplantation or regeneration biology to mention just a few important areas of medical sciences. In addition, their cognitive capabilities make the rat an excellent model in neurobiological research including neurodegenerative diseases. Progress in many of these existing fields has been hampered by technical limitations to target the rat genome in a defined way. With the release of the sequence of the rat genome and recent technological developments for targeting genes and gene activities, which include genetic modification via sequence-specific zinc-finger nucleases (Geurts et al., 2009) and controlled expression of defined shRNAs (Kotnik et al., 2009), a ‘‘renaissance of the rat’’ (Abbott, 2004) as animal model will likely become reality in the not too distant future. In 2008, two research groups described the generation of germ line competent rat ES cells (Buehr et al., 2008; Li et al., 2008), though animals engineered via homologous recombination in ES cells have not yet been reported. Several publications demonstrated the principle functionality of the Tet technology in transgenic rats (Barton et al., 2002; Konopka et al., 2009; Sheng et al., 2010; Zhou et al., 2009). First encouraging results for tight tissue-specific regulation even in the rat brain come from the laboratory of D. Bartsch and K. Scho¨nig, Mannheim, where a first rat line expressing tTA under the CamKIIa promoter allows to regulate an indicator genes in a spatially and temporally defined manner comparable to the CamK-tTA mouse line (D. Bartsch and K. Scho¨nig, unpublished results). There can be little doubt that once the technology for modifying the rat genome is further advanced, conditional mutants based on the principle of Tet technology will make a wide spectrum of questions amenable to genetic analysis.

8. Future Perspectives Early after its adaptation to transgenic mice, the Tet technology has been combined with other genetic approaches. Most prominent among these are recombinase-based systems for chromosome engineering. For example, controlling CRE expression via tTA/rtTA allows to limit the recombinase action to a defined time window, thereby also avoiding

The Power of Reversibility

445

deleterious effects of constitutive recombinase activity during the life span of an animal (Garcia and Mills, 2002). A most useful tool for this approach is the well-characterized LC-1 mouse, where CRE synthesis is tightly controlled by Ptet (Scho¨nig et al., 2002). Furthermore, temporally defined cell-typespecific cell ablation via Tet-controlled expression of the diphtheria toxin as originally described by the group of Fishman (Lee et al., 1998) allows to study the role of specific cell populations within an organ or tissue and to explore questions concerning regeneration potentials and irreversible damage. Also, the possibility to generate conditionally immortalized primary cells from Tet-transgenic animals, as first demonstrated by Efrat et al. (1995), exemplifies an application of Tet control, which undoubtedly will continue to have considerable impact in investigating and exploiting primary cells. Here, we like to point out some further combinations of technologies, which hold considerable promise for addressing exiting questions in novel in vivo models.

8.1. Conditional knockdown via RNAi The discovery of RNA interference (RNAi) as novel regulatory principle was quickly exploited for developing tools for conditional gene control. In principle, it should allow researchers to down regulate the expression of any endogenous gene for which sequence information is available. Promising results from nonmammalian organisms and successes in mammalian cell lines sparked considerable interest in adapting the underlying methods to mice. With synthetic short interfering RNAs (siRNA), the major problem still to be solved is the reliable delivery of siRNAs to defined tissues or cell types in vivo. Here, we will focus on the controlled mediation of RNAi by self-processing shRNAs (Chang et al., 2006). Especially for transgenic mice this concept has distinct advantages, as these shRNAs can be transcribed from RNA polymerase II (pol II)-type promoters and thus the ‘‘Tet Zoo’’ of existing mouse lines tissue specifically expressing rtTA or tTA can be exploited. Thus the problem is primarily reduced to designing proper shRNAs, which efficiently and specifically mediate the gene knockdown via RNAi. In cell lines, this approach has been convincingly shown (Weidenfeld et al., 2009). A striking demonstration of the potential of this approach in vivo is the work by Dickins et al. (2007). They explored the Dox-dependent knockdown of Trp53 in transgenic murine cancer models, in an analysis taking full advantage of various Tet transactivator lines and the reversibility of the shRNA-mediated knockdown. However, we are not aware of a great number of other successful applications of this approach (see also below). For a more comprehensive discussion about the current status of conditional RNAi approaches, we refer to reviews such as Lee and Kumar (2009).

446

¨nig et al. Kai Scho

Conditional production of shRNAs can also be achieved by pol IIIspecific promoters, such as H1 or U6 promoters equipped with tetO sequences. These can be regulated by TetR- or Tet-controlled transcriptional silencers (Deuschle et al., 1995; Freundlieb et al., 1997). Even though this approach is functional in vivo (Herold et al., 2008; Seibler et al., 2007; Szulc et al., 2006), it needs to be considered that it lacks tissue specificity and requires consistent and ubiquitous expression of TetR or tTS (Gossen, 2006).

8.2. Integrating the binary Tet-control system into genomes A genuine problem in stably setting up Tet regulation in the ‘‘classical mode’’ is the random and unpredictable integration of transgenes in genomes. This is the major reason for inconsistent and unspecific expression patterns observed at the cellular as well as at the organismal level. The factors affecting the expression characteristics include the integration site of the transgene within the genome, the number and orientation of integrates, but also the sequence composition of the integrate itself (Martin and Whitelaw, 1996). The problem applies to both integrates of the binary system, the Ptet-controlled expression unit as well as the transcription units in which tTA or rtTA encoding sequences are placed under the control of a supposedly cell-typespecific promoter. Despite of these shortcomings, numerous cell and mouse lines have been generated where Tet regulation functions impressively allowing for a range of regulation of 105–106-fold in cell lines (Gossen and Bujard, 1992) as well as in mice (Kistner et al., 1996), with no measurable background activity in the ‘‘OFF’’ state. We have termed chromosomal loci, which allow for such regulation of Ptet ‘‘silent but activatable’’ (s/a) loci. In HeLa cells (constitutively expressing rtTA), such s/a loci were functionally identified and made retargetable via the Flp-site-specific recombinase (Weidenfeld et al., 2009). In mice, a functionally similar locus, the LC-1 locus, allows for tight control of CRE recombinase (Scho¨nig et al., 2002). This latter locus was isolated and integrated into a bacterial artificial chromosome (BAC) for the transfer of Ptet-controlled expression units (see below). The s/a strategy has also been realized in ES cells, which have been used to generate transgenic animals. Two approaches have made use of precharacterized genetic loci, the type I collagen gene (ColA1) locus (Beard et al., 2006) and the hypoxanthine phosphoribosyltransferase (HPRT) locus (Palais et al., 2009). Both gene loci have been previously shown to support high levels of transgene expression in many cell types. A different route has been taken by Zeng et al. (2008), who screened for optimal s/a loci in ES cells. After random integration of Tet-inducible retroviruses, they identified a very promising integration site in the vicinity of the ubiquitously expressed gene Carm1 on chromosome 9 qA3. Tightly controlled expression from this locus was demonstrated in various

The Power of Reversibility

447

tissues of adult animals after breeding ES cell-derived mice to different tTA-expressing lines. The results of these approaches are very encouraging and have already led to impressive results (Carey et al., 2010; Foudi et al., 2009). It remains to be seen, however, whether the s/a loci identified up to date will function in all cell types of interest. Nevertheless, there can be little doubt that in the future suitable loci, which can be targeted by homologous recombination, will allow to circumvent much of the unpredictable nature of generating suitable Tet-responder lines. Similarly, transfer of BAC constructs carrying Ptet-controlled transcription units, or even the entire binary regulation system with the gene of interest, will significantly facilitate the setting up of Tet-control in vivo, particularly in species like the rat where homologous recombination at the ES cell level is not (yet) feasible. The production of transactivator mouse lines with predictable cell-type specificity can still be a problem, as respective promoters, even when well characterized in situ, are still sensitive to chromosomal position effects when used ectopically. On the other hand, exactly these local effects often result in exiting artificial specificities, which are highly valuable (though not necessarily to the creator of the line!). It is therefore important that the researchers in the field at least deposit some of these lines to make them available to their colleagues. Considering the recent progress in these fields, it is attractive to speculate that the targeted transfer of Ptet-controlled transcription units to proper genomic loci will also improve the chances to obtain animals with tightly controlled expression of shRNAs. This will result in double transgenic mice, where in principle any endogenous gene can be functionally targeted by knockdown, without touching its genomic locus. Tissue or cell-type selectivity of this conditional knockdown will depend on the Tet transactivator line used. As in the past, we can expect a wealth of new information and many surprises particularly due to the potential of the system to reversibly perturb gene functions in vivo. Here, as in all the other applications covered in this review, the true reversibility as the hallmark of the Tet-control principle will allow further paradigm changing insights in biological processes.

REFERENCES Abbott, A. (2004). Laboratory animals: The Renaissance rat. Nature 428, 464–466. Alvarez-Saavedra, M., et al. (2007). Cell-specific expression of wild-type MeCP2 in mouse models of Rett syndrome yields insight about pathogenesis. Hum. Mol. Genet. 16, 2315–2325. Anastassiadis, K., Glaser, S., Kranz, A., Berhardt, K., and Stewart, A. F. (2010). A practical summary of site-specific recombination, conditional mutagenesis, and tamoxifen induction of CreERT2. Methods Enzymol. 477, 109–123.

448

¨nig et al. Kai Scho

Barco, A., et al. (2002). Expression of constitutively active CREB protein facilitates the late phase of long-term potentiation by enhancing synaptic capture. Cell 108, 689–703. Baron, U., Freundlieb, S., Gossen, M., and Bujard, H. (1995). Co-regulation of two gene activities by tetracycline via a bidirectional promoter. Nucleic Acids Res. 23, 3605–3606. Barton, M. D., Dunlop, J. W., Psaltis, G., Kulik, J., DeGennaro, L., and Kwak, S. P. (2002). Modified GFAP promoter auto-regulates tet-activator expression for increased transactivation and reduced tTA-associated toxicity. Brain Res. Mol. Brain Res. 101, 71–81. Beard, C., Hochedlinger, K., Plath, K., Wutz, A., and Jaenisch, R. (2006). Efficient method to generate single-copy transgenic mice by site-specific integration in embryonic stem cells. Genesis 44, 23–28. Bejar, R., Yasuda, R., Krugers, H., Hood, K., and Mayford, M. (2002). Transgenic calmodulin-dependent protein kinase II activation: Dose-dependent effects on synaptic plasticity, learning, and memory. J. Neurosci. 22, 5719–5726. Berens, C., and Hillen, W. (2003). Gene regulation by tetracyclines. Constraints of resistance regulation in bacteria shape TetR for application in eukaryotes. Eur. J. Biochem. 270, 3109–3121. Brambrink, T., Foreman, R., Welstead, G. G., Lengner, C. J., Wernig, M., Suh, H., and Jaenisch, R. (2008). Sequential expression of pluripotency markers during direct reprogramming of mouse somatic cells. Cell Stem Cell 2, 151–159. Buehr, M., Meek, S., Blair, K., Yang, J., Ure, J., Silva, J., McLay, R., Hall, J., Ying, Q. L., and Smith, A. (2008). Capture of authentic embryonic stem cells from rat blastocysts. Cell 135, 1287–1298. Carey, B. W., Markoulaki, S., Beard, C., Hanna, J., and Jaenisch, R. (2010). Single-gene transgenic mouse strains for reprogramming adult somatic cells. Nat. Methods 7, 56–59. Celikel, T., et al. (2007). Select overexpression of homer1a in dorsal hippocampus impairs spatial working memory. Front Neurosci. 1, 97–110. Chang, K., Elledge, S. J., and Hannon, G. J. (2006). Lessons from nature: microRNA-based shRNA libraries. Nat. Methods 3, 707–714. Chen, J., et al. (1998). Transgenic animals with inducible, targeted gene expression in brain. Mol. Pharmacol. 54, 495–503. Chikama, T., et al. (2005). Characterization of tetracycline-inducible bitransgenic Krt12rtTA/þ/tet-O-LacZ mice. Invest. Ophthalmol. Vis. Sci. 46, 1966–1972. Chin, L., et al. (1999). Essential role for oncogenic Ras in tumour maintenance. Nature 400, 468–472. Das, A. T., Zhou, X., Vink, M., Klaver, B., Verhoef, K., Marzio, G., and Berkhout, B. (2004). Viral evolution as a tool to improve the tetracycline-regulated gene expression system. J. Biol. Chem. 279, 18776–18782. Deuschle, U., Meyer, W. K., and Thiesen, H. J. (1995). Tetracycline-reversible silencing of eukaryotic promoters. Mol. Cell. Biol. 15, 1907–1914. Deutsch, U., et al. (2008). Inducible endothelial cell-specific gene expression in transgenic mouse embryos and adult mice. Exp. Cell Res. 314, 1202–1216. Diamond, I., et al. (2000). Conditional gene expression in the epidermis of transgenic mice using the tetracycline-regulated transactivators tTA and rTA linked to the keratin 5 promoter. J. Invest. Dermatol. 115, 788–794. Dickins, R. A., McJunkin, K., Hernando, E., Premsrirut, P. K., Krizhanovsky, V., Burgess, D. J., Kim, S. Y., Cordon-Cardo, C., Zender, L., Hannon, G. J., and Lowe, S. W. (2007). Tissue-specific and reversible RNA interference in transgenic mice. Nat. Genet. 39, 914–921. Doring, A., et al. (2007). E- and P-selectin are not required for the development of experimental autoimmune encephalomyelitis in C57BL/6 and SJL mice. J. Immunol. 179, 8470–8479.

The Power of Reversibility

449

Dudai, Y. (2002). Memory from A to Z. Keywords, Concepts and Beyond. Oxford University Press, Oxford. Dunbar, M. E., et al. (2001). Temporally regulated overexpression of parathyroid hormonerelated protein in the mammary gland reveals distinct fetal and pubertal phenotypes. J. Endocrinol. 171, 403–416. Eckermann, K., et al. (2007). The beta-propensity of Tau determines aggregation and synaptic loss in inducible mouse models of tauopathy. J. Biol. Chem. 282, 31755–31765. Efrat, S., Fusco-DeMane, D., Lemberg, H., Al Emran, O., and Wang, X. (1995). Conditional transformation of a pancreatic beta-cell line derived from transgenic mice expressing a tetracycline-regulated oncogene. Proc. Natl. Acad. Sci. USA 92, 3576–3580. Felsher, D. W. (2008). Reversing cancer from inside and out: Oncogene addiction, cellular senescence, and the angiogenic switch. Lymphat. Res. Biol. 6, 149–154. Felsher, D. W., and Bishop, J. M. (1999). Reversible tumorigenesis by MYC in hematopoietic lineages. Mol. Cell 4, 199–207. Foudi, A., Hochedlinger, K., Van Buren, D., Schindler, J. W., Jaenisch, R., Carey, V., and Hock, H. (2009). Analysis of histone 2B-GFP retention reveals slowly cycling hematopoietic stem cells. Nat. Biotechnol. 27, 84–90. Freundlieb, S., Baron, U., Bonin, A. L., Gossen, M., and Bujard, H. (1997). Use of tetracycline-controlled gene expression systems to study mammalian cell cycle. Methods Enzymol. 283, 159–173. Fridmacher, V., et al. (2003). Forebrain-specific neuronal inhibition of nuclear factorkappaB activity leads to loss of neuroprotection. J. Neurosci. 23, 9403–9408. Garcia, E. L., and Mills, A. A. (2002). Getting around lethality with inducible Cre-mediated excision. Semin. Cell Dev. Biol. 13, 151–158. Geurts, A. M., Cost, G. J., Freyvert, Y., Zeitler, B., Miller, J. C., Choi, V. M., Jenkins, S. S., Wood, A., Cui, X., Meng, X., Vincent, A., Lam, S., et al. (2009). Knockout rats via embryo microinjection of zinc-finger nucleases. Science 325, 433. Gimenez, E., et al. (2004). A transgenic mouse model with inducible Tyrosinase gene expression using the tetracycline (Tet-on) system allows regulated rescue of abnormal chiasmatic projections found in albinism. Pigment Cell Res. 17, 363–370. Gogos, J. A., et al. (2000). Genetic ablation and restoration of the olfactory topographic map. Cell 103, 609–620. Gossen, M. (2006). Conditional gene expression: Intelligent designs. Gene Ther. 13, 1251–1252. Gossen, M., and Bujard, H. (1992). Tight control of gene expression in mammalian cells by tetracycline-responsive promoters. Proc. Natl. Acad. Sci. USA 89, 5547–5551. Gossen, M., and Bujard, H. (2002). Studying gene function in eukaryotes by conditional gene inactivation. Annu. Rev. Genet. 36, 153–173. Gossen, M., Bonin, A. L., and Bujard, H. (1993). Control of gene activity in higher eukaryotic cells by prokaryotic regulatory elements. Trends Biochem. Sci. 18, 471–475. Gossen, M., Freundlieb, S., Bender, G., Muller, G., Hillen, W., and Bujard, H. (1995). Transcriptional activation by tetracyclines in mammalian cells. Science 268, 1766–1769. Gross, C., et al. (2002). Serotonin1A receptor acts during development to establish normal anxiety-like behaviour in the adult. Nature 416, 396–400. Gunther, E. J., et al. (2002). A novel doxycycline-inducible system for the transgenic analysis of mammary gland biology. FASEB J. 16, 283–292. Guo, Y., et al. (1999). Conversion of C57Bl/6 mice from a tumor promotion-resistant to a -sensitive phenotype by enhanced ornithine decarboxylase expression. Mol. Carcinog. 26, 32–36. Herold, M. J., van den Brandt, J., Seibler, J., and Reichardt, H. M. (2008). Inducible and reversible gene silencing by stable integration of an shRNA-encoding lentivirus in transgenic rats. Proc. Natl. Acad. Sci. USA 105, 18507–18512.

450

¨nig et al. Kai Scho

Hess, J., et al. (2001). Induction of pre-B cell proliferation after de novo synthesis of the pre-B cell receptor. Proc. Natl. Acad. Sci. USA 98, 1745–1750. Ju, H., et al. (2001). Conditional and targeted overexpression of vascular chymase causes hypertension in transgenic mice. Proc. Natl. Acad. Sci. USA 98, 7469–7474. Kerrison, J. B., et al. (2005). A system for inducible gene expression in retinal ganglion cells. Invest. Ophthalmol. Vis. Sci. 46, 2932–2939. Kholodilov, N., et al. (2004). Regulation of the development of mesencephalic dopaminergic systems by the selective expression of glial cell line-derived neurotrophic factor in their targets. J. Neurosci. 24, 3136–3146. Kiskinis, E., and Eggan, K. (2010). Progress toward the clinical application of patient-specific pluripotent stem cells. J. Clin. Invest. 120, 51–59. Kistner, A., Gossen, M., Zimmermann, F., Jerecic, J., Ullmer, C., Lubbert, H., and Bujard, H. (1996). Doxycycline-mediated quantitative and tissue-specific control of gene expression in transgenic mice. Proc. Natl. Acad. Sci. USA 93, 10933–10938. Klein, C. A. (2008). Cancer. The metastasis cascade. Science 321, 1785–1787. Knostman, K. A., et al. (2007). Creation and characterization of a doxycycline-inducible mouse model of thyroid-targeted RET/PTC1 oncogene and luciferase reporter gene coexpression. Thyroid 17, 1181–1188. Kolber, B. J., et al. (2010). Transient early-life forebrain corticotropin-releasing hormone elevation causes long-lasting anxiogenic and despair-like changes in mice. J. Neurosci. 30, 2571–2581. Konopka, W., Duniec, K., Klejman, A., Wawrzyniak, M., Owczarek, D., Gawrys, L., Maleszewski, M., Mallet, J., and Kaczmarek, L. (2009). Tet system in the brain: Transgenic rats and lentiviral vectors approach. Genesis 47, 274–280. Koschmieder, S., et al. (2005). Inducible chronic phase of myeloid leukemia with expansion of hematopoietic stem cells in a transgenic model of BCR-ABL leukemogenesis. Blood 105, 324–334. Kotnik, K., Popova, E., Todiras, M., Mori, M. A., Alenina, N., Seibler, J., and Bader, M. (2009). Inducible transgenic rat model for diabetes mellitus based on shRNA-mediated gene knockdown. PLoS One 4, e5124. Krestel, H. E., Shimshek, D. R., Jensen, V., Nevian, T., Kim, J., Geng, Y., Bast, T., Depaulis, A., Scho¨nig, K., Schwenk, F., Bujard, H., Hvalby, O., et al. (2004). A genetic switch for epilepsy in adult mice. J. Neurosci. 24, 10568–10578. Lee, S. K., and Kumar, P. (2009). Conditional RNAi: Towards a silent gene therapy. Adv. Drug Deliv. Rev. 61, 650–664. Lee, P., Morley, G., Huang, Q., Fischer, A., Seiler, S., Horner, J. W., Factor, S., Vaidya, D., Jalife, J., and Fishman, G. I. (1998). Conditional lineage ablation to model human diseases. Proc. Natl. Acad. Sci. USA 95, 11371–11376. Lee, H. G., et al. (2009). The neuronal expression of MYC causes a neurodegenerative phenotype in a novel transgenic mouse. Am. J. Pathol. 174, 891–897. Leenders, H., et al. (2000). Role of the forkhead transcription family member, FKHR, in thymocyte differentiation. Eur. J. Immunol. 30, 2980–2990. Li, P., Tong, C., Mehrian-Shai, R., Jia, L., Wu, N., Yan, Y., Maxson, R. E., Schulze, E. N., Song, H., Hsieh, C. L., Pera, M. F., and Ying, Q. L. (2008). Germline competent embryonic stem cells derived from rat blastocysts. Cell 135, 1299–1310. Licht, T., et al. (2010). VEGF is required for dendritogenesis of newly born olfactory bulb interneurons. Development 137, 261–271. Mack, V., et al. (2001). Conditional restoration of hippocampal synaptic potentiation in Glur-A-deficient mice. Science 292, 2501–2504. Mansuy, I. M., et al. (1998a). Inducible and reversible gene expression with the rtTA system for the study of memory. Neuron 21, 257–265.

The Power of Reversibility

451

Mansuy, I. M., et al. (1998b). Restricted and regulated overexpression reveals calcineurin as a key component in the transition from short-term to long-term memory. Cell 92, 39–49. Markoulaki, S., Hanna, J., Beard, C., Carey, B. W., Cheng, A. W., Lengner, C. J., Dausman, J. A., Fu, D., Gao, Q., Wu, S., Cassady, J. P., and Jaenisch, R. (2009). Transgenic mice with defined combinations of drug-inducible reprogramming factors. Nat. Biotechnol. 27, 169–171. Martin, D. I., and Whitelaw, E. (1996). The vagaries of variegating transgenes. Bioessays 18, 919–923. Mayford, M., Bach, M. E., Huang, Y. Y., Wang, L., Hawkins, R. D., and Kandel, E. R. (1996). Control of memory formation through regulated expression of a CaMKII transgene. Science 274, 1678–1683. Mehrad, B., et al. (2002). Transient lung-specific expression of the chemokine KC improves outcome in invasive aspergillosis. Am. J. Respir. Crit. Care Med. 166, 1263–1268. Michalon, A., Koshibu, K., Baumgartel, K., Spirig, D. H., and Mansuy, I. M. (2005). Inducible and neuron-specific gene expression in the adult mouse brain with the rtTA2S-M2 system. Genesis 43, 205–212. Mocanu, M. M., et al. (2008). The potential for beta-structure in the repeat domain of tau protein determines aggregation, synaptic decay, neuronal loss, and coassembly with endogenous Tau in inducible mouse models of tauopathy. J. Neurosci. 28, 737–748. Nakashiba, T., et al. (2008). Transgenic inhibition of synaptic transmission reveals role of CA3 output in hippocampal learning. Science 319, 1260–1264. Nicholls, R. E., et al. (2006). mGluR2 acts through inhibitory Galpha subunits to regulate transmission and long-term plasticity at hippocampal mossy fiber-CA3 synapses. Proc. Natl. Acad. Sci. USA 103, 6380–6385. Nie, T., et al. (2007). Genetic disruption of protein kinase A anchoring reveals a role for compartmentalized kinase signaling in theta-burst long-term potentiation and spatial memory. J. Neurosci. 27, 10278–10288. Nuber, S., et al. (2008). Neurodegeneration and motor dysfunction in a conditional model of Parkinson’s disease. J. Neurosci. 28, 2471–2484. Palais, G., Nguyen Dinh Cat, A., Friedman, H., Panek-Huet, N., Millet, A., Tronche, F., Gellen, B., Mercadier, J. J., Peterson, A., and Jaisser, F. (2009). Targeted transgenesis at the HPRT locus: An efficient strategy to achieve tightly controlled in vivo conditional expression with the tet system. Physiol. Genomics 37, 140–146. Passman, R. S., and Fishman, G. I. (1994). Regulated expression of foreign genes in vivo after germline transfer. J. Clin. Invest. 94, 2421–2425. Paulson, J. B., et al. (2008). Amyloid plaque and neurofibrillary tangle pathology in a regulatable mouse model of Alzheimer’s disease. Am. J. Pathol. 173, 762–772. Podsypanina, K., Du, Y. C., Jechlinger, M., Beverly, L. J., Hambardzumyan, D., and Varmus, H. (2008). Seeding and propagation of untransformed mouse mammary cells in the lung. Science 321, 1841–1844. Pot, C., et al. (2002). Nogo-A expressed in Schwann cells impairs axonal regeneration after peripheral nerve injury. J. Cell Biol. 159, 29–35. Proctor, B. M., et al. (2008). Requirement for p38 mitogen-activated protein kinase activity in neointima formation after vascular injury. Circulation 118, 658–666. Radomska, H. S., et al. (2002). Transgenic targeting with regulatory elements of the human CD34 gene. Blood 100, 4410–4419. Ramsden, M., et al. (2005). Age-dependent neurofibrillary tangle formation, neuron loss, and memory impairment in a mouse model of human tauopathy (P301L). J. Neurosci. 25, 10637–10647. Roth, S., et al. (2009). Generation of a tightly regulated doxycycline-inducible model for studying mouse intestinal biology. Genesis 47, 7–13.

452

¨nig et al. Kai Scho

Santacruz, K., et al. (2005). Tau suppression in a neurodegenerative mouse model improves memory function. Science 309, 476–481. Sarao, R., and Dumont, D. J. (1998). Conditional transgene expression in endothelial cells. Transgenic Res. 7, 421–427. Scho¨nig, K., and Bujard, H. (2003). Generating conditional mouse mutants via tetracyclinecontrolled gene expression. Methods Mol. Biol. 209, 69–104. Scho¨nig, K., Schwenk, F., Rajewsky, K., and Bujard, H. (2002). Stringent doxycycline dependent control of CRE recombinase in vivo. Nucleic Acids Res. 30, e134. Seibler, J., Kleinridders, A., Kuter-Luks, B., Niehaves, S., Bruning, J. C., and Schwenk, F. (2007). Reversible gene knockdown in mice using a tight, inducible shRNA expression system. Nucleic Acids Res. 35, e54. Shakya, R., et al. (2005). The role of GDNF in patterning the excretory system. Dev. Biol. 283, 70–84. Sheng, Y., Lin, C. C., Yue, J., Sukhwani, M., Shuttleworth, J. J., Chu, T., and Orwig, K. E. (2010). Generation and characterization of a Tet-On (rtTA-M2) transgenic rat. BMC Dev. Biol. 10, 17. Shimizu, E., et al. (2000). NMDA receptor-dependent synaptic reinforcement as a crucial process for memory consolidation. Science 290, 1170–1174. Stadtfeld, M., Nagaya, M., Utikal, J., Weir, G., and Hochedlinger, K. (2008). Induced pluripotent stem cells generated without viral integration. Science 322, 945–949. Szulc, J., Wiznerowicz, M., Sauvain, M. O., Trono, D., and Aebischer, P. (2006). A versatile tool for conditional gene expression and knockdown. Nat. Methods 3, 109–116. Takahashi, K., and Yamanaka, S. (2006). Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663–676. Teng, P. I., et al. (2002). Inducible and selective transgene expression in murine vascular endothelium. Physiol. Genomics 11, 99–107. Thomas, M. K., et al. (2001). Development of diabetes mellitus in aging transgenic mice following suppression of pancreatic homeoprotein IDX-1. J. Clin. Invest. 108, 319–329. Tichelaar, J. W., et al. (2000). Conditional expression of fibroblast growth factor-7 in the developing and mature lung. J. Biol. Chem. 275, 11858–11864. Urlinger, S., Baron, U., Thellmann, M., Hasan, M. T., Bujard, H., and Hillen, W. (2000). Exploring the sequence space for tetracycline-dependent transcriptional activators: Novel mutations yield expanded range and sensitivity. Proc. Natl. Acad. Sci. USA 97, 7963–7968. Valencik, M. L., and McDonald, J. A. (2001). Codon optimization markedly improves doxycycline regulated gene expression in the mouse heart. Transgenic Res. 10, 269–275. Viosca, J., et al. (2009). Enhanced CREB-dependent gene expression increases the excitability of neurons in the basal amygdala and primes the consolidation of contextual and cued fear memory. Learn. Mem. 16, 193–197. Vitale-Cross, L., et al. (2004). Conditional expression of K-ras in an epithelial compartment that includes the stem cells is sufficient to promote squamous cell carcinogenesis. Cancer Res. 64, 8804–8807. Wehr, M., et al. (2009). Transgenic silencing of neurons in the mammalian brain by expression of the allatostatin receptor (AlstR). J. Neurophysiol. 102, 2554–2562. Weidenfeld, I., Gossen, M., Low, R., Kentner, D., Berger, S., Gorlich, D., Bartsch, D., Bujard, H., and Scho¨nig, K. (2009). Inducible expression of coding and inhibitory RNAs from retargetable genomic loci. Nucleic Acids Res. 37, e50. Weinberg, R. A. (2008). Leaving home early: Reexamination of the canonical models of tumor progression. Cancer Cell 14, 283–284. Wernig, M., Lengner, C. J., Hanna, J., Lodato, M. A., Steine, E., Foreman, R., Staerk, J., Markoulaki, S., and Jaenisch, R. (2008). A drug-inducible transgenic system for direct reprogramming of multiple somatic cell types. Nat. Biotechnol. 26, 916–924.

The Power of Reversibility

453

West, J., et al. (2004). Pulmonary hypertension in transgenic mice expressing a dominantnegative BMPRII gene in smooth muscle. Circ. Res. 94, 1109–1114. Wiesner, S. M., et al. (2005). Repressible transgenic model of NRAS oncogene-driven mast cell disease in the mouse. Blood 106, 1054–1062. Woltjen, K., Michael, I. P., Mohseni, P., Desai, R., Mileikovsky, M., Hamalainen, R., Cowling, R., Wang, W., Liu, P., Gertsenstein, M., Kaji, K., Sung, H. K., et al. (2009). piggyBac transposition reprograms fibroblasts to induced pluripotent stem cells. Nature 458, 766–770. Xie, W., et al. (1999). Conditional expression of the ErbB2 oncogene elicits reversible hyperplasia in stratified epithelia and up-regulation of TGFalpha expression in transgenic mice. Oncogene 18, 3593–3607. Yamamoto, A., et al. (2000). Reversal of neuropathology and motor dysfunction in a conditional model of Huntington’s disease. Cell 101, 57–66. Ye, L., et al. (2001). Regulated expression of the human CFTR gene in epithelial cells. Mol. Ther. 3, 723–733. Ye, X., et al. (2008). Divergent roles of endothelial NF-kappaB in multiple organ injury and bacterial clearance in mouse models of sepsis. J. Exp. Med. 205, 1303–1315. Yu, H. M., et al. (2005). Development of a unique system for spatiotemporal and lineagespecific gene expression in mice. Proc. Natl. Acad. Sci. USA 102, 8615–8620. Yu, H. M., et al. (2007). Impaired neural development caused by inducible expression of Axin in transgenic mice. Mech. Dev. 124, 146–156. Zeng, H., Horie, K., Madisen, L., Pavlova, M. N., Gragerova, G., Rohde, A. D., Schimpf, B. A., Liang, Y., Ojala, E., Kramer, F., Roth, P., Slobodskaya, O., et al. (2008). An inducible and reversible mouse genetic rescue system. PLoS Genet. 4, e1000069. Zhou, X., Vink, M., Klaver, B., Berkhout, B., and Das, A. T. (2006). Optimization of the Tet-On system for regulated gene expression through viral evolution. Gene Ther. 13, 1382–1390. Zhou, H., Huang, C., Yang, M., Landel, C. P., Xia, P. Y., Liu, Y. J., and Xia, X. G. (2009). Developing tTA transgenic rats for inducible and reversible gene expression. Int. J. Biol. Sci. 5, 171–181. Zhou, M., et al. (2010). Forebrain overexpression of CK1delta leads to down-regulation of dopamine receptors and altered locomotor activity reminiscent of ADHD. Proc. Natl. Acad. Sci. USA 107, 4401–4406. Zhu, P., Aller, M. I., Baron, U., Cambridge, S., Bausen, M., Herb, J., Sawinski, J., Cetin, A., Osten, P., Nelson, M. L., Kugler, S., Seeburg, P. H., et al. (2007). Silencing and un-silencing of tetracycline-controlled genes in neurons. PLoS One 2, e533.

C H A P T E R

T W E N T Y- T H R E E

Gene Expression Profiling of Mouse Oocytes and Preimplantation Embryos Francesca E. Duncan* and Richard M. Schultz† Contents 1. Introduction 2. Experimental Design Considerations 3. RNA Isolation and cRNA Target Preparation Protocol (Based on the Affymetrix Protocol for Eukaryotic Small Sample Target Labeling Assay Version II) 3.1. Isolation of total RNA and synthesis of double-stranded cDNA (day 1) 3.2. Recovery of double-stranded cDNA and first cycle of amplification (day 2) 3.3. Synthesis of double-stranded cDNA from first round of amplified cRNA (day 3) 3.4. Recovery of double-stranded cDNA, second cycle of amplification, and cRNA labeling (day 4) 3.5. cRNA fragmentation and assessment by gel electrophoresis (day 5) 4. Quality Control Assessment of cRNA 5. Methods for Microarray Data Analysis 6. Validation Methods 7. Archiving Results 8. Future Directions Acknowledgments References

458 463

464 466 469 470 471 472 473 474 476 476 476 478 479

Abstract Gene expression profiling using microarray technology is a robust, efficient, and cost-effective approach to assay a cell or tissue’s transcriptome at a particular time or under a specific condition. Application of this technology to oocytes and * Department of Obstetrics and Gynecology, Northwestern University, Chicago, Illinois, USA Department of Biology, University of Pennsylvania, Philadelphia, Pennsylvania, USA

{

Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77023-3

#

2010 Elsevier Inc. All rights reserved.

457

458

Francesca E. Duncan and Richard M. Schultz

preimplantation embryos has been limited largely because this biological material is difficult to acquire in sufficient quantities. We describe here a protocol to isolate and amplify mRNA from oocytes and preimplantation embryos that is suitable for microarray experiments. This protocol is based on a linear two-step amplification protocol using T7 RNA polymerase-based in vitro transcription and has been used to isolate more than 80 mg of cRNA from only 20 oocytes or preimplantation embryos. Gene expression profiling has provided insight into the molecular mechanisms of meiotic maturation, fertilization, and preimplantation embryo development. It has also been used to characterize female gametes and embryos from animals harboring gene-specific knockouts or knockdowns. Finally, this technology has been useful in evaluating how various Assisted Reproductive Technologies impact global patterns of gene expression in resulting embryos.

1. Introduction Since the completion of the Human Genome Project in 2003 (Schmutz et al., 2004), there have been significant advances in sequencing technologies that have allowed the complete or partial genome sequencing of many diverse model organisms (Genomes Online Database, GOLD, www.genomesonline.org). The challenge now is to translate this large quantity of sequence data into meaningful biological information. The quest to decipher the meaning of genomic sequences has led to the field of transcriptomics or the study of the transcriptome, the complete set and quantity of transcripts in a tissue or cell at a given developmental stage and under a specific condition. High-throughput microarray technology is one of the most commonly used methods to assay global patterns of gene expression. In general terms, microarray platforms are plates, slides, or membranes spotted with gene-specific DNA probes that are either oligonucleotides or full-length cDNAs. In a microarray experiment, mRNA is extracted from experimental samples and converted into target cRNA (complementary RNA) labeled with, for example, biotin. When the labeled cRNA is fragmented and hybridized to the microarray platform, the target sequences bind to the corresponding probes through base pair complementarity. A fluorescent dye that binds to biotin is then washed over the platform, and the complementary matching between targets and probes is detected as fluorescent spots. The fluorescence intensity is proportional to the number of target cRNA molecules bound to the probe (DalmaWeiszhausz et al., 2006). Standard statistical methods are then used to convert the fluorescence intensities into usable biological information. This technology has many potential applications that range from basic science to clinical use and patient care. For example, gene expression profiling studies of breast cancer have provided clinicians and researchers a

Gene Expression Profiling

459

deeper understanding of the disease’s subtypes, progression, recurrence, and response to therapeutics (Sotiriou and Piccart, 2007). It has been difficult to apply microarray technology to the study of oocytes and preimplantation embryos, however, because the amount of material is limiting in quantity. A fully grown mouse oocyte, for example, is approximately 80 mm in diameter and contains about 80 pg of RNA. Depending on the strain and age of mouse, a hyperstimulated female will yield only between 10 and 80 oocytes. Thus, this amount of material is not sufficient for use in microarray protocols, which require microgram quantities of cRNA. We describe a protocol, largely based on the Affymetrix protocol for Eukaryotic Small Sample Target Labeling Assay Version II that we have optimized to obtain large quantities of highquality cRNA from small numbers of mouse oocytes and preimplantation embryos (Table 23.1). This technique involves a linear two-step amplification protocol using T7 RNA polymerase-based in vitro transcription (Van Gelder et al., 1990) (Fig. 23.1). Briefly, cDNA is synthesized from total extracted RNA using a primer containing oligo(dT) coupled to the T7 RNA polymerase promoter sequence. Following Second Strand synthesis, a double-stranded cDNA copy of each mRNA is made that contains the T7 RNA polymerase promoter binding site. Thus, T7 polymerase can then be used for in vitro transcription to generate amplified cRNA. To amplify substantial quantities of cRNA from less than 100 ng of starting material, as is the case when using oocytes and preimplantation embryos, a second round of in vitro transcription amplification is required (Dalma-Weiszhausz et al., 2006) (Fig. 23.1). During the second in vitro transcription reaction, biotinylated nucleotides are incorporated into the target. Typical cRNA yields following amplification from 20 oocytes or preimplantation embryos range from 30 to 80 mg (Table 23.1). This high quality biotinylated cRNA is then suitable for hybridization on microarray platforms. Microarrays on oocytes and preimplantation embryos using this specific RNA isolation and amplification protocol have allowed us to: (1) advance our knowledge of the basic biology of oogenesis, fertilization, and preimplantation embryo development; (2) characterize the gene expression profiles of gametes from mice with gene-specific knockouts or knockdowns; and (3) assess the effect of various assisted reproductive technologies (ART) on global patterns of gene expression in the resulting embryos (Table 23.1 and reviewed in Hamatani et al., 2006, 2008). This technology also has the potential to identify key molecular biomarkers to assess the health and quality of female gametes and preimplantation embryos. We focus primarily on the RNA isolation and amplification steps of a microarray experiment as the quality of target cRNA is one of the most crucial factors in an experiment’s success or failure. We also highlight

Table 23.1 Representative microarray experiments using oocytes or preimplantation embryos

Purpose of microarray experiment (reference)

Basic biology of oocytes and preimplantation embryos To characterize global patterns of gene expression that accompanies major morphological transitions during preimplantation development (Zeng et al., 2004).

Sample

GV-intact oocytes and preimplantation embryos (one-, two-, and eight-cell and blastocyst) MII-arrested eggs, one- and two-cell embryos

No. of oocytes or embryos/ sample

cRNA yield after 1st amplification (mg)

cRNA yield after 2nd amplification (mg)

80–480

2.0–10.8

62–119

0.03–0.06 0.07–0.20 0.20–0.25

8–12 28
5 42
2

0.30–0.80

45–70

0.35–1.20

30–70

0.30–1.10

40–70

325–380 To identify genes that are activated during zygotic gene activation by comparing the gene expression profiles of one- and two- cell embryos that were either treated or not treated with the RNA polymerase II-inhibitor, a-amanitin (Zeng and Schultz, 2005). 5 To characterize changes in gene expression patterns that GV-intact oocytes 10 occur during mouse oogenesis using oocytes isolated 20 from primordial, primary, secondary, small antral, and large antral follicles. The transcript profiles of oocytes developed in vitro and following hormonal stimulation were also determined (Pan et al., 2005). To determine the effect of altering the Ca2þ oscillatory Blastocysts 20 patterns following fertilization on gene expression profiles in the resulting blastocysts (Ozil et al., 2006). GV-intact oocytes and MII-arrested 25–35 To compare gene expression profiles of GV-intact eggs oocytes and MII-arrested eggs from young and old mice (Pan et al., 2008). Characterization of oocytes and preimplantation embryos from knockout or knockdown experiments 30 To confirm that use of long dsRNA transgenic RNAi GV-intact oocytes approaches in oocytes does not have off-targeting effects (Stein et al., 2005).

Two-cell embryos 20 To confirm that preimplantation embryos from conditional Zp3-Cre-mediated knockout of Brg1 arrest at the 2-cell stage due to zygotic genome activation defects (Bultman et al., 2006). GV-intact oocytes 25 To determine the effect of basonuclin deficiency on RNA polymerase II transcription in oocytes from mice transgenic for a Zp3-mediated RNAi hairpin targeting basonuclin (Ma et al., 2006). MI oocytes and MII-arrested eggs 15–20 To determine if maternal transcripts are misregulated in in vitro matured oocytes from Zp3-Cre mediated conditional Dicer knockout mice (Murchison et al., 2007). GV-intact oocytes 25 To examine transcription misregulation in oocytes depleted of CTCF by Zp3-mediated transgenic RNAi (Wan et al., 2008). 20 To determine if maternal transcripts are misregulated in GV-intact oocytes fully grown oocytes from Zp3-Cre mediated conditional Dicer knockout mice (Ma et al., 2010). Examining the effect of assisted reproduction technologies on oocytes and preimplantation embryos 80 To determine the effect of embryo culture in different Blastocysts culture media, from the one-cell to the blastocyst stage, on global patterns of gene expression in the resulting embryos (Rinaudo and Schultz, 2004). 80 To compare gene expression profiles of preimplantation Blastocysts embryos cultured either under atmospheric (20%) or physiologic (5%) oxygen concentrations (Rinaudo et al., 2006). Blastocysts 20 To determine the effect of the blastomere biopsy procedure, inherent to preimplantation genetic testing, on global patterns of gene expression in the resulting blastocysts (Duncan et al., 2009).

0.25–0.90

45–70

0.34–1.80

40–77

0.20–0.70

30–60

0.30–1.10

35–65

0.42–0.96

64–87

1.5–5

74–138

1.2–7.4

74–115

0.12–0.51

34–78

462

Francesca E. Duncan and Richard M. Schultz

5¢

Total RNA AAAAA

3¢

1. RNA isolation; Day 1

3¢

2. First strand cDNA synthesis; Day 1

+T7-oligo(dT) primer 5¢

AAAAA TTTTT-T7

3¢

5¢

5¢ AAAAA-T7 TTTTT-T7

3¢

3¢

3. Second strand cDNA synthesis; Day 1

5¢

+T7 polymerase and unlabeled ribonucleotides UUUUU

3¢

5¢

4. First round of amplification by in vitro transcription; Day 2

+Random primers 5¢

AAAAA UUUUU

3¢

3¢ 5¢

5. First strand cDNA synthesis from first round of amplification; Day 3

+T7-oligo(dT) primer 3¢

5¢

AAAAA-T7 TTTTT-T7 5¢

3¢

+T7 polymerase and biotinylated ribonucleotides 3¢

******** UUUUU 5¢

6. Second strand cDNA synthesis from first round of amplification; Day 3

7. Second round of amplification by in vitro transcription and cRNA labeling; Day 4

RNA DNA Metal-induced hydrolysis of amplified and biotinylated cRNA

T7

***

T7 promoter sequence Biotin

Figure 23.1 Schematic of the linear two-step amplification protocol for gene expression profiling sample preparation. Total RNA is isolated from samples, and reverse transcription using a primer containing a stretch of polythymines (oligo(dT)) coupled to the T7 RNA polymerase promoter sequence is used to synthesize double-stranded cDNA (Steps 1–3). In the first round of amplification, T7 polymerase is used in an in vitro transcription reaction with the cDNA as a template to create amplified cRNA (Step 4). The amplified cRNA is converted into cDNA, which is then used as the template in a second round of amplification (Steps 5 and 6). The second in vitro transcription reaction with T7 polymerase is performed with biotinylated ribonucleotides so that the final cRNA sample is biotin-labeled (Step 7). This final amplified and labeled product is subjected to metal-induced hydrolysis, and the fragmented cRNA is hybridized to a microarray platform.

Gene Expression Profiling

463

other key experiment components including experimental design, quality control parameters, data analysis and validation, and data archiving. Hybridization methods and data acquisition are beyond the scope of this review because they are highly dependent on the platform and equipment used and are usually outsourced to core facilities. For more detailed information on statistical analyses of microarray data, we refer the reader to the review by White and Salamonsen (2005) or suggest that the reader consults a bioinformatics expert.

2. Experimental Design Considerations Several experimental design factors must be considered prior to commencing a microarray experiment, and consultation with a statistician or bioinformatics expert is warranted. Biological replicates should be included in a microarray experiment to account for variation and to allow for statistically significant analyses to be performed. Biological variation, or the natural variation that occurs between patients, animals, tissues, or cells, tends to be a greater source of variation compared to technical variation (White and Salamonsen, 2005). Four biological replicates provide sufficient statistical power and confidence levels to detect a 1.4-fold difference in transcript abundance (Zeng et al., 2004). Furthermore, although we can isolate, amplify, and label RNA from as few as five oocytes (2.5 ng of total RNA), we find that using between 10 and 20 oocytes or preimplantation embryos per each biological replicate is optimal (Pan et al., 2005). Experimental variation should be minimized by avoiding confounding factors, such as using different batches of culture medium, or using mice of different ages. In addition, the RNA isolation and amplification steps for all samples should be performed at the same time using the same reagents. Finally, the choice of microarray platform is another important experimental design consideration. Although there are several commercially available platforms available, we have used the Affymetrix MOE430 v2.0 GeneChips with mouse oocytes and preimplantation embryos (reviewed in Dalma-Weiszhausz et al., 2006). This platform covers most of the mouse genome, with 45,000 probe sets corresponding to more than 39,000 transcripts and variants, and likely contains most of the transcripts in mouse oocytes (Pan et al., 2005). Nevertheless, newer gene chips are now available, for example, the Affymetrix Mouse 1.0 ST gene chip. This gene chip represents 28,853 genes, in which each gene on the array is assayed by 27 probes spread across the entire length of the gene. This strategy provides a more complete and accurate picture of gene expression than 30 -based expression array designs.

464

Francesca E. Duncan and Richard M. Schultz

3. RNA Isolation and cRNA Target Preparation Protocol (Based on the Affymetrix Protocol for Eukaryotic Small Sample Target Labeling Assay Version II) This protocol is based on RNA isolation and amplification from 20 oocytes or preimplantation embryos. For a complete listing of equipment and reagents required for this protocol refer to Tables 23.2 and 23.3, respectively.

Table 23.2 Equipment needed for RNA isolation and cRNA target preparation

Equipment

42  C heat block Microcentrifuge Thermal cycler/PCR machine Vacuum centrifuge Refrigerated microcentrifuge 37  C water bath NanoDrop ND-1000 Spectrophotometer (Thermo Scientific, Wilmington, DE) 37  C heat block 70  C heat block 95  C heat block 94  C heat block 16  C heat block 65  C heat block  80  C freezer  20  C freezer Gel electrophoresis apparatus and power source Microwave UV lamp or gel documentation system 500-ml Erlenmeyer flask Filter tips 0.2-, 0.5-, and 1.5-ml RNase-, DNase-free tubes

Day Day Day Day Day 1 2 3 4 5

* * * *

*

* *

* * * *

* * * * *

* * * * * * *

*

* *

* *

* *

* *

* *

* *

* * * * *

Note: DNase- and RNase-free tubes and filter tips should be used for this protocol. We also recommend wiping all work surfaces and pipetmen with RNase ZAP wipes prior to use.

465

Gene Expression Profiling

Table 23.3 Reagents needed for RNA isolation and cRNA target preparation

Reagent

RNase ZAP wipes

Source and Catalog Number

Applied Biosystems/ Ambion, Austin, TX, #AM9786 Molecular Devices, PicoPure RNA Sunnyvale, CA, Isolation Kit # KIT0204 (Arcturus) T7 oligo(dT) promoter Affymetrix, Santa primer, 50 mM Clara, CA, #900375 DEPC–H2O (500 ml) Applied Biosystems/ Ambion, #AM9920 Invitrogen, Carlsbad, Superscript II CA, #18064-071 (200 U/ml); includes 5 First Strand Buffer and 0.1 M DTT 10 mM dNTP mix Invitrogen, #18427013 RNase inhibitor Applied Biosystems/ (40 U/ml) Ambion, #AM2684 5 Second Strand Invitrogen, #10812Buffer 014 E. coli DNA ligase Invitrogen, #18052(10 U/ml) 019 E. coli DNA polymerase Invitrogen, #18010I (10 U/ml) 025 T4 DNA polymerase Invitrogen, #18005(5 U/ml) 025 Glycogen (5 mg/ml) Applied Biosystems/ Ambion, #AM9510 5 M ammonium acetate Applied Biosystems/ Ambion, #AM9070G Absolute ethanol Pharmco-AAPER #111ACS200 MEGAscript T7 kit Applied Biosystems/ Ambion, #AM1334

Day Day Day Day Day 1 2 3 4 5

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

* *

*

*

*

*

*

*

*

*

*

*

*

*

(continued)

466

Francesca E. Duncan and Richard M. Schultz

Table 23.3 (continued)

Reagent

Qiagen RNeasy Mini Kit Random primers, 3 mg/ml RNase H (2 U/ml) Enzo BioArray High Yield RNA Transcript Labeling Kit (T7) 5 Fragmentation Buffer Ultrapure agarose

Source and Catalog Number

Qiagen, Valencia, CA, #74104 Invitrogen, #48190011 Invitrogen, #18021071 Enzo Life Sciences, Plymouth Meeting, PA #ENZ-42655-40 Affymetrix, #900371

Invitrogen, #16500500 Applied Biosystems/ NorthernMax Ambion, 10 Denaturing #AM8676 Gel Buffer Applied Biosystems/ NorthernMax 10 Ambion, MOPS Gel Running #AM8671 Buffer MilliQ-H2O Millipore, Billerica, MA RNA Millennium Applied Biosystems/ Markers Ambion, #AM7150 Gel Loading Buffer Applied Biosystems/ Ambion, #AM8546G Ethidium bromide BioRad, Hercules, solution (10 mg/ml) CA, #161-0433

Day Day Day Day Day 1 2 3 4 5

*

* * * *

* * *

*

* *

*

*

Note: DNase- and RNase-free tubes and filter tips should be used for this protocol. We also recommend wiping all work surfaces and pipetmen with RNase ZAP wipes prior to use.

3.1. Isolation of total RNA and synthesis of double-stranded cDNA (day 1) 3.1.1. Collection of samples 1. Transfer samples for RNA isolation in less than a 1-ml volume into a 0.5ml tube containing 10 ml of Extraction Buffer from the PicoPure RNA Isolation Kit.

Gene Expression Profiling

467

2. Use the samples immediately for RNA extraction or snap freeze them in an ethanol/dry ice bath and store them until use at 80  C. 3.1.2. RNA extraction and isolation Use the PicoPure RNA Isolation Kit according to manufacturer’s instructions. In brief: 1. Thaw cell extracts on ice and incubate at 42  C for 30 min.1 2. Precondition the RNA purification columns: (a) Pipet 250 ml Conditioning Buffer to the purification columns and incubate for 5 min at room temperature (RT) (b) Centrifuge at 16,000g for 1 min 3. Pipet 10 ml of 70% ethanol into the cell extracts and mix well by pipetting (do not centrifuge). 4. Add the cell extract/ethanol mixtures to the purification columns and centrifuge at 94g for 2 min and then at 16,400g for 30 s.2 5. Add 100 ml Wash Buffer 1 to the purification columns and centrifuge at 11,400g for 1 min. 6. Add 100 ml Wash Buffer 2 to the purification columns and centrifuge at 11,400g for 1 min. 7. Add 100 ml Wash Buffer 2 to the purification columns and centrifuge at 16,400g for 2 min. 8. Remove the flow through and centrifuge at 16,400g for 1 min. 9. Transfer the purification columns to elution tubes and pipet 11 ml Elution Buffer directly onto the purification column membranes. 10. Incubate the Elution Buffer on the purification columns for 5 min at RT. 11. Centrifuge for 1 min at 1,200g and then for 1 min at 16,400g. 3.1.3. First Strand synthesis 1. Reduce the volume of the eluted total RNA to 4 ml using a vacuum centrifuge with low heat (30  C takes 5–10 min). 2. In a 0.2-ml tube, mix 4 ml RNA with 1 ml T7 oligo(dT) promoter primer (diluted to 5 mM with DEPC–H2O). Incubate the samples at 70  C for 6 min.3 3. Cool samples to 4  C for 2 min and centrifuge briefly to collect the samples. 1 2 3

If a 42  C heat block is not available, transfer samples to 0.2-ml tubes and use a PCR machine set at 42  C with a heated lid. It is only necessary to remove the flow through from the purification columns when they are full, not after each spin. Use the following program on a thermal cycler with a heated lid: 70  C—6 min; 4  C—pause; 42  C—60 min; 70  C—10 min; 4  C—pause.

468

Francesca E. Duncan and Richard M. Schultz

4. Prepare the First Strand master mix as follows: 1 First Strand master mix #1 2 ml 5 First Strand Buffer 1 ml 0.1 M DTT 0.5 ml 10 mM dNTP 0.5 ml RNase inhibitor (40 U/ml) 1 ml Superscript II (200 U/ml) (5 ml total volume) 5. Add 5 ml of the master mix to each sample (total volume will now be 10 ml). 6. Mix components gently without vortexing, centrifuge briefly, and incubate at 42  C for 1 h. 7. Heat the samples to 70  C for 10 min to inactivate the Superscript. 8. Cool to 4  C and centrifuge briefly.

3.1.4. Second Strand synthesis 1. Prepare the Second Strand master mix as follows: 1 Second Strand master mix #1 45 ml DEPC–H2O 15 ml 5 Second Strand Buffer 1.5 ml 10 mM dNTP 0.5 ml E. coli DNA ligase (10 U/ml) 2 ml E. coli DNA polymerase I (10 U/ml) 0.5 ml RNase H (2 U/ml) (64.5 ml total volume) 2. Add 64.5 ml of the Second Strand master mix to each First Strand reaction. 3. Mix by pipetting and centrifuge briefly. Incubate at 16  C for 2 h.4 4. Add 1 ml of T4 DNA polymerase (5 U/ml) to each reaction and mix components gently without vortexing. 5. Incubate at 16  C for 10 min.

3.1.5. Ethanol precipitation 1. Transfer the samples to 1.5 ml tubes. 2. To each sample add the following: 40 ml DEPC–H2O, 2 ml glycogen (5 mg/ml), 72 ml 5 M ammonium acetate, and 480 ml of cold 100% ethanol. 3. Mix and precipitate samples overnight at 20  C. 4

Use the following program on a thermal cycler without a heated lid: 16  C—120 min; 4  C—pause; 16  C— 10 min; 4  C—pause.

Gene Expression Profiling

469

3.2. Recovery of double-stranded cDNA and first cycle of amplification (day 2) 3.2.1. Recovery of double-stranded cDNA 1. Centrifuge samples at 18,400g for 30 min at 4  C (a small white pellet the size of a pinhead should be observed). 2. Remove the supernatant and wash the pellet with 800 ml of cold 70% ethanol (diluted with DEPC–H2O). 3. Centrifuge the samples at 18,400g for 10 min at 4  C. 4. Remove the ethanol and dry pellet in a vacuum centrifuge without heat. Note: Do not over dry the samples. Check them every 2–3 min until there is about 0.5–1 ml volume remaining (this will take 10 min). 3.2.2. First cycle in vitro transcription 1. Prepare the in vitro transcription master mix at RT using reagents from the MEGAscript T7 kit as follows: 1 in vitro transcription master mix 4 ml DEPC–H2O 4 ml premixed NTPs (18.75 mM each) 1 ml 10 Reaction Buffer 1 ml 10 enzyme mix (10 ml total volume) 2. Add 10 ml of the master mix directly to each dried cDNA pellet and pipet up and down to mix. 3. Centrifuge briefly to collect the samples and incubate them for 6 h in a 37  C water bath. Tap samples gently to mix and centrifuge briefly every hour. 3.2.3. Purification of first cycle amplified cRNA Use the RNeasy Mini clean up kit according to manufacturer’s instructions. In brief: Add 90 ml RNase-free water to each cRNA sample. Add 350 ml Buffer RLT and mix well. Add 250 ml 100% ethanol and mix well (do not centrifuge samples). Transfer samples to RNeasy Mini spin columns placed in collection tubes. Centrifuge for 15 s at 9400g. Discard the flow through. 5. Add 500 ml Buffer RPE and centrifuge for 15 s at 9400g. Discard the flow through 6. Add 500 ml Buffer RPE and centrifuge for 2 min at 9400g. 7. Place the spin columns into new 2 ml collection tubes and centrifuge at maximum speed for 1 min. 1. 2. 3. 4.

470

Francesca E. Duncan and Richard M. Schultz

8. Place the spin columns in clean collection tubes and elute the cRNA samples with 30 ml RNase-free water. Note: The water should be added directly to the membrane and the columns should be allowed to sit for 5–10 min at RT. Centrifuge the columns at 9400g for 1 min to elute cRNA. Repeat with an additional 20 ml RNase-free water. 9. Determine the yield and purity (A260:A280) of the cRNA by analyzing 1 ml of each sample on a NanoDrop ND-1000 Spectrophotometer. 10. Reduce the volume of the eluted cRNA to 4 ml using a vacuum centrifuge with low heat (30  C takes 25 min). 11. cRNA can be stored overnight at 20  C. Wrap the top of each tube with parafilm.

3.3. Synthesis of double-stranded cDNA from first round of amplified cRNA (day 3) 3.3.1. First Strand synthesis Note: A maximum of 500 ng of amplified cRNA should be used for the second cycle of amplification. 1. To each tube with cRNA, add 1 ml random primers (diluted to 0.2 mg/ml with DEPC–H2O), mix, and centrifuge briefly to collect the samples. 2. Incubate samples at 70  C for 10 min.5 3. Cool samples to 4  C for 2 min and centrifuge briefly to collect the samples. 4. Prepare the First Strand master mix as follows: 1 First Strand master mix #2 2 ml 5 First Strand Buffer 1 ml 0.1 M DTT 0.5 ml 10 mM dNTP 0.5 ml RNase inhibitor (40 U/ml) 1 ml Superscript II (200 U/ml) (5 ml total volume) 5. Add 5 ml of the master mix to each sample (total volume will now be 10 ml). 6. Mix components gently without vortexing, centrifuge briefly, and incubate at 42  C for 1 h. 7. Add 1 ml of RNase H (2 U/ml) to each sample and incubate for 20 min at 37  C. 8. Heat samples at 95  C for 5 min to inactivate RNase H. 9. Chill samples on ice for 2 min and centrifuge briefly to collect the samples. 5

Use heat blocks set to 70, 42, 37, and 95  C. Alternatively, cRNA samples can be transferred to 0.2-ml tubes and reactions can be performed in a thermal cycler with a heated lid using the following program: 70  C—10 min; 4  C—pause; 42  C—60 min; 4  C—pause (add RNase H); 37  C—20 min; 95  C—5 min; 4  C—pause.

Gene Expression Profiling

471

3.3.2. Second Strand synthesis 1. To each sample, add 2 ml T7 oligo(dT) promoter primer (diluted to 5 mM with DEPC–H2O), mix, and centrifuge briefly to collect the samples. 2. Incubate the samples at 70  C for 6 min, chill on ice briefly, and centrifuge briefly.6 3. Prepare the Second Strand master mix as follows: 1 Second Strand master mix #2 43.5 ml DEPC 15 ml 5 Second Strand Buffer 1.5 ml 10 mM dNTP 2 ml E. coli DNA polymerase I (10 U/ml) (62 ml total volume) 4. Add 62 ml of the Second Strand master mix to each First Strand reaction (total volume will now be 75 ml). 5. Mix, centrifuge briefly, and incubate at 16  C for 2 h. 6. Add 2 ml T4 DNA polymerase (5 U/ml) to each sample and mix thoroughly. 7. Incubate samples at 16  C for 10 min.

3.3.3. Ethanol precipitation 1. Transfer the samples to 1.5 ml tubes. 2. To each sample add the following: 40 ml DEPC–H2O, 2 ml glycogen (5 mg/ml), 72 ml 5 M ammonium acetate, and 480 ml of cold 100% ethanol. 3. Mix and precipitate samples overnight at 20  C.

3.4. Recovery of double-stranded cDNA, second cycle of amplification, and cRNA labeling (day 4) 3.4.1. Recovery of double-stranded cDNA 1. Centrifuge samples at 18,400g for 40 min at 4  C. 2. Remove the supernatant and wash the pellet with 800 ml of cold 70% ethanol (diluted with DEPC–H2O). 3. Centrifuge the samples at 18,400g for 10 min at 4  C. 4. Remove the ethanol and dry pellet in a vacuum centrifuge without heat. Note: Do not over dry the samples. Check them every 2–3 min until there is about 0.5–1 ml volume remaining (this will take 10 min). A pellet should be observed. If not, it may be on the side of the tube. 6

Use 16 and 70  C heat blocks or a thermal cycler with heated lid for 70  C.

472

Francesca E. Duncan and Richard M. Schultz

3.4.2. Second cycle in vitro transcription and labeling Use the Enzo BioArray High Yield T7 RNA Transcript Labeling kit according to manufacturer’s instructions. In brief: 1. Prepare the master mix for in vitro transcription at RT as follows: 1 in vitro transcription and labeling master mix 22 ml DEPC–H2O 4 ml HY Reaction Buffer 4 ml 10 biotin-labeled ribonucleotides 4 ml 10 DTT 4 ml 10 RNase inhibitor mix 2 ml 20 T7 RNA polymerase (40 ml total volume) 2. Add 40 ml of the master mix to each pellet, mix (vortex briefly to make sure liquid coats the sides), centrifuge briefly, and incubate in a 37  C water bath for 5 h. Mix and centrifuge briefly every hour to collect the samples. 3.4.3. Purification of second cycle amplified and labeled cRNA 1. Add 60 ml RNase-free water to each cRNA sample. 2. Purify the cRNA using the Qiagen RNeasy Mini Kit according to the protocol as written in Section 3.2.3, Steps 2–8. 3. Determine the yield and purity (A260:A280) of the cRNA by analyzing 1 ml of each sample on the NanoDrop ND-1000 Spectrophotometer. 4. cRNA can be stored at 80  C until fragmentation. Wrap the top of each tube with parafilm.

3.5. cRNA fragmentation and assessment by gel electrophoresis (day 5) 3.5.1. cRNA fragmentation 1. Assemble the fragmentation reactions as follows: 22 mg cRNA 8.8 ml 5 Fragmentation Buffer DEPC–H2O to 44 ml The concentration of fragmented cRNA will be 0.5 mg/ml 2. Incubate samples at 94  C for 35 min and cool on ice briefly.7 3. Set aside 4 ml (2 mg) of each fragmented sample to run on an agarose gel alongside unfragmented samples. 4. Wrap tubes containing fragmented cRNA with parafilm and store at 80  C until hybridization. 7

Use 94  C heat block.

Gene Expression Profiling

473

3.5.2. Gel electrophoresis of fragmented cRNA 1. Pour a 1.5% formaldehyde gel: (a) Mix 3 g agarose þ 180 ml DEPC–H2O in a 500-ml Erlenmeyer flask and heat in a microwave until the agarose is completely dissolved. Note: Stop the microwave every  20–30 s and swirl the solution to prevent boiling over. (b) While the agarose solution is still hot, add 20 ml NorthernMax 10 Denaturing Gel Buffer to the agarose solution, swirl to mix well, and pour directly into a gel casting mold. Allow the gel to dry under a hood. Note: Do not pour gels too thick ( 0.5 cm). 2. Make 1 Running Buffer: Mix 1350 ml MilliQ-H2O þ 150 ml NorthernMax 10 MOPS Gel Running Buffer 3. Prepare the samples for the formaldehyde gel: Sample 2 mg cRNA (fragmented or unfragmented) 3 ml Gel Loading Buffer þ 50 mg/ml ethidium bromide8 DEPC–H2O to 9 ml MW marker 4 ml Millenium Markers 3 ml Gel Loading Buffer þ 50 mg/ml ethidium bromide8 2 ml DEPC–H2O 4. Heat the samples at 65  C for 10 min and cool to RT. 5. Centrifuge briefly to collect the samples and load the entire volume on the gel. 6. Run the gel in 1 Running Buffer at 50 V until the dye front is 2/3 of the way down. 7. Image the RNA using UV light or a gel documentation system.

4. Quality Control Assessment of cRNA At several points during the isolation, amplification, and labeling procedures the quantity and quality of the material can be gauged. Following the first and second amplifications, the cRNA quantity can be measured by absorbance at 260 nm. Average cRNA yields following each amplification step for experiments involving different numbers of oocytes or preimplantation embryos are reported in Table 23.1. The absorbance ratio at 260 nm:280 nm is an indicator of cRNA quality and should be between 2.0 8

Make a 1:200 dilution of the 10 mg/ml ethidium bromide stock in the Gel Loading Buffer for a final concentration of 50 mg/ml.

474

Francesca E. Duncan and Richard M. Schultz

Figure 23.2 Denaturing gel electrophoresis of samples before and after metal-induced hydrolysis. Samples of amplified and biotinylated cRNA before (unfragmented) and after (fragmented) metal-induced hydrolysis were analyzed on a denaturing, formaldehyde gel (2 mg cRNA/lane). Unfragmented cRNA typically appears as a smear ranging in size from 300 to 1500 bp and fragmented cRNA is approximately 35–200 bp.

and 2.4 following the second amplification. The average sizes of the amplified products before and after fragmentation are also good indicators of cRNA quality. Prior to fragmentation, the amplified and labeled cRNA will appear as a smear ranging from approximately 300 to 1500 bp on a denaturing agarose gel and following fragmentation, the products will be between 35 and 200 bp (Fig. 23.2).

5. Methods for Microarray Data Analysis Following hybridization and data acquisition, microarray data analysis occurs in two steps. First the raw scanned image data are processed and then the processed data are analyzed using statistical methods. DAT files, comprised of pixel values, are the raw optical images of the hybridized chip. DAT files are processed into CEL files that contain the intensity values associated with the probesets (Dalma-Weiszhausz et al., 2006). The intensity values stored in the CEL files are quantified and normalized using algorithms, such as MAS5, RMA, or GCRMA (Lim et al., 2007). These algorithms transform the data into biologically relevant gene expression values, stored as CHP files, which can then be used for analysis. Standard array quality control parameters should be examined to determine the overall sample quality. These parameters include the Noise (Raw Q), Background, and Scaling Factor. In addition, the Percent Present Call refers to how many of the probe sets were detected and

Gene Expression Profiling

475

should be approximately 40% or greater. Probes on the Affymetrix Gene Chips are designed to hybridize to the 30 end of transcripts. However, probe sets corresponding to the 30 , middle, and 50 regions of specific maintenance genes, such as GAPDH and actin, are typically included as controls to monitor the target cRNA quality and to ensure that skewing to the 30 end did not occur during amplification steps. The 30 –50 Signal Intensity Ratio for these controls should be less than 10. Statistical analysis of microarray data is usually done using specialized software, such as GeneSpring GX 10 (Agilent Technologies, Placerville, CA) or Partek Genomics Suite (St. Louis, MO), to generate lists of genes with statistically significant differences in expression. Significant foldchange differences between samples are calculated typically using a 5% false discovery rate (FDR). The gene lists can then be analyzed to determine the relationships between experimental samples. For example, Principle Components Analysis is an unsupervised clustering method and visualization tool that distributes samples into a two-dimensional space based on the variance in gene expression (Joliffe and Morgan, 1992). Thus, experimental samples with similar trends in gene expression profiles will cluster together. Hierarchical clustering analysis can also be used to assess how similar replicate samples across all experimental groups are in terms of gene expression. In this type of analysis, a heat-map with horizontal bars representing each transcript is generated, and the bar color corresponds to the relative abundance of each transcript. There are also many ways to elucidate the biological meaning of microarray data. For example, Expression Analysis Systematic Explorer (EASE) is a program that uses statistical methods to test for overrepresentation of gene ontology classifications in gene lists using previously published annotation databases (Hosack et al., 2003). This overrepresentation means that, irrespective of gene expression levels, genes with a specific similar biological function (i.e., cytoskeleton components or transcription factors) are enriched in a given gene list compared to a normal distribution of all the genes assayed. EASE can, therefore, be used to identify biological themes within gene lists. The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database (www.genome.jp/kegg/pathway.html) as well as software such as Ingenuity Pathway Analysis (Ingenuity, Redwood City, CA) can be used to classify gene lists based on molecular interactions, pathways, and reactions. Finally, the Database for Annotation, Visualization and Integrated Discovery (DAVID) is a useful resource consisting of a large set of functional annotation tools geared to uncover the biological meanings behind large gene lists (http://david.abcc. ncifcrf.gov) (Dennis et al., 2003). The reader should refer to the manuscript by Huang for a step-by-step protocol for how to use DAVID (Huang da et al., 2009).

476

Francesca E. Duncan and Richard M. Schultz

6. Validation Methods Data validation must be performed as a final step of any microarray experiment. Significant gene expression differences should be confirmed by an independent method such as quantitative Real Time PCR. The biological validity of the data can be assessed by examining the expression levels of various control transcripts. For example, if the experimental samples are denuded oocytes, somatic cell-specific genes should not be expressed (e.g., Fst, Bmp2, Bmp4, Bmp7). Likewise, if the experimental samples are derived from oocytes or preimplantation embryos harboring a gene-specific knockout or knockdown, transcripts for that gene should be absent or dramatically lower compared to control samples, respectively. Transcript profiling only provides a measure of RNA abundance that can be affected by levels of transcription as well as levels of RNA processing or degradation. Therefore, the transcript level of a given gene does not necessarily correspond to its protein level, modification, activity, or localization. Follow up studies, using either immunocytochemistry and/or immunoblot analysis, should be performed to characterize genes-of-interest at the protein level. Finally, to place the results in a meaningful biological context, functional assays using, for example, overexpression, dominant negative, or RNA interference approaches should be performed.

7. Archiving Results At the time of publication, microarray data should be submitted to a public repository that accepts, holds, and distributes microarray data compliant with Minimal Information About a Microarray Experiment (MIAME) standards (Brazma et al., 2001) (Table 23.4). The three most used repositories are: Gene Expression Omnibus (GEO) at the National Center for Biotechnology Information (USA) (www.ncbi.nlm.nih.gov/geo), ArrayExpress at the European Bioinformatics Institute (UK) (www.ebi.ac.uk/microarray-as/ae), and Center for Information Biology Gene Expression Database (CIBEX) at the DNA Data Bank of Japan ( Japan) (www.cibex.nig.ac.jp).

8. Future Directions The future in genomics research is to combine single-cell whole genome amplification with transcriptome profiling using next generation deepsequencing methods. In fact, this technology has been applied successfully to single mouse blastomeres and oocytes (Tang et al., 2009). RNA sequencing

477

Gene Expression Profiling

Table 23.4 Six key elements of MIAME (Brazma et al., 2001) Sl. MIAME no. element

(1) Experimental design

Relevant information to be included in submission  Type of experiment (normal vs. diseased comparison,   

(2) Array design



(3) Samples





(4) Hybridizations 

(5) Measurements



(6) Normalization controls



wild type vs. knockout, time course, dose response, etc.) Experimental variables examined (time, dose, genetic variation, response to a treatment or compound, etc.) Types of replicates and controls used Sample data relationships (i.e., specification of which raw data files correlate to which experimental samples) A systematic definition of the arrays used in the experiment (gene identifiers, genomic coordinates, probe sequences, or catalog number of a commercially available array platform) Description of the biological sample including the source (organism taxonomy, cell type, etc.) and treatments applied to the sample Laboratory protocols used for the technical extraction and labeling of the nucleic acids Description of the conditions under which the sample-array hybridizations were performed (hybridization solution, type of blocking agent, wash procedure, quantity of labeled target used, etc.) Experimental results progressing from raw to processed data including (1) the original image scans of the array, (2) the microarray quantification matrices based on analysis of the raw data, and (3) the final gene expression matrix following normalization Explanation of normalization strategy employed and algorithms used

(RNA-Seq) is a high-throughput sequence-based approach that directly determines cDNA sequence with single-base resolution and allows both the mapping and quantification of entire genomes (reviewed in Wang et al., 2009). Using the mRNA-Seq assay on individual mouse blastomeres, the expression of 75% more genes were detected compared to microarray techniques and previously unknown transcript variants were identified (Tang et al., 2009). Deep-sequencing technologies, such as RNA-Seq, are advantageous for several reasons. First, they do not rely on preexisting genomic sequence for transcript detection. Second, in addition to providing information about

478

Francesca E. Duncan and Richard M. Schultz

transcript abundance, these methods provide important information about transcript structure including, transcription boundaries and single nucleotide polymorphisms. Finally, transcript expression levels separated by up to five orders of magnitude can be detected accurately in a single experiment (Wang et al., 2009). Although deep-sequencing methods will ultimately replace microarray technology, there are several limitations, such as high cost and appreciable sequencing error rates, that make its widespread use impractical at this time. Furthermore, novel bioinformatics methods still need to be developed to store, retrieve, and analyze the large volumes of data that will be generated using this technology. In addition to deep-sequencing methods, direct single molecule RNA-Seq without prior conversion of RNA to cDNA was recently reported and is an exciting new frontier for probing the transcriptome (Ozsolak et al., 2009). It is simply a matter of time when these aforementioned approaches will become the method of choice to analyze gene expression. An additional future goal in transcriptomics will be to mine existing microarray data by using meta-analytic methods to integrate gene expression profiles from different published experiments (Choi et al., 2003; RodriguezZas et al., 2008a,b). In 2004, three laboratories independently examined the gene expression profiles of mouse preimplantation embryos using three different microarray platforms: Affymetrix MOE430, Affymetrix U74, and NIA 22K (Hamatani et al., 2004; Wang et al., 2004; Zeng et al., 2004). The results from these published experiments were merged to identify candidate maternal-effect genes, which were defined as those genes that were predominantly expressed in metaphase II-arrested eggs and one-cell embryos but not during later stages of preimplantation embryo development (Mager et al., 2006). This combined microarray analysis uncovered 37 putative maternaleffect genes of which 21 were previously not known to function during early development. These results highlight the power of in silico analysis of existing microarray data to provide novel biological insights. In conclusion, although gene expression profiling will likely be replaced by more advanced technologies, it currently remains the most reliable, robust, and cost-effective manner of analyzing the transcriptome of oocytes and preimplantation embryos.

ACKNOWLEDGMENTS F. E. D. thanks Paula Stein and Karen Schindler for critically reading the manuscript, Hua Pan for technical advice, and John Tobias for help with bioinformatics and microarray analysis. The authors also would like to acknowledge MDS Analytical Technologies (Sunnyvale, CA), Qiagen (Valencia, CA), and Affymetrix (Santa Clara, CA) for allowing the use of their modified protocols in this chapter. Arcturus and PicoPure are registered trademarks of MDS Analytical Technologies (US) Inc. Research described in this manuscript was supported by grants from the NIH (HD22681 and HD22732) to R. M. S. and F. E.D. was supported by an NIH training grant (T32 HD 007305).

Gene Expression Profiling

479

REFERENCES Brazma, A., Hingamp, P., Quackenbush, J., Sherlock, G., Spellman, P., Stoeckert, C., Aach, J., Ansorge, W., Ball, C. A., Causton, H. C., Gaasterland, T., Glenisson, P., et al. (2001). Minimum information about a microarray experiment (MIAME)-toward standards for microarray data. Nat. Genet. 29, 365–371. Bultman, S. J., Gebuhr, T. C., Pan, H., Svoboda, P., Schultz, R. M., and Magnuson, T. (2006). Maternal BRG1 regulates zygotic genome activation in the mouse. Genes Dev. 20, 1744–1754. Choi, J. K., Yu, U., Kim, S., and Yoo, O. J. (2003). Combining multiple microarray studies and modeling interstudy variation. Bioinformatics 19(Suppl. 1), i84–i90. Dalma-Weiszhausz, D. D., Warrington, J., Tanimoto, E. Y., and Miyada, C. G. (2006). The affymetrix GeneChip platform: An overview. Methods Enzymol. 410, 3–28. Dennis, G., Jr., Sherman, B. T., Hosack, D. A., Yang, J., Gao, W., Lane, H. C., and Lempicki, R. A. (2003). DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol. 4, P3. Duncan, F. E., Stein, P., Williams, C. J., and Schultz, R. M. (2009). The effect of blastomere biopsy on preimplantation mouse embryo development and global gene expression. Fertil. Steril. 91, 1462–1465. Hamatani, T., Carter, M. G., Sharov, A. A., and Ko, M. S. (2004). Dynamics of global gene expression changes during mouse preimplantation development. Dev. Cell 6, 117–131. Hamatani, T., Ko, M., Yamada, M., Kuji, N., Mizusawa, Y., Shoji, M., Hada, T., Asada, H., Maruyama, T., and Yoshimura, Y. (2006). Global gene expression profiling of preimplantation embryos. Hum. Cell 19, 98–117. Hamatani, T., Yamada, M., Akutsu, H., Kuji, N., Mochimaru, Y., Takano, M., Toyoda, M., Miyado, K., Umezawa, A., and Yoshimura, Y. (2008). What can we learn from gene expression profiling of mouse oocytes? Reproduction 135, 581–592. Hosack, D. A., Dennis, G., Jr., Sherman, B. T., Lane, H. C., and Lempicki, R. A. (2003). Identifying biological themes within lists of genes with EASE. Genome Biol. 4, R70. Huang da, W., Sherman, B. T., and Lempicki, R. A. (2009). Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat. Protoc. 4, 44–57. Joliffe, I. T., and Morgan, B. J. (1992). Principal component analysis and exploratory factor analysis. Stat. Methods Med. Res. 1, 69–95. Lim, W. K., Wang, K., Lefebvre, C., and Califano, A. (2007). Comparative analysis of microarray normalization procedures: Effects on reverse engineering gene networks. Bioinformatics 23, i282–i288. Ma, J., Zeng, F., Schultz, R. M., and Tseng, H. (2006). Basonuclin: A novel mammalian maternal-effect gene. Development 133. Ma, J., Flemr, M., Stein, P., Berninger, P., Malik, R., Zavolan, M., Svoboda, P., and Schultz, R. M. (2010). microRNA activity is suppressed in mouse oocytes. Curr. Biol. 9, 265–270. Mager, J., Schultz, R. M., Brunk, B. P., and Bartolomei, M. S. (2006). Identification of candidate maternal-effect genes through comparison of multiple microarray data sets. Mamm. Genome 17, 941–949. Murchison, E. P., Stein, P., Xuan, Z., Pan, H., Zhang, M. Q., Schultz, R. M., and Hannon, G. J. (2007). Critical roles for Dicer in the female germline. Genes Dev. 21, 682–693. Ozil, J. P., Banrezes, B., Toth, S., Pan, H., and Schultz, R. M. (2006). Ca2þ oscillatory pattern in fertilized mouse eggs affects gene expression and development to term. Dev. Biol. 300, 534–544.

480

Francesca E. Duncan and Richard M. Schultz

Ozsolak, F., Platt, A. R., Jones, D. R., Reifenberger, J. G., Sass, L. E., McInerney, P., Thompson, J. F., Bowers, J., Jarosz, M., and Milos, P. M. (2009). Direct RNA sequencing. Nature 461, 814–818. Pan, H., OBrien, M. J., Wigglesworth, K., Eppig, J. J., and Schultz, R. M. (2005). Transcript profiling during mouse oocyte development and the effect of gonadotropin priming and development in vitro. Dev. Biol. 286, 493–506. Pan, H., Ma, P., Zhu, W., and Schultz, R. M. (2008). Age-associated increase in aneuploidy and changes in gene expression in mouse eggs. Dev. Biol. 316, 397–407. Rinaudo, P., and Schultz, R. M. (2004). Effects of embryo culture on global pattern of gene expression in preimplantation mouse embryos. Reproduction 128, 301–311. Rinaudo, P. F., Giritharan, G., Talbi, S., Dobson, A. T., and Schultz, R. M. (2006). Effects of oxygen tension on gene expression in preimplantation mouse embryos. Fertil. Steril. 86, 1252–1265, 1265e1–1336e1. Rodriguez-Zas, S. L., Ko, Y., Adams, H. A., and Southey, B. R. (2008a). Advancing the understanding of the embryo transcriptome co-regulation using meta-, functional, and gene network analysis tools. Reproduction 135, 213–224. Rodriguez-Zas, S. L., Schellander, K., and Lewin, H. A. (2008b). Biological interpretations of transcriptomic profiles in mammalian oocytes and embryos. Reproduction 135, 129–139. Schmutz, J., Wheeler, J., Grimwood, J., Dickson, M., Yang, J., Caoile, C., Bajorek, E., Black, S., Chan, Y. M., Denys, M., Escobar, J., Flowers, D., et al. (2004). Quality assessment of the human genome sequence. Nature 429, 365–368. Sotiriou, C., and Piccart, M. J. (2007). Taking gene-expression profiling to the clinic: When will molecular signatures become relevant to patient care? Nat. Rev. Cancer 7, 545–553. Stein, P., Zeng, F., Pan, H., and Schultz, R. M. (2005). Absence of non-specific effects of RNA interference triggered by long double-stranded RNA in mouse oocytes. Dev. Biol. 286, 464–471. Tang, F., Barbacioru, C., Wang, Y., Nordman, E., Lee, C., Xu, N., Wang, X., Bodeau, J., Tuch, B. B., Siddiqui, A., Lao, K., and Surani, M. A. (2009). mRNA-Seq wholetranscriptome analysis of a single cell. Nat. Methods 6, 377–382. Van Gelder, R. N., von Zastrow, M. E., Yool, A., Dement, W. C., Barchas, J. D., and Eberwine, J. H. (1990). Amplified RNA synthesized from limited quantities of heterogeneous cDNA. Proc. Natl. Acad. Sci. USA 87, 1663–1667. Wan, L. B., Pan, H., Hannenhalli, S., Cheng, Y., Ma, J., Fedoriw, A., Lobanenkov, V., Latham, K. E., Schultz, R. M., and Bartolomei, M. S. (2008). Maternal depletion of CTCF reveals multiple functions during oocyte and preimplantation embryo development. Development 135, 2729–2738. Wang, Q. T., Piotrowska, K., Ciemerych, M. A., Milenkovic, L., Scott, M. P., Davis, R. W., and Zernicka-Goetz, M. (2004). A genome-wide study of gene activity reveals developmental signaling pathways in the preimplantation mouse embryo. Dev. Cell 6, 133–144. Wang, Z., Gerstein, M., and Snyder, M. (2009). RNA-Seq: A revolutionary tool for transcriptomics. Nat. Rev. Genet. 10, 57–63. White, C. A., and Salamonsen, L. A. (2005). A guide to issues in microarray analysis: Application to endometrial biology. Reproduction 130, 1–13. Zeng, F., and Schultz, R. M. (2005). RNA transcript profiling during zygotic genome activation in the preimplantation embryo. Dev. Biol. 283, 40–57. Zeng, F., Baldwin, D. A., and Schultz, R. M. (2004). Transcript profiling during preimplantation mouse development. Dev. Biol. 272, 483–496.

C H A P T E R

T W E N T Y- F O U R

Interrogating the Transcriptome of Oocytes and Preimplantation Embryos Anne E. Peaston,* Joel H. Graber,* Barbara B. Knowles,† and Wilhelmine N. de Vries* Contents 1. Introduction 1.1. Biochemical analyses revealed significant decrease in total RNA quantity during oogenesis and early embryogenesis 1.2. Molecular studies demonstrated sequence control of mRNA fate 1.3. Large libraries of expressed sequences identified genes controlling oocyte maturation and preimplantation embryo development 2. RNA Relative Quantification in Oocytes and Preimplantation Embryos 2.1. Inappropriate normalization results in artifactual increases in apparent transcript abundance during transcriptional silence 2.2. Normalization to an exogenous control eliminates apparent expression increase during transcriptional silence 2.3. Choices in both computational and experimental normalization can lead to artifacts 3. Comparative Analysis of Large-Scale Expression Analyses 3.1. Comparison of results between large-scale experiments is complicated by differences in both experimental and computational approaches 3.2. Comparison of FGO and OO sequencing datasets 3.3. Are the transcripts identified in FGO but absent from OO biologically meaningful?

482 483 485

486 487 487 490 493 495

495 496 497

* The Jackson Laboratory, Bar Harbor, Maine, USA Institute of Medical Biology, A*STAR, Immunos, Singapore

{

Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77024-5

#

2010 Elsevier Inc. All rights reserved.

481

482

Anne E. Peaston et al.

4. Taking Isoform-Specific Changes into Account 5. Concluding Remarks Acknowledgments References

500 502 503 506

Abstract During its growth phase, a mouse oocyte accumulates RNA that is the sole template for new protein synthesis in the transcriptionally silent interval between growth completion and transcriptional activation of the embryonic genome. Over this transcriptionally silent interval, almost half the quantity of RNA accumulated in the full-grown oocyte is degraded, while stable messages undergo major transcript-specific polyadenylation fluctuations associated with timely translation of new proteins. These processes, in the background of substantial RNA degradation, create unique pitfalls for transcriptome analysis. Three particular challenges are discussed herein. (1) Systematic errors of relative quantification occur if standard approaches are used, wherein samples are normalized to a constant quantity of RNA, or when computational analyses are normalized to an apparent ‘‘constant’’ endogenous to the sample. We show that use of a fixed quantity of exogenous RNA per oocyte or embryo alleviates this problem. (2) Comparison of large-scale expression analyses from widely disparate platforms highlights how the differing protocols produce correspondingly different lists of genes with significant changes in transcript abundance. Only with careful attention to the differences among experiments can such discrepancies be understood. (3) The complete assessment of changes in expression requires correspondingly comprehensive assessment of the role of isoform-specific changes.

1. Introduction Formation of the new mammalian embryo from male and female gametes is the result of a naturally occurring nuclear reprogramming process that converts differentiated cells to totipotency. The transition from gamete to embryo requires oocyte maturation and completion of oocyte meiotic divisions, fertilization, nuclear reprogramming of both gametes, the first mitosis, and activation of the embryonic genome, in addition to metabolic housekeeping processes. Ongoing reprogramming accompanies the embryo’s development through the cleavage stages to its initial cellular differentiation and morphogenetic events starting in the morula and blastocyst. Understanding the molecular mechanisms driving events in this period in mice is important to reproductive science as it translates to human health. The unique nuclear reprogramming capability of the oocyte is of fundamental interest to the stem cell field, given that the oocyte is able to convert differentiated gametes (or somatic cell nuclei) into the totipotent cells of

Interrogating the Transcriptome of Oocytes and Preimplantation Embryos

483

the early cleavage-stage embryo, which can be viewed as the ultimate mammalian stem cells. Mouse gametes and preimplantation embryos are readily isolated at discrete stages of reprogramming and subsequent differentiation, and offer an easily accessible mammalian system in which to study these processes at the molecular and cellular level. However, obtaining sufficient material to characterize molecular events is challenging since only limited numbers of gametes and preimplantation embryos can be acquired from a single mouse. Here, we touch on some important historical data that established, in general terms, the dynamics of RNA biology in mouse oocytes and early embryos, and demonstrate and discuss some of the pitfalls of assaying the dynamics of gene expression in this milieu. This chapter is largely conceptual, and we hope will be useful as an aid to thinking about how to approach quantitative experiments. We have not attempted to include instructions for most of the assays discussed, and these may be found elsewhere in this volume or in other literature cited. However, since we include here previously unpublished qPCR data, we provide instructions for this procedure as an appendix at the end of the chapter.

1.1. Biochemical analyses revealed significant decrease in total RNA quantity during oogenesis and early embryogenesis To understand relative changes in gene expression in mouse oocytes and preimplantation embryos, it is important to appreciate quantitative aspects of RNA biology during this time in development (Table 24.1). The developing mouse oocyte synthesizes and stockpiles RNA until it is fully grown. Thereafter, apart from a minor quantity of new transcription in the one-cell embryo, there is no further RNA synthesis until after the first cleavage division about 48 h later (Clegg and Piko, 1983a,b; Piko and Clegg, 1982). This indicates that new protein synthesis required for oocyte progression through maturation, meiosis, fertilization, and the first embryonic cleavage division is dependent on maternal stores of RNA. Measurement of the RNA content in mouse full-grown oocytes (FGOs) and ovulated oocytes (OO) demonstrated that, during oocyte maturation, about 80 pg total RNA is degraded of which about 60 pg represents rRNA and about 20 pg represents polyadenlyated (polyA) mRNA. A further 30 pg undergoes substantial shortening or loss of long polyA tails (Paynton et al., 1988, and references therein). In the one-cell embryo, new polyadenylation of a similar quantity of mRNA suggested readenylation of much of the previously deadenylated mRNA component (Clegg and Piko, 1983b), although some deadenylated transcripts were degraded (Paynton et al., 1988). By the late two-cell embryo stage, after initiation of embryonic transcription, about 60% of maternally supplied total RNA is still present

Table 24.1 RNA content of oocytes and preimplantation embryos FGO

Total (pg) PolyA mRNA (pg) Poly A mRNA (molecules
107)h PolyA mRNA (% of total) a b c d e f g h i j k

Bachvarova et al. (1985). Sternlicht and Schultz (1981). Kaplan et al. (1982). Olds et al. (1973). Piko and Clegg (1982). Bachvarova (1974). Bachvarova (1985). Clegg and Piko (1983a). Brower et al. (1981). De Leon et al. (1983). Bachvarova and De Leon (1980).

OO a,b,c

430–600 50–90a,g – 19–20a,i,j

300–550 25–35a,g 1.7 10a,j,k

Zygote a,c,d,e,f

345–540 23e 2.4 –

Late two-cell d,e

236–390 – 0.7 –

d,e

8–16-cell

450–690 – 1.3 –

d,e

Blast (32 cells)

1370–1420d,e – 3.4 –

Interrogating the Transcriptome of Oocytes and Preimplantation Embryos

485

in the embryo, of which approximately 24 pg is maternally derived polyA RNA. In the early blastocyst, about 30% of maternally derived total RNA remains, at which point it contributes somewhat less than 10% of the total blastocyst RNA (Bachvarova and De Leon, 1980; Piko and Clegg, 1982). From these biochemical studies, it would appear that maternal messages may have the potential to play important roles throughout the preimplantation development. Later studies of individual genes exposed a complex picture, where RNA stability and translation were found to be sequence-specific and related to the processes of polyadenylation and deadenylation. For example, during oocyte maturation, some genes are deadenylated and subsequently degraded, while others follow deadenylation with subsequent cytoplasmic polyadenylation and translation (Gebauer et al., 1994; Huarte et al., 1987; Oh et al., 2000; Paynton et al., 1988; West et al., 1996). Two broad questions emerge from this work: how does RNA sequence determine maternal mRNA fate? and which maternal and embryonic transcripts are important for accomplishing oocyte growth and maturation, the oocyte to embryo transition, and preimplantation embryo development?

1.2. Molecular studies demonstrated sequence control of mRNA fate Stage-specific translation of new proteins during oocyte maturation and the oocyte to embryo transition is dependent on cytoplasmic polyadenylation of stored maternal messages. Deletion studies of individual mRNA 30 -UTRs, coupled with biochemical analyses of 30 -UTR binding proteins and measurements of protein synthesis, have been the mainstay for discovery and characterization of cis elements associated with mRNA fate. A proposed system regulating mRNA deadenylation and storage in the Xenopus laevis growing oocyte, and polyadenylation and translation during oocyte meiotic progression, depends on the number and position of specific motifs in mRNA 30 -UTRs, and interactions of their bound-specific protein complexes (Belloc et al., 2008). A very similar system is likely to operate in mice and other vertebrates, based on consideration of known shared mechanisms, expression in mouse and other vertebrate oocytes of orthologs of genes encoding the specific proteins involved, and correlations between mRNA 30 -UTR motifs and transcript stability, as briefly reviewed by Evsikov and Marin de Evsikova (2009b). However, the proposed X. laevis 30 -UTR code does not perfectly predict translational activation and repression effects of different human and mouse 30 -UTRs (Pique et al., 2008), suggesting that other 30 -UTR motifs and other factors such as mRNA secondary structure, which can modulate the effects of the known motifs, will have to be explored.

486

Anne E. Peaston et al.

Small noncoding RNAs also play an important part in transcript fate in mouse oocytes and early embryos. Oocyte deficiencies of Dicer1, or Eif2c2 (Ago2), which encode proteins critical for production of microRNAs and small-interfering RNAs, alter transposon and protein-coding transcript abundance (Murchison et al., 2007; Watanabe et al., 2008).

1.3. Large libraries of expressed sequences identified genes controlling oocyte maturation and preimplantation embryo development cDNA library construction and analysis lies at the heart of the last roughly two decades of effort to map, at a large scale, the changing RNA landscape of the oocyte and early preimplantation embryo. Construction of bacterially cloned primary cDNA libraries from ovulated oocytes and preimplantation embryos was described in this series in 1993 (Rothstein et al., 1993). Advances in automated sequencing made possible the array and sequencing of thousands of cloned transcripts per cDNA library, enabling genome-wide transcriptome surveys, large-scale molecular snapshots of stage-specific biochemical activities, and provided a window into evolutionary conservation of gene expression during that period (Evsikov and Marin de Evsikova, 2009a; Evsikov et al., 2004, 2006; Ko et al., 2000). Depending on the method of library construction, sequence representation in a library may directly reflect sequence abundance in the starting material. Direct comparison of quantitative sequence representation in large primary mouse nonamplified cDNA libraries has led to insight into molecules involved in transit from one stage to another (Evsikov et al., 2004, 2006; Peaston et al., 2004). Similarly, subtractive hybridization and related techniques have been used to identify and clone genes with stage-specific expression (e.g., Hwang et al., 1996, 1997, 1999; Oh et al., 1997; Rothstein et al., 1992; Zeng and Schultz, 2003). Collecting the quantities of oocytes or embryos required to extract the large amount of RNA needed for any global analysis of gene expression is daunting. However, collection of large numbers of oocytes and embryos can be avoided by amplifying the cDNA prepared from small samples. The two basic approaches to cDNA amplification are exponential amplification of the cDNA by polymerase chain reaction (PCR), and linear isothermal amplification by in vitro transcription of the cDNA (Brady and Iscove, 1993; Van Gelder et al., 1990). Lively discussion continues on the advantages and disadvantages of these methods and their numerous derivatives for particular experimental goals and situations (see, e.g., Baugh et al., 2001; Iscove et al., 2002; Ji et al., 2004; Klur et al., 2004; Lang et al., 2009; Subkhankulova and Livesey, 2006). High-throughput sequencing of tagged ‘‘molecular cDNA libraries’’ avoids the need for bacterial cDNA cloning and provides sequences from hundreds of thousands to millions of molecules. Alone or

Interrogating the Transcriptome of Oocytes and Preimplantation Embryos

487

in combination with sample amplification, this methodology has made possible large gains in cataloguing the transcriptomes of oocytes and embryos. For example, high-throughput sequencing of size-selected molecular cDNA libraries was important in discovery of small noncoding RNAs in oocytes (Tam et al., 2008; Watanabe et al., 2008). There are many variants to each of these methods, and choosing which method to use involves juggling multiple trade-offs between sensitivity and specificity of transcript detection, and various biases associated with sample collection and processing. Independent of the complications specific to oogenesis and early embryogenesis described here, major sources of bias include, but are not limited to, transcript or cDNA size selection, subtractive hybridization, and linear or exponential amplification of RNA or cDNA. cDNA sequencing techniques have produced a large quantity of data and are fertile territory for suggesting hypotheses regarding molecules and molecular mechanisms controlling oocyte maturation, the oocyte to embryo transition, and preimplantation development. Compared to hybridization-based assays of gene expression, the data produced is unbiased by preselection of primers or probes, and uniquely provides opportunity for discovery of previously unknown coding and noncoding transcripts and transcript variants. Bacterially cloned cDNA libraries provided the initial essential catalogs to guide the development of transcript-derived probesets for hybridization-based high-throughput global expression analyses, that is, microarray analysis of gene expression (Tanaka et al., 2000). Global transcriptome profiling by stage-specific microarray analysis of gene expression has vastly expanded our understanding of what genes are expressed during oocyte maturation, the oocyte-to-embryo transition, and preimplantation development (Hamatani et al., 2004; Wang et al., 2004; Zeng and Schultz, 2003). Different groups of genes display distinct patterns of temporally correlated gene expression, piquing curiosity as to how these patterns are established, and what is their function.

2. RNA Relative Quantification in Oocytes and Preimplantation Embryos 2.1. Inappropriate normalization results in artifactual increases in apparent transcript abundance during transcriptional silence A number of studies, including our own (Evsikov et al., 2004), have reported increased signals corresponding to certain genes in OOs and one-cell embryos (zygotes) compared with FGO when assaying developmentally regulated gene expression. Given the transcriptional silence at this period in the mouse, we hypothesized that the apparent increase in gene

488

Anne E. Peaston et al.

expression pattern must be an artifact. Does polyadenylation or some other transcript alteration of specific genes during oocyte maturation affect the efficiency of their conversion from RNA to cDNA? Does oligo(dT)priming lead to more efficient conversion of polyadenylated transcripts than poorly adenylated transcripts, and thus their apparent increase reflects adenylation status rather than abundance? Alternatively, normalization problems associated with the changing total RNA content between FGOs and OOs could produce such an artifact. Here, we outline our reasoning and experiments supporting the hypothesis that this pattern arises primarily as an artifact of normalization. Using quantitative amplification and dot blotting (QADB), we and others have reported an apparent increased abundance of Ing1 and Btg4 and Gnai2 (Gai2) transcripts during the oocyte to embryo transition (Evsikov et al., 2004; Rambhatla et al., 1995). QADB was developed to allow determination of patterns of expression and relative copy number of many different mRNAs from a common set of samples using small numbers of embryos (Rambhatla et al., 1995). The methodology relies on exponential amplification of cDNA. Interestingly, linear amplification methodologies were used in microarray experiments showing increased transcript abundance during transcriptional silence (Fig. 24.1) (Su et al., 2007; Wang et al., 2004). However, the possibility that amplification had skewed QADB or microarray results seemed low since cDNA from both linear and exponential amplification have been well documented to reflect relative abundances of mRNAs in the primary material (Iscove et al., 2002; Van Gelder et al., 1990). Another potential source of error could be inefficient oligo(dT)-primed cDNA synthesis, particularly if coupled with specific gene detection with probes or primers located in the 50 region of transcripts (Fig. 24.2). Wang et al. (2004) proposed that, in their microarray experiments, increased specific transcript abundance in the MII-stage oocyte (OO) might be explained as an artifact due to increased efficiency of oligo(dT)-primed cDNA synthesis of mRNAs that undergo cytoplasmic polyadenylation during oocyte maturation. If this were so, it would be difficult to be certain whether transcripts with decreased abundance in the OO had undergone degradation or were stable deadenylated transcripts with relatively short polyA tails. In this context, it is also worth noting that deadenylation of stable transcripts in growing oocytes is not necessarily complete, and polyA tail lengths of 40–90 nt have been reported (Bachvarova, 1992). It is unlikely that this length polyA tail would be inefficiently reverse transcribed in comparison with, for example, a 150 nt polyA tail in the presence of excess oligo(dT) primer. Nonetheless, cDNA priming strategy, among other factors, has been shown to significantly influence the yield of reverse transcription reactions (e.g., see Stahlberg et al., 2004; Stangegaard et al., 2006). Lequarre et al. (2004) observed significant differences in apparent

Normalized intensity (bits) Normalized intensity (bits) Normalized intensity (bits) Normalized intensity (bits)

Cluster 1 (1182 probesets)

0.6 0.5 0.4 0.3 0.2 0.1 0

Cluster 2 (1460 probesets)

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

Cluster 3 (514 probesets

0.5 0.4 0.3 0.2 0.1 0

Cluster 4 (538 probesets)

0.6 0.5 0.4 0.3 0.2 0.1 0 FGO

OO

Zyg

E2C

Figure 24.1 Relative gene expression patterns during the period of transcriptional silence in mouse oocytes and early embryos. Raw microarray data from Wang et al. (2004) (ArrayExpress; http://www.ebi.ac.uk/arrayexpress) was reanalyzed using RMA background correction and quantile normalization, followed by a k-means clustering. Cluster 1 clearly shows a decreased signal between FGO and OO, while clusters 2, 3, and 4 show increased signal at different points. A fifth cluster (data not shown) showed a roughly uniform signal for oocytes and embryos. FGO, full-grown oocyte; OO, ovulated oocyte; Zyg, zygote; E2C, early two-cell embryo prior to transcriptional activation of embryonic genome.

490

Anne E. Peaston et al.

A

B

5¢

3¢

8

8

5¢

3¢

1

3

8

Figure 24.2 Schematic of the interacting effects of inefficient cDNA synthesis and primer or probe position on transcript quantitation. (A) Efficient synthesis of full-length cDNA, both 50 and 30 PCR products or probes will detect the same copy number. (B) Inefficient synthesis of full-length cDNA, leading to discordant results for expression quantitation depending on the position of the probe/primer. Thin black lines represent full-length (left) or 50 -truncated (right) cDNA. Thick black lines represent PCR products or probes targeted to different regions of the cDNA. Numbers below thick black bars represent relative cDNA copy number detected by the specific probe.

transcript abundance in maturing bovine oocytes, depending on usage of oligo(dT)- or random hexamer priming for cDNA synthesis prior to quantitative RT-PCR (qPCR) (Lequarre et al., 2004). T was avoided in QADB because cDNA preparation and amplification were designed to capture, reverse transcribe, and amplify only approximately 300–900 nt from transcript 30 -ends, and dot blot probes were designed to hybridize within the 30 -end of the gene. Moreover, despite using random priming for cDNA preparation, Su et al. (2007) also observed increased transcript abundance in microarray experiments comparing OOs with FGOs. Taken together, these considerations suggested that, while efficiency of oligo(dT)-priming of cDNA synthesis might play a role, particularly in settings of whole scale polyadenylation alterations, other factors have greater influence on the apparent increase of some transcripts during the transcriptionally silent period of oocyte maturation.

2.2. Normalization to an exogenous control eliminates apparent expression increase during transcriptional silence One factor in common between the QADB analyses and the microarray analyses of Wang et al. (2004) and Su et al. (2007) was a set of normalization procedures inadvertently failing to account for the decreasing total RNA content of oocytes during the period of transcriptional silence. In the case of QADB, specific hybridization signals were normalized to the total cDNA per dot blot, a derivative of the total mass of RNA per sample. In the

Interrogating the Transcriptome of Oocytes and Preimplantation Embryos

491

microarray experiments, microarray chips were loaded with an equal quantity of RNA per developmental stage. To test the hypothesis that normalization was the major cause of the artifactual increase, rather than inefficiencies of cDNA synthesis related to oligo(dT) priming, we used qPCR to compare gene expression patterns in oocytes and preimplantation embryos. There is no endogenous transcript that is generally agreed to be completely satisfactory for normalization over this period of development. Indeed, endogenous genes whose transcripts are typically used as internal controls are unreliable during this developmental period in mice as their transcript levels are unstable (Mamo et al., 2007). Therefore, we normalized to a luciferase control in the same qPCR. A constant number of oocytes or embryos, spiked with a constant quantity of polyadenylated luciferase RNA right before RNA extraction, was used to isolate RNA (detailed in the Appendix). It has been argued that normalization to oocyte or embryo number is the ideal approach to adequately compensate for the progressively decreasing total RNA (Evsikov and Marin de Evsikova, 2009b). Since we added an invariant amount of luciferase RNA per oocyte or embryo, its measurement acted as a surrogate for oocyte or embryo number. Additionally, it was a necessary control for technical variation in qPCR since small variations at any step of RNA extraction, cDNA synthesis and PCR assembly and run can lead to large variations in the results, complicating comparisons between samples. cDNA was synthesized either from equal amounts of RNA per developmental stage, or from equal numbers of oocytes or embryos at each stage, using either oligo(dT) or random priming. We selected Dbf4, Ing1, Elavl1, Ctnnb1, and Ccndbp1 for qPCR analysis. Transcripts of Ctnnb1 and Ccndbp1 (formerly known as Maid) are polyadenylated during oocyte maturation or in the zygote; Ing1 was previously shown by QADB, and Ctnnb1 (Catnb) by microarray analysis, to have significantly increased signals in OOs or zygotes (Evsikov et al., 2004; Oh et al., 2000; Wang et al., 2004). When reanalyzing the microarray data of Wang et al. (2004), we found increased signal from Dbf4 in OOs and zygotes (Cluster 3, Fig. 24.1). Elavl1 signals did not reach the threshold for detection in the data of Su et al. (2007) although in the processed data from Wang et al. (2004) its pattern of expression fits Cluster 2 (Fig. 24.1), that is, increased in OOs. With normalization to luciferase qPCR, results using cDNA prepared from equal amounts of RNA per sample (Fig. 24.3) were very similar to those using cDNA from equal numbers of oocytes or embryos per sample (not shown). No statistically significant increase in transcript signal was observed for any tested gene during oocyte maturation or later stages of transcriptional silence using either oligo(dT) or random priming. Instead, there was no statistically significant difference in the signals for any transcript between the FGO and OO, although transcripts were degraded in zygotes and two-cell embryos. This indicated that, under these conditions, the

Anne E. Peaston et al.

ND

1.2 1 0.8 0.6 0.4 0.2 0

ND

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 1.2 1 0.8 0.6 0.4 0.2 0

ND

FGO

OO

Zyg

ND e2C

Dbf4

1.2 1 0.8 0.6 0.4 0.2 0

Ctnnb1

1.2 1 0.8 0.6 0.4 0.2 0

1.2 1 0.8 0.6 0.4 0.2 0

Ing1

ND

RP

1.4 1.2 1 0.8 0.6 0.4 0.2 0

1.4 1.2 1 0.8 0.6 0.4 0.2 0 1.2 1 0.8 0.6 0.4 0.2 0

Ccndbp1

dT

1.2 1 0.8 0.6 0.4 0.2 0

Elavl1

RQ

RQ

RQ

RQ

RQ

492

FGO

OO

Zyg

e2C

Figure 24.3 qPCR assay of relative gene expression in oocytes and preimplantation embryos normalized to exogenously provided luciferase control. cDNA was prepared from equal quantities of RNA from each batch using either oligo(dT)18 (left column of histograms) or random decamer priming (right column of histograms). The histograms show quantitation of gene expression relative to FGO, bars represent standard error of the mean calculated from three independent experiments with triplicate measurements. X-axis labels as for Fig. 24.1.

increased poly(A) tail length that occurs with some transcripts during oocyte maturation does not lead to significantly increased transcript detection in oligo(dT)-primed cDNA. These observations were similar to those of

Interrogating the Transcriptome of Oocytes and Preimplantation Embryos

493

Su et al. (2007) whose qPCR analyses, normalizing to rabbit beta globin mRNA spiked into oocyte or embryo samples at the beginning of RNA extraction, and using random primers for cDNA synthesis, showed stable or slightly decreased expression of genes that appeared to be increased in OO by microarray analysis (Su et al., 2007). Together, these results indicate that normalization methods that make insufficient allowance for the background RNA degradation occurring during oocyte maturation can lead to artifacts in microarray and other analyses of gene expression.

2.3. Choices in both computational and experimental normalization can lead to artifacts Artifacts will appear whenever data is normalized to an internal standard that varies between samples, and this includes total mass of RNA in the case of maturing oocytes and preimplantation embryos. Normalization can be explicit in the experimental procedures, or implicit in the assumptions underlying standard computational analysis of microarray or deep sequencing data. An important goal of standard microarray normalization methods is to ‘‘detect and correct systematic differences between chips so that data from different chips can be directly compared’’ (Gautier et al., 2004). These normalization methods rest on the assumption that, while individual gene expression may change between samples, the absolute RNA content, as well as the distribution among the various transcripts will be roughly equivalent. The assumption informs the general practice of hybridizing an equal mass of cDNA per sample per chip. As described above, these assumptions are violated during mouse oocyte maturation and also during development through the early cleavage stages. Wang et al. (2004) normalized microarray data using the invariant set method of dChip software, which chooses a subset of perfect match probes with small within-subset rank difference between multiple arrays, as the basis for fitting a normalization curve (Li and Hung Wong, 2001; Wang et al., 2004). Su et al. (2007) recognized that standard normalization methods would be inappropriate. Instead, they used quantile normalization ‘‘based on a group of nondifferentially expressed genes that were sampled based on estimated normal mixture model using expectation maximization . . . algorithm’’ (Su et al., 2007). The results were compared with published data showing unchanged expression of specific genes, and were verified experimentally by multiple qPCR experiments. Table 24.2 further illustrates this point, showing results of a thought experiment measuring the expression of Gene A in FGO compared with oocytes that have been harvested at six arbitrary times (OO1–OO6) during the course of maturation. Sample OO1 is equivalent in every respect to the FGO, but the total RNA per oocyte of the remaining samples progressively

494

Anne E. Peaston et al.

Table 24.2 Quantitation artifacts predicted from varying abundance of a specific transcript in context of varying total RNA content

Sample

Total RNA Gene A

a

b c

b

True abundance

Measured abundance for decreasing total mRNA per OOa

FGO

OO1

OO

100.0 Y 10.0 10.0 10.0 9.0 10.0 8.0 10.0 7.0 10.0 6.0 10.0 5.0 10.0 4.0 10.0 3.0 10.0 2.0 10.0 1.0

100.0 10.0 c 9.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0

OO2

OO3

OO4

OO5

OO6

90.0 11.1 10.0 8.9 7.8 6.7 5.6 4.4 3.3 2.2 1.1

80.0 12.5 11.3 10.0 8.8 7.5 6.3 5.0 3.8 2.5 1.3

70.0 14.3 12.9 11.4 10.0 8.6 7.1 5.7 4.3 2.9 1.4

60.0 16.7 15.0 13.3 11.7 10.0 8.3 6.7 5.0 3.3 1.7

50.0 20.0 18.0 16.0 14.0 12.0 10.0 8.0 6.0 4.0 2.0

Where the amount of OO RNA used to take the measurement is normalized to the amount in the FGO, Equation (3): Measured value ¼ N
Y/FGO, where N is true quantity of Gene A transcripts in OO, Y is total RNA in OO, FGO is total RNA in FGO and assuming perfect proportional representation of specific gene transcripts in the cDNA and perfect quantitative detection of specific transcripts in cDNA from all samples. The true total amount of RNA per oocyte. The unit of measurement is unimportant to the general case presented here. The values in italics indicate a range of conditions under which transcript abundance will appear stable although its true abundance is decreased in OO. In cells to the right of the boxes, transcript abundance falsely appears to be increased in the OO. In cells to the left, the relative decrease is underestimated except in OO1 which contains the same amount of RNA as FGO.

decreases, and in sample OO6 each oocyte contains only 50% of the total RNA of an FGO. If expression measurements were carried out on each sample using the same amount of RNA as in a single FGO, the result in sample OO6 would be the analysis of total RNA of two oocytes. Therefore, if Gene A’s real expression remains stable at 10 transcripts per oocyte, the measured value in OO6 will be double its level in the FGO. If Gene A’s real expression had declined to five transcripts per oocyte in sample OO6, the measurement will indicate unchanged expression. In the context of a significant change in the total RNA content between samples, the assessment of increased or decreased abundance of specific transcripts (or genes) is explicitly dependent upon the use of appropriate normalizations in both experimental and computational procedures. In the absence of these, such assessments will be explicitly dependent upon the relationship between the change in the actual transcript abundance and the overall change in total RNA content.

Interrogating the Transcriptome of Oocytes and Preimplantation Embryos

495

3. Comparative Analysis of Large-Scale Expression Analyses 3.1. Comparison of results between large-scale experiments is complicated by differences in both experimental and computational approaches The attempts of Su et al. (2007) to compensate for loading equal cDNA mass per microarray were not entirely successful, although the list of unstable transcripts reported by them was likely to be highly accurate. It would be interesting to compare these data with independent quantitative data on transcriptome differences between FGO and OO, obtained by some other means. The only other large datasets available for analyzing mouse FGO and OO are cDNA libraries. Indeed, comparative analyses of expressed sequence tags (ESTs) from bacterially cloned cDNA libraries from FGO and two-cell embryos have mapped temporal changes in gene expression during the mouse oocyte to embryo transition (Evsikov et al., 2004, 2006; Sharov et al., 2003). Unfortunately, similar comparative analysis using available sequence information from FGO and mature OOs presents difficulties, as discussed below, because of major differences in library preparation and data analysis. We surveyed the FGO transcriptome through 50 single pass sequencing of a cDNA library classically constructed using polyA mRNA extracted from 12,000 FGOs from inbred C57BL/6J females. Nearly 19,000 ESTs were generated from individual clones with average insert size of 1531 bp, representing transcription from approximately 4800 annotated genes and also 600 transposable element and unannotated loci (Evsikov et al., 2006). Because the cDNA was not PCR-amplified before cloning, and the library was not normalized, the number of ESTs per gene should approximately represent the relative abundance of its transcripts. On the other hand, the OO transcriptome was surveyed recently by high-throughput sequencing of two molecular cDNA libraries constructed using polyA mRNA extracted from two independent OOs from outbred MF1 females (Tang et al., 2009). The molecular libraries underwent a total of 37–39 cycles of PCR before sequencing with the SOLiD system, generating millions of 50 nt reads. Transcript fragments from 13,619 annotated genes were identified through at least one read assigned to the RefSeq transcript(s) of that gene. We compared the data for each oocyte and found that the two OOs shared expression of over 11,000 genes, but each also exhibited mutually exclusive expression of about 1200 genes (Fig. 24.4, left Venn diagram and see below for method). These differences may have been associated with individual differences in oocytes from outbred mice. Alternatively, since most of the genes expressed in the mutually exclusive sets were represented

496

Anne E. Peaston et al.

A

B

C

FGO 260 37

44

FGO

FGO

211

96

49

115

2645

16160

4423 1162 OO#1

6809

1144 OO#2

OO

OO

Figure 24.4 Comparisons of annotated gene content in large-scale expression analyses. (A) High-throughput sequencing of FGO bacterially cloned cDNA library compared with two OO molecular cDNA libraries. (B) Genes found exclusively in FGO in the previous comparison were compared with the set of genes downregulated in OO in the microarray experiment of Su et al. (2007). (C) Unique gene probesets on the Affymetrix mouse 430v2 GeneChip system, Entrez Genes were obtained from a custom chip definition file (Dai et al., 2005) available at http://brainarray.mhri.med.umich. edu/Brainarray/Database/CustomCDF/genomic_curated_CDF.asp and converted to MGI IDs. Genes exclusive to FGO and excluded from OO molecular libraries and from the set of microarray-identified downregulated genes were compared with the set of genes uniquely identified on Affy430v2 chip.

by a low number of reads, their differential presence in the libraries is most likely due simply to sampling differences. Accordingly, an increase in the cutoff for positive gene identification to 10 reads in at least one OO resulted in the identified expression of 11,079 genes, with few genes exclusive to one library or the other.

3.2. Comparison of FGO and OO sequencing datasets To prepare a comparative analysis of the sequencing datasets, data for the FGO was extracted from supplementary files EvsikovSuppFGO_Library_1_0.xls (Evsikov et al., 2006), and data for the OO from Supplementary Table 1 of Tang et al. (2009), with all genes converted to Mouse Genome Informatics (MGI) gene/marker identification (ID) numbers (MGI IDs) for ease of comparison (see Appendix). As would be predicted from the biology of oocyte maturation, there was substantial overlap between the FGO and OO transcriptomes, with transcripts of 4504 annotated genes found in common (Fig. 24.4, left Venn diagram). Many more genes were represented in the OO data than in the FGO data. This is also unsurprising given the significant increase in sampling depth of SOLiD sequencing, which is estimated to be capable of detecting transcripts present at less than one copy per cell (Anonymous, 2008).

Interrogating the Transcriptome of Oocytes and Preimplantation Embryos

497

3.3. Are the transcripts identified in FGO but absent from OO biologically meaningful? Just over 5% (260) of FGO-identified annotated genes were absent from the list of genes represented in the OOs. Evsikov et al. (2006) estimated that, in the FGO cDNA library, transcripts present in fewer than 1000 copies per FGO are less than 50% likely to be represented in the library. This suggests that many if not most genes represented in this library have abundant transcripts in FGO and, arguing from this, are likely to be found in the OO unless they are explicitly targeted for rapid degradation in the transition from FGO to OO. In the absence of targeted degradation, possible explanations for the absence of FGO transcripts in the RNA-Seq OO data might include technical issues associated with sample preparation, or with bioinformatic analysis. We looked for independent evidence of expression of these transcripts in OO using the MGI Batch Query tool to access gene expression data in the MGI databases. Evidence from PCR, microarray, and cDNA library clones documented expression of 48 of the 260 genes in OOs, suggesting technical failure to amplify transcripts of these genes (Table 24.3). The comparison of lists of genes and/or transcripts resulting from disparate analyses such as these can also be greatly complicated by differences in the methodology used to associate specific cDNA or RNA-Seq data with genes. The FGO library was manually annotated, associating ESTs with known genes where possible, but keeping all sequenced transcripts within the final list of expressed genes/transcripts. Tang et al. (2009) only reported reads overlapping with mouse RefSeq transcripts. Accordingly, genes identified in the FGO library that currently lack a mouse RefSeq would be excluded from identification in the RNA-Seq data, for example, D14Ertd426e. These genes might be identified in the OO data by reanalysis of the raw reads using other genome annotation tracks, or using future updates of RefSeq annotations. A biologically interesting alternative explanation for the absence of FGO transcripts in the OO sample lies in the possibility of transcript cytoplasmic deadenylation, with or without degradation during oocyte maturation. Indeed, firm evidence confirms expression of the zona pellucida glycoprotein 4 pseudogene (Zp4-ps) in FGO but not in OO (Evsikov et al., 2006). To further test this hypothesis, we returned to the microarray data specifically designed to determine gene expression differences between the mouse FGO and OO (Su et al., 2007). A list of 2694 genes downregulated in OO (kindly supplied by Dr. John Eppig, The Jackson Laboratory) were compared with the 260 genes that we found to be FGO-specific, and 49 FGO genes were included in the OO reduced abundance set

498

Anne E. Peaston et al.

Table 24.3 Forty-eight genes found in FGO cDNA library but not in deeply sequenced OO libraries MGI ID

Gene symbol

Source of expression dataa

MGI:109298 MGI:1354961

Mapkapk2 Synj1

MGI:1919715

Zfp131

MGI:1921424

Gtsf1

MGI:1921627 MGI:2149033 MGI:2444036 MGI:2444431

Hsf2bp Obox2 Srcap Rab11fip3

MGI:2684949

Tdrd5

MGI:97515 MGI:3700744 MGI:2684995 MGI:1261919 MGI:2141773 MGI:2385865 MGI:1924217 MGI:1923517 MGI:3650906 MGI:1917364 MGI:1918249 MGI:2142163 MGI:1889333 MGI:3644989 MGI:2685916 MGI:2139883 MGI:2140708 MGI:3036240 MGI:3028594 MGI:107735 MGI:2140313 MGI:2183559 MGI:1921895 MGI:3611448 MGI:1859645 MGI:2145569 MGI:1923420

Pcsk5 Nlrp4g Gm149 D6ErtD527e E330021D16Rik BC018507 2700049A03Rik Fads1 Zfp616 Prrg1 Naa40 A730082K24Rik AA545190 Gm5946 Gm1070 Zbtb10 C87977 9430025M13Rik Zfp422-rs1 Myo9a AI481877 Arfgap1 Fam46c Bin2 Usp27x AA536875 Tmem29

Zygote, cDNA, MGI:1868910 2-cell embryo, cDNA, MGI:568000 MII oocyte, qRTPCR, MGI:3813811 MII oocyte, qRTPCR, MGI:3769551 MII oocyte, cDNA, MGI:3155178 MII oocyte, cDNA, MGI:3155346 MII oocyte, PCR, MGI:3816489 MII oocyte and zygote, microarray, ArrayExpress MII oocyte and zygote, microarray, GEO GDS814 MII oocyte, cDNA, MGI:3064022 MII oocyte, cDNA, MGI:3155336 MII oocyte, cDNA, MGI:3113803 MII oocyte, cDNA, MGI:3113824 MII oocyte, cDNA, MGI:3113947 MII oocyte, cDNA, MGI:3114223 MII oocyte, cDNA, MGI:3114440 MII oocyte, cDNA, MGI:3114599 MII oocyte, cDNA, MGI:3114744 MII oocyte, cDNA, MGI:3114837 MII oocyte, cDNA, MGI:3155358 MII oocyte, cDNA, MGI:3155358 MII oocyte, cDNA, MGI:3115146 MII oocyte, cDNA, MGI:3116338 MII oocyte, cDNA, MGI:3116807 MII oocyte, cDNA, MGI:3155669 MII oocyte, cDNA, MGI:3116844 MII oocyte, cDNA, MGI:3116994 MII oocyte, cDNA, MGI:3117036 MII oocyte, cDNA, MGI:3117421 MII oocyte, cDNA, MGI:3118247 MII oocyte, cDNA, MGI:3119015 MII oocyte, cDNA, MGI:3156464 MII oocyte, cDNA, MGI:3119576 MII oocyte, cDNA, MGI:3227871 MII oocyte, cDNA, MGI:3228238 MII oocyte, cDNA, MGI:3228658

Interrogating the Transcriptome of Oocytes and Preimplantation Embryos

499

Table 24.3 (continued)

a

MGI ID

Gene symbol

Source of expression dataa

MGI:109424 MGI:2676368 MGI:2142361 MGI:2141341 MGI:2144989 MGI:1926059 MGI:2140466 MGI:1924248 MGI:1919565 MGI:1921356 MGI:3651308 MGI:1913628

Abca4 Dnajc13 C86187 C87414 Plekhh1 4930562C15Rik AU022252 Fbxw27 Ccdc74a Dcaf6 Gm15698 2310028O11Rik

MII oocyte, cDNA, MGI:3229762 MII oocyte, cDNA, MGI:3230007 MII oocyte, cDNA, MGI:495770 MII oocyte, cDNA, MGI:4417521 MII oocyte, cDNA, MGI:496146 MII oocyte, cDNA, MGI:3064042 MII oocyte, cDNA, MGI:3064143 MII oocyte, cDNA, MGI:3064189 MII oocyte, cDNA, MGI:3064791 MII oocyte, cDNA, MGI:3064996 MII oocyte, cDNA, MGI:1863717 MII oocyte, cDNA, MGI:1865137

Source data includes only whether gene expressed in oocyte or embryo during transcriptional silence, the nature of the evidence (whether cDNA clone, PCR data, etc.), and one MGI or other identifier for expression evidence in MGI or other database.

(Fig. 24.4, middle Venn diagram). While this finding supports the possibility that the FGO-specific set may be degraded during oocyte maturation, there are still 211 genes to be accounted for. A trivial explanation might be that the FGO library contains a number of transcripts not found in other libraries, and therefore not represented on the microarray. Indeed, comparison of the remaining 211 FGO-specific genes with the list of genes represented by unique probes on the Affymetrix mouse 430v2 GeneChip revealed that 96 were not targeted with probes (Fig. 24.4, right Venn diagram). The remaining 115 genes were represented on the chip but most were either classified as not present in either sample (26 genes) or did not pass the authors’ stringent requirements for classification as significantly degraded. In short, a rigorous comparison of the relative abundance of gene representation in the FGO cDNA library and the OO mRNA-Seq datasets would require extensive reanalysis of the raw read data. The extent to which stage-specific differences in transcripts’ polyadenylation status could be responsible for discrepancies in the lists of genes identified as differentially present is unknown. A probe-level reanalysis and comparison of the random-primed microarray data from Su et al. (2007) and the oligo(dT)primed data from Murchison et al. (2007) identified several hundred transcripts with evidence of change in polyadenylation status between the FGO and OO stages (Salisbury et al., 2009). In light of the discussion above, such an effect would likely require an exceedingly short or absent polyA tail, a possibility that remains to be directly tested.

500

Anne E. Peaston et al.

4. Taking Isoform-Specific Changes into Account A final complication to analysis of gene expression dynamics during oogenesis and early embryogenesis lies in the potential for differential adenylation and degradation of the different transcript isoforms of a gene. Most existing analyses do not treat isoforms separately, and instead report summarized, averaged, or even arbitrary expression dynamics if multiple isoforms are expressed. However, the probe-level analysis of Salisbury et al. (2009) makes it clear that the differences in expression among isoforms can be significant, and that the collapse of multiple isoforms into a single category of increased or decreased abundance can be problematic. Within the OO RNA-Seq data, ambiguities of exon-to-transcript alignment lead to identical read counts for different transcript isoforms of the same gene. Two examples are illustrated in Table 24.4. The transcript read counts differ between the oocytes as would be expected. However, read counts for different transcripts of a single gene are identical or near identical in both oocytes. Many more such examples can be found in the data. It is very unlikely that so many different isoforms should have identical read counts in one oocyte, let alone two, suggesting this is an artifact of the analysis. In addition, alignment-based analysis of expression data frequently includes automatic masking of retrotransposons before mapping reads to the genome, a step that complicates identification of transcripts with retrotransposon-derived alternative exons, of which there are thousands documented (Faulkner et al., 2009). For example, the very abundantly expressed gene Spin1 has three different 30 -UTR alternative transcripts, approximately 4.1, 1.7, and 0.8 kb, associated with use, in oocytes, of three different polyadenylation and cleavage sites in the 30 -terminal exon (Oh et al., 2000). In addition, it has two alternative 50 isoforms, both expressed in oocytes but Table 24.4 Read numbers for alternative transcripts in ovulated oocyte data Read #a Gene symbol

RefSeq identifier

Oocyte # 1

Oocyte # 2

Slc12a1

NM_011137 NM_198934 NM_198932 NM_198933 NM_001077264 NM_007458

2547 2546 2527 2526 171 171

2366 2366 2335 2335 236 236

Ap2a1 a

The number is the sum of 50-mer reads for the transcript, and is taken from Supplementary Table 1 of Tang et al. (2009).

501

Interrogating the Transcriptome of Oocytes and Preimplantation Embryos

in one the first exon is entirely within an endogenous retrovirus. The transcription of this single gene has the potential for six alternative transcripts in oocytes (Peaston et al., 2004). Isoform diversity of Spin1 is reflected by only two sequences in RefSeq database (Fig. 24.5). Isoform 1, derived from a transcript isolated from testis, lacks the long 30 -UTR seen in oocytes, and depends for its certain identification on reads mapping to exons 1, 2, and 3. Reads mapping to the distal 30 -UTR in exon 6 would automatically be assigned to isoform 2. However, in the oocyte, distal 30 -UTR reads can equally represent transcripts with a 50 segment identical to isoform 1 but with longer 30 -UTRs. Moreover, positive identification of isoform 2 is not possible if the retrotransposon was masked before mapping. Finally, reads mapping to exons 4, 5 and the 50 end of exon 6 will be assigned to both transcripts since they cannot be reliably assigned to either transcript, thereby affecting quantification. Better quantification of relative gene representation might be achieved by reanalysis of the OO data using reads per kilobase of exon model per million mapped reads, as was indeed suggested by Tang et al. (2009). This method normalizes for transcript length and total read number per transcript, thereby reflecting transcript molar concentration in the original sample (Mortazavi et al., 2008). Moreover, read-level and exon-level analyses might allow relative quantification of specific alternative transcript usage, opening the way for functional insights, much as 30 -UTR probelevel analysis of microarray data has been used to analyze alternative polyadenylation and cleavage site use (Salisbury et al., 2009). Finally, for transcript isoform identification there is as yet no perfect substitute for cloning and sequencing full-length mRNAs from the cell or tissue of interest. 2

4136 nt

1

1064 nt

1

2

3

4

5

6

Figure 24.5 Spin transcript schematic (not to scale) with exons numbered. Alternative first exon from MT retrotransposon (black box) and three alternative polyadenylation and cleavage sites (arrowheads in exon 6). Thin black line 1 represents Spin1 isoform 1 RefSeq # NM_011462, and thick black line 2 represents Spin1 isoform 2 RefSeq # NM_146043. Short black lines indicate notional short sequencing reads. Reads mapping to exons 1, 2, and 3 will identify Spin1 isoform 2, but in the event of their absence, this isoform cannot be confidently identified. Short reads mapped to exons 4 and 5 will potentially be assigned to both transcripts. Short reads mapping to most of exon 6 will be assigned to Spin1 isoform 1, but may represent variants of isoform 2 with longer 30 -UTR.

502

Anne E. Peaston et al.

5. Concluding Remarks The normalization problem might also be expected to come into play when the total mRNA of the organism is increasing, for example, a stable transcript might appear to decrease as the mRNA content of the embryo doubles between the two-cell stage and the eight-cell stage (Table 24.1). Normalization problems can be circumvented in two basic ways. For qPCR experiments, spiking the original samples with equal quantities per oocyte or embryo of exogenous RNA such as rabbit globin or luciferase will act not only as a control for technical variation between samples but is also a convenient yardstick for relative quantification. Alternatively, using template equivalent to a defined number of oocytes or embryos will allow undistorted relative measurements. In principle, the same type of solution could be applied to design of microarray and high-throughput sequencing assays. Inclusion on microarray chips of a suitable range of exogenous probe sets for a standard mRNAs spiked into samples before RNA extraction and assay would be useful under many conditions. However what constitutes ‘‘suitable’’ is far from clear, and normalization for these types of assays remains a significant problem for assaying relative gene expression in a background of changing cellular total RNA content. Comparative quantitative analysis of large nonnormalized bacterially cloned cDNA libraries from oocytes and embryos at different stages by definition avoids normalization problems but, even with technical advances and reduced prices of high-throughput sequencing, library construction and analysis is time and resource-intensive. However, these libraries are invaluable for generating long sequences, and creating physical resources of unequivocal full-length transcripts for future reference and manipulation. On the other hand, high-throughput sequencing, from pooled cell samples down to the single cell level, is still a young technology undergoing rapid technical developments. Sample preparation for some applications, although still technically demanding, has been simplified through commercially available kits, but data storage and analysis present significant challenges. Nonetheless, the technology offers the ability to survey the transcriptome of a cell in much greater depth than is usually covered by classical cDNA libraries, albeit with challenges associated with interpreting the meaning of a single read of extensively amplified material. This brief outline of some systematic artifacts associated with commonly used gene expression assays, and puzzles emerging from comparison of large datasets from whole transcriptome analyses may be of assistance in critical appraisal of the scientific literature, and careful experimental design for transcriptome analysis in oocytes and preimplantation embryos.

Interrogating the Transcriptome of Oocytes and Preimplantation Embryos

503

ACKNOWLEDGMENTS The authors thank Dr. John Eppig, Dr. Carol Bult, and Dr. Constance Smith for critical review of the manuscript, and are grateful to Dr. John Eppig and Dr. You-Qiang Su for sharing data with us. The work was funded by National Cancer Institute grant CA120204 (to A. E. P.), National Institutes of Health grant HD37102 (to B. B. K.), and NIH GM 07276 (to J. H. G.).

Appendix. Quantitative Evaluation of Transcript Abundance Using Real-Time PCR Oocyte and embryo isolation Specific stages of oocytes and embryos were from B6D2F1/J female mice (The Jackson Laboratory, Bar Harbor, ME). Equipment and materials required and the full method were described previously (Nagy et al., 2003). In brief, the following important timing steps are involved: 1. Synchronize estrous cycles of females at 3 weeks of age by intraperitoneal (IP) injection of 5 IU pregnant mare serum gonadotrophin (eCG, Sigma, Cat. # G4527). 2. Collect full-grown germinal vesicle-intact (GV stage) oocytes 46–48 h after eCG injection. 3. Stimulate ovulation by IP injection of 5 IU human chorionic gonadotrophin (hCG, Sigma, Cat. # C1063) 46–48 h after eCG injection. 4. Collect ovulated oocytes (MII stage) 13–17 h after hCG injection. 5. For embryo isolation, mate females 12 h post-hCG for exactly 1.5 h. 6. Isolate zygotes 20–24 h after hCG injection. 7. Isolate early two-cell stage embryos 32–39 h after hCG injection. This is prior to transcriptional activation of the embryonic genome. In later stage embryos, new transcription from the embryonic genome will confound measurement of maternal transcripts by qPCR. Collect FGOs/cumulus cell complexes into M2 medium supplemented with 0.2 mM 3-isobutyl-l-methyl xanthine (IBMX) to inhibit oocyte meiotic maturation during sample collection, and gently denude them of granulosa cells by pipetting. Remove adherent granulosa cells from OOs and zygotes by incubating them in hyaluronidase (400 IU/ml in M2 medium). Then remove the zona pellucida and any remaining granulosa cells stuck to it by incubating oocytes and embryos in pronase (0.5% in M2 medium). After zona removal, rinse the oocytes or embryos thoroughly in drops of PBS-PVP (0.4% polyvinyl pyrrolidone in phosphate buffered saline). Those that appear imperfect or to be the incorrect stage should be rejected and the remainder transferred in minimum volume to Eppendorf tubes, snap frozen, and stored at 70  C until RNA extraction.

504

Anne E. Peaston et al.

RNA extraction and cDNA synthesis For qPCR experiments, prepare samples of RNA from aliquots of 55 oocytes or embryos using the PicoPure RNA Isolation Kit (Arcturus; Cat. # KIT0202). Fifty-five oocytes or embryos produce enough cDNA for triplicate qPCR evaluations of 6–10 genes. When comparing a series of samples from different oocyte or embryo stages, it is best to prepare RNA and cDNA from the whole series at one time to minimize technical variation. To provide an internal control for RNA isolation, cDNA synthesis, and PCR, add 0.04 ng polyadenylated luciferase RNA (Promega, Cat. # L4561) to each sample immediately prior to RNA extraction. After extraction, treat the RNA with DNaseI (Qiagen, Cat. # 79254) as described in the Arcturus manual, yielding a final volume of 11 ml RNA. Synthesize cDNA using the MessageSensor kit (Ambion, Cat. # 1745) according to manufacturer’s instructions with either oligo(dT)18 or random decamer primers. cDNA may be diluted to a convenient volume using a suitable buffer such as 10 mM Tris pH 8.3. Test the quality of cDNA by PCR, using primers for mitochondrial ATP synthase (mt-Atp6) and cDNA template equivalent to 0.5–1 oocyte or embryo. cDNA quality is tested using mt-Atp6, as mt-Atp6 is highly expressed in all oocyte and embryo stages (Harwood et al., 2008). Weak or absent mt-Atp6 amplification indicates a problem with cDNA synthesis and the sample should be discarded before further experimentation. cDNA samples are stored at 80  C.

Quantitative PCR For qPCR experiments, we use the ABI 7500 Real-time PCR System (Applied Biosystems, Foster City, CA) in conjunction with TaqManÒ Universal PCR Master Mix (PE Applied Biosystems, Cat. # 4304437), and gene-specific FAM-TAMRA-probes and primers at 5 and 20 mM, respectively (PE Applied Biosystems). Primers are designed to span or flank gene introns, where possible near the 30 -end of the gene (Table 24.5). For valid calculation of relative expression changes, approximately equal amplification efficiencies of the primer and probe sets for different genes and the luciferase internal control are required. Efficiencies are tested essentially as previously described (Baker and O’Shaughnessy, 2001). In short, RNA and cDNA are prepared from testis samples as described for oocytes and embryos, except that testis samples are spiked with luciferase RNA after tissue lysis. Serial dilutions of testis cDNA are used in qPCR reactions. The log of the relative testis cDNA concentration is plotted against the threshold cycle (Ct) for each gene, regression analysis is used to generate the best-fit line, and the slope of the line is used to determine the efficiency of each PCR reaction. cDNA clones of the transcript, obtained from the American Type Culture Collection, can be

Table 24.5 Primers and probes for qPCR Gene symbol

Probe (6FAM. . .TAMRA)

Forward primer (50 –30 )

Reverse primer (50 –30 )

Ask

TGACAGCACACCGGAAC ACCAAGTG AGCTCGGACCCTCTGAGCCC TAGTCA AATGTGACCCGTACAGCGGC CTCTT TGCAGCCAATCCCAACCAGA ACAAAA AGCGTCTCCTGCCGACCT TTCCC AGCTCACTTGCCCACTTCC TTCCACA

GGATTGAAGGATGTGA CAGGAAA GCTGATATTGACGGGC AGTATG GCCGCAGCCCTTT TGAT GTTCCTCCGAGCCC ATCA AAACGCTCCAAGGC CAAAG TAATTACAGGCTTCC GACAC AAAC TCGAAGTATTCCGCGT ACGTG

GGTCTACTCTCTGTT CTCCGTCATT TGCCCTCATCTAGC GTCTCA GTGGCTGTCGGAGTTG TCTTC CCTAGCAGGCGAGTGG TACAG CCGTAGGAGACTCTGGTTGCA

Ctnnb1 Ccndbp1 Elavl1 Ing1 mt-Atp6 Luciferasea

a

TGTTCACCTCGATATGTGC ATCTGTA AAAGCA

Luciferase primer and probes as published (Baker and O’Shaughnessy, 2001).

CTAATGCCATTGGTTGAATA AATAGG GCCCTGGTTCCTGGAACAA

506

Anne E. Peaston et al.

used as positive controls for gene-specific primers. PCR products of each reaction should be checked against the positive control by gel electrophoresis for production of a single correctly sized band, and the band sequenced to confirm its identity. The widely used comparative CT method (Applied Biosystems; User Bulletin # 2) to determine the expression level (RQ) of each gene relative to that in the FGO, normalizing the target gene signal to that of the Luciferase reference. Real-time quantitative PCR experiments are repeated at least three times, each using a different set of oocytes and embryos, with triplicate measurements on all samples per experiment. The results are presented as RQ  S.E. of the mean. For the qPCR experiments reported in this work, Student’s t-test was used to assess the difference in gene expression between FGO and OOs, and a p-value > 0.05 was taken to indicate statistically insignificant differences between these stages.

Generation of sequencing datasets for comparison of FGO and OO FGO data were extracted from the supplementary file EvsikovSuppFGO_ Library_1_0.xls (Evsikov et al., 2006). Data for the OO were extracted from Supplementary Table 1 of Tang et al. (2009). FGO and OO datasets were presented as Microsoft Excel files with unique genes identified by gene symbol (FGO) or by gene symbol and RefSeq_ID (OO). To compare the unique annotated gene content of the two datasets, it was necessary to use unambiguous gene identification. Since gene symbols are subject to change, and since one gene may have more than one RefSeq_ID, we opted to use the MGI IDs for each gene, as this number does not change, and is associated with all transcript isoforms of a gene. We used the MGI Batch Query tool (http://www.informatics.jax.org/javawi2/servlet/WIFetch? page¼batchQF) to convert unique annotated gene symbols of the FGO cDNA library and the RefSeq_IDs from each wild-type OO to their matching unique MGI ID. Gene symbols lacking an associated gene MGI ID were discarded from further analysis. Duplicates within each list were removed, and the lists were compared by constructing Venn diagrams using the web-based GeneVenn utility (http://www.bioinformatics.org/gvenn/; Pirooznia et al., 2007).

REFERENCES Anonymous. (2008). Application Note SOLiDTM System High-Throughput Analysis of Differential Gene Expression. Applied Biosystems, Foster City, CA. USA. Bachvarova, R. (1974). Incorporation of tritiated adenosine into mouse ovum RNA. Dev. Biol. 40, 52–58.

Interrogating the Transcriptome of Oocytes and Preimplantation Embryos

507

Bachvarova, R. (1985). Gene expression during oogenesis and oocyte development in mammals. In ‘‘Developmental Biology a Comprehensive Synthesis, Oogenesis,’’ (L. W. Browder, ed.), Vol. 1, p. 453. Plenum Press, New York. Bachvarova, R. F. (1992). A maternal tail of poly(A): The long and the short of it. Cell 69, 895–897. Bachvarova, R., and De Leon, V. (1980). Polyadenylated RNA of mouse ova and loss of maternal RNA in early development. Dev. Biol. 74, 1–8. Bachvarova, R., De Leon, V., Johnson, A., Kaplan, G., and Paynton, B. V. (1985). Changes in total RNA, polyadenylated RNA, and actin mRNA during meiotic maturation of mouse oocytes. Dev. Biol. 108, 325–331. Baker, P. J., and O’Shaughnessy, P. J. (2001). Expression of prostaglandin D synthetase during development in the mouse testis. Reproduction 122, 553–559. Baugh, L. R., Hill, A. A., Brown, E. L., and Hunter, C. P. (2001). Quantitative analysis of mRNA amplification by in vitro transcription. Nucleic Acids Res. 29, E29. Belloc, E., Pique, M., and Mendez, R. (2008). Sequential waves of polyadenylation and deadenylation define a translation circuit that drives meiotic progression. Biochem. Soc. Trans. 36, 665–670. Brady, G., and Iscove, N. N. (1993). Construction of cDNA libraries from single cells. Methods Enzymol. 225, 611–623. Brower, P. T., Gizang, E., Boreen, S. M., and Schultz, R. M. (1981). Biochemical studies of mammalian oogenesis: Synthesis and stability of various classes of RNA during growth of the mouse oocyte in vitro. Dev. Biol. 86, 373–383. Clegg, K. B., and Piko, L. (1983a). Poly(A) length, cytoplasmic adenylation and synthesis of poly(A) þ RNA in early mouse embryos. Dev. Biol. 95, 331–341. Clegg, K. B., and Piko, L. (1983b). Quantitative aspects of RNA synthesis and polyadenylation in 1-cell and 2-cell mouse embryos. J. Embryol. Exp. Morphol. 74, 169–182. Dai, M., Wang, P., Boyd, A. D., Kostov, G., Athey, B., Jones, E. G., Bunney, W. E., Myers, R. M., Speed, T. P., Akil, H., Watson, S. J., and Meng, F. (2005). Evolving gene/transcript definitions significantly alter the interpretation of GeneChip data. Nucleic Acids Res. 33(20), e175. De Leon, V., Johnson, A., and Bachvarova, R. (1983). Half-lives and relative amounts of stored and polysomal ribosomes and poly(A) þ RNA in mouse oocytes. Dev. Biol. 98, 400–408. Evsikov, A. V., and Marin de Evsikova, C. (2009a). Evolutionary origin and phylogenetic analysis of the novel oocyte-specific eukaryotic translation initiation factor 4E in Tetrapoda. Dev. Genes Evol. 219, 111–118. Evsikov, A. V., and Marin de Evsikova, C. (2009b). Gene expression during the oocyte-toembryo transition in mammals. Mol. Reprod. Dev. 76, 805–818. Evsikov, A. V., de Vries, W. N., Peaston, A. E., Radford, E. E., Fancher, K. S., Chen, F. H., Blake, J. A., Bult, C. J., Latham, K. E., Solter, D., and Knowles, B. B. (2004). Systems biology of the 2-cell mouse embryo. Cytogenet. Genome Res. 105, 240–250. Evsikov, A. V., Graber, J. H., Brockman, J. M., Hampl, A., Holbrook, A. E., Singh, P., Eppig, J. J., Solter, D., and Knowles, B. B. (2006). Cracking the egg: Molecular dynamics and evolutionary aspects of the transition from the fully grown oocyte to embryo. Genes Dev. 20, 2713–2727. Faulkner, G. J., Kimura, Y., Daub, C. O., Wani, S., Plessy, C., Irvine, K. M., Schroder, K., Cloonan, N., Steptoe, A. L., Lassmann, T., Waki, K., Hornig, N., et al. (2009). The regulated retrotransposon transcriptome of mammalian cells. Nat. Genet. 41, 563–571. Gautier, L., Cope, L., Bolstad, B. M., and Irizarry, R. A. (2004). Affy—Analysis of Affymetrix GeneChip data at the probe level. Bioinformatics 20, 307–315.

508

Anne E. Peaston et al.

Gebauer, F., Xu, W., Cooper, G. M., and Richter, J. D. (1994). Translational control by cytoplasmic polyadenylation of c-mos mRNA is necessary for oocyte maturation in the mouse. EMBO J. 13, 5712–5720. Hamatani, T., Carter, M. G., Sharov, A. A., and Ko, M. S. (2004). Dynamics of global gene expression changes during mouse preimplantation development. Dev. Cell 6, 117–131. Harwood, B. N., Cross, S. K., Radford, E. E., Haac, B. E., and De Vries, W. N. (2008). Members of the WNT signaling pathways are widely expressed in mouse ovaries, oocytes, and cleavage stage embryos. Dev. Dyn. 237, 1099–1111. Huarte, J., Belin, D., Vassalli, A., Strickland, S., and Vassalli, J. D. (1987). Meiotic maturation of mouse oocytes triggers the translation and polyadenylation of dormant tissue-type plasminogen activator mRNA. Genes Dev. 1, 1201–1211. Hwang, S., Benjamin, L. E., Oh, B., Rothstein, J. L., Ackerman, S. L., Beddington, R. S., Solter, D., and Knowles, B. B. (1996). Genetic mapping and embryonic expression of a novel, maternally transcribed gene Mem3. Mamm. Genome 7, 586–590. Hwang, S. Y., Oh, B., Fuchtbauer, A., Fuchtbauer, E. M., Johnson, K. R., Solter, D., and Knowles, B. B. (1997). Maid: A maternally transcribed novel gene encoding a potential negative regulator of bHLH proteins in the mouse egg and zygote. Dev. Dyn. 209, 217–226. Hwang, S. Y., Oh, B., Zhang, Z., Miller, W., Solter, D., and Knowles, B. B. (1999). The mouse cornichon gene family. Dev. Genes Evol. 209, 120–125. Iscove, N. N., Barbara, M., Gu, M., Gibson, M., Modi, C., and Winegarden, N. (2002). Representation is faithfully preserved in global cDNA amplified exponentially from subpicogram quantities of mRNA. Nat. Biotechnol. 20, 940–943. Ji, W., Zhou, W., Gregg, K., Lindpaintner, K., and Davis, S. (2004). A method for gene expression analysis by oligonucleotide arrays from minute biological materials. Anal. Biochem. 331, 329–339. Kaplan, G., Abreu, S. L., and Bachvarova, R. (1982). rRNA accumulation and protein synthetic patterns in growing mouse oocytes. J. Exp. Zool. 220, pp. 361–370 (calculated from ribosomal RNA content), quoted in Bachvarova (1985). Klur, S., Toy, K., Williams, M. P., and Certa, U. (2004). Evaluation of procedures for amplification of small-size samples for hybridization on microarrays. Genomics 83, 508–517. Ko, M. S., Kitchen, J. R., Wang, X., Threat, T. A., Hasegawa, A., Sun, T., Grahovac, M. J., Kargul, G. J., Lim, M. K., Cui, Y., Sano, Y., Tanaka, T., et al. (2000). Large-scale cDNA analysis reveals phased gene expression patterns during preimplantation mouse development. Development 127, 1737–1749. Lang, J. E., Magbanua, M. J., Scott, J. H., Makrigiorgos, G. M., Wang, G., Federman, S., Esserman, L. J., Park, J. W., and Haqq, C. M. (2009). A comparison of RNA amplification techniques at sub-nanogram input concentration. BMC Genomics 10, 326. Lequarre, A. S., Traverso, J. M., Marchandise, J., and Donnay, I. (2004). Poly(A) RNA is reduced by half during bovine oocyte maturation but increases when meiotic arrest is maintained with CDK inhibitors. Biol. Reprod. 71, 425–431. Li, C., and Hung Wong, W. (2001). Model-based analysis of oligonucleotide arrays: Model validation, design issues and standard error application. Genome Biol. 2, RESEARCH0032.1–RESEARCH0032.11. Mamo, S., Gal, A. B., Bodo, S., and Dinnyes, A. (2007). Quantitative evaluation and selection of reference genes in mouse oocytes and embryos cultured in vivo and in vitro. BMC Dev. Biol. 7, 14. Mortazavi, A., Williams, B. A., McCue, K., Schaeffer, L., and Wold, B. (2008). Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat. Methods 5, 621–628.

Interrogating the Transcriptome of Oocytes and Preimplantation Embryos

509

Murchison, E. P., Stein, P., Xuan, Z., Pan, H., Zhang, M. Q., Schultz, R. M., and Hannon, G. J. (2007). Critical roles for Dicer in the female germline. Genes Dev. 21, 682–693. Nagy, A., Gertsenstein, M., Vintersten, K., and Behringer, R. (eds.), (2003). Manipulating the Mouse Embryo a Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Oh, B., Hwang, S. Y., Solter, D., and Knowles, B. B. (1997). Spindlin, a major maternal transcript expressed in the mouse during the transition from oocyte to embryo. Development 124, 493–503. Oh, B., Hwang, S., McLaughlin, J., Solter, D., and Knowles, B. B. (2000). Timely translation during the mouse oocyte-to-embryo transition. Development 127, 3795–3803. Olds, P. J., Stern, S., and Biggers, J. D. (1973). Chemical estimates of the RNA and DNA contents of the early mouse embryo. J. Exp. Zool. 186, 39–46. Paynton, B. V., Rempel, R., and Bachvarova, R. (1988). Changes in state of adenylation and time course of degradation of maternal mRNAs during oocyte maturation and early embryonic development in the mouse. Dev. Biol. 129, 304–314. Peaston, A. E., Evsikov, A. V., Graber, J. H., de Vries, W. N., Holbrook, A. E., Solter, D., and Knowles, B. B. (2004). Retrotransposons regulate host genes in mouse oocytes and preimplantation embryos. Dev. Cell 7, 597–606. Piko, L., and Clegg, K. B. (1982). Quantitative changes in total RNA, total poly(A), and ribosomes in early mouse embryos. Dev. Biol. 89, 362–378. Pique, M., Lopez, J. M., Foissac, S., Guigo, R., and Mendez, R. (2008). A combinatorial code for CPE-mediated translational control. Cell 132, 434–448. Pirooznia, M., Nagarajan, V., and Deng, Y. (2007). GeneVenn—A web application for comparing gene lists using Venn diagrams. Bioinformation 1, 420–422. Rambhatla, L., Patel, B., Dhanasekaran, N., and Latham, K. E. (1995). Analysis of G protein alpha subunit mRNA abundance in preimplantation mouse embryos using a rapid, quantitative RT-PCR approach. Mol. Reprod. Dev. 41, 314–324. Rothstein, J. L., Johnson, D., DeLoia, J. A., Skowronski, J., Solter, D., and Knowles, B. (1992). Gene expression during preimplantation mouse development. Genes Dev. 6, 1190–1201. Rothstein, J. L., Johnson, D., Jessee, J., Skowronski, J., DeLoia, J. A., Solter, D., and Knowles, B. B. (1993). Construction of primary and subtracted cDNA libraries from early embryos. Methods Enzymol. 225, 587–610. Salisbury, J., Hutchison, K. W., Wigglesworth, K., Eppig, J. J., and Graber, J. H. (2009). Probe-level analysis of expression microarrays characterizes isoform-specific degradation during mouse oocyte maturation. PLoS One 4, e7479. Sharov, A. A., Piao, Y., Matoba, R., Dudekula, D. B., Qian, Y., VanBuren, V., Falco, G., Martin, P. R., Stagg, C. A., Bassey, U. C., Wang, Y., Carter, M. G., et al. (2003). Transcriptome analysis of mouse stem cells and early embryos. PLoS Biol. 1, E74. Stahlberg, A., Hakansson, J., Xian, X., Semb, H., and Kubista, M. (2004). Properties of the reverse transcription reaction in mRNA quantification. Clin. Chem. 50, 509–515. Stangegaard, M., Dufva, I. H., and Dufva, M. (2006). Reverse transcription using random pentadecamer primers increases yield and quality of resulting cDNA. Biotechniques 40, 649–657. Sternlicht, A. L., and Schultz, R. M. (1981). Biochemical studies of mammalian oogenesis: Kinetics of accumulation of total and poly(A)-containing RNA during growth of the mouse oocyte. J. Exp. Zool. 215, 191–200. Su, Y. Q., Sugiura, K., Woo, Y., Wigglesworth, K., Kamdar, S., Affourtit, J., and Eppig, J. J. (2007). Selective degradation of transcripts during meiotic maturation of mouse oocytes. Dev. Biol. 302, 104–117.

510

Anne E. Peaston et al.

Subkhankulova, T., and Livesey, F. J. (2006). Comparative evaluation of linear and exponential amplification techniques for expression profiling at the single-cell level. Genome Biol. 7, R18. Tam, O. H., Aravin, A. A., Stein, P., Girard, A., Murchison, E. P., Cheloufi, S., Hodges, E., Anger, M., Sachidanandam, R., Schultz, R. M., and Hannon, G. J. (2008). Pseudogenederived small interfering RNAs regulate gene expression in mouse oocytes. Nature 453, 534–538. Tanaka, T. S., Jaradat, S. A., Lim, M. K., Kargul, G. J., Wang, X., Grahovac, M. J., Pantano, S., Sano, Y., Piao, Y., Nagaraja, R., Doi, H., Wood, W. H., III, et al. (2000). Genome-wide expression profiling of mid-gestation placenta and embryo using a 15,000 mouse developmental cDNA microarray. Proc. Natl. Acad. Sci. USA 97, 9127–9132. Tang, F., Barbacioru, C., Wang, Y., Nordman, E., Lee, C., Xu, N., Wang, X., Bodeau, J., Tuch, B. B., Siddiqui, A., Lao, K., and Surani, M. A. (2009). mRNA-Seq wholetranscriptome analysis of a single cell. Nat. Methods 6, 377–382. Van Gelder, R. N., von Zastrow, M. E., Yool, A., Dement, W. C., Barchas, J. D., and Eberwine, J. H. (1990). Amplified RNA synthesized from limited quantities of heterogeneous cDNA. Proc. Natl. Acad. Sci. USA 87, 1663–1667. Wang, Q. T., Piotrowska, K., Ciemerych, M. A., Milenkovic, L., Scott, M. P., Davis, R. W., and Zernicka-Goetz, M. (2004). A genome-wide study of gene activity reveals developmental signaling pathways in the preimplantation mouse embryo. Dev. Cell 6, 133–144. Watanabe, T., Totoki, Y., Toyoda, A., Kaneda, M., Kuramochi-Miyagawa, S., Obata, Y., Chiba, H., Kohara, Y., Kono, T., Nakano, T., Surani, M. A., Sakaki, Y., et al. (2008). Endogenous siRNAs from naturally formed dsRNAs regulate transcripts in mouse oocytes. Nature 453, 539–543. West, M. F., Verrotti, A. C., Salles, F. J., Tsirka, S. E., and Strickland, S. (1996). Isolation and characterization of two novel, cytoplasmically polyadenylated, oocyte-specific, mouse maternal RNAs. Dev. Biol. 175, 132–141. Zeng, F., and Schultz, R. M. (2003). Gene expression in mouse oocytes and preimplantation embryos: Use of suppression subtractive hybridization to identify oocyte- and embryospecific genes. Biol. Reprod. 68, 31–39.

C H A P T E R

T W E N T Y- F I V E

Gene Expression Profiling of Mouse Embryos with Microarrays Alexei A. Sharov, Yulan Piao, and Minoru S. H. Ko Contents 1. Introduction 2. Considerations for Methods of Gene Expression Profiling 2.1. Noncoding RNAs, microRNAs, and proteins 2.2. Spatial resolution in complex tissues and organs 2.3. Assessing absolute mRNA abundance with RNA-seq 3. Experimental Strategies 3.1. Check first existing data sets in the public database 3.2. Analysis of embryonic materials without experimental manipulation 3.3. Analysis of embryonic materials after experimental manipulation 3.4. Number of replications required for the microarray analysis 4. Expression Profiling of Small Amounts of RNAs 4.1. Protocol 5. QC of Microarray Results 5.1. Rank-Plot 5.2. Correlation between replications 5.3. Calibration of microarray signal intensities 5.4. Cooperative hybridization issues 6. Analysis and Interpretation of Microarray Data 6.1. Statistical analysis of microarray data 6.2. Finding a set of statistically significant genes 6.3. Principal component analysis 6.4. Functional annotation: Gene ontology, pathways, and transcription factor binding sites 7. Submitting the Data to the Public Database Acknowledgments References

512 513 513 513 514 515 515 515 517 519 520 521 526 526 529 529 530 531 531 532 533 535 536 537 537

Developmental Genomics and Aging Section, Laboratory of Genetics, National Institute on Aging, NIH, Baltimore, Maryland, USA Methods in Enzymology, Volume 477 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)77025-7

511

512

Alexei A. Sharov et al.

Abstract Global expression profiling by DNA microarrays provides a snapshot of cell and tissue status and becomes an essential tool in biological and medical sciences. Typical questions that can be addressed by microarray analysis in developmental biology include: (1) to find a set of genes expressed in a specific cell type; (2) to identify genes expressed commonly in multiple cell types; (3) to follow the timecourse changes of gene expression patterns; (4) to demonstrate cell’s identity by showing similarities or differences among two or multiple cell types; (5) to find regulatory pathways and/or networks affected by gene manipulations, such as overexpression or repression of gene expression; (6) to find downstream target genes of transcription factors; (7) to find downstream target genes of cell signaling; (8) to examine the effects of environmental manipulation of cells on gene expression patterns; and (9) to find the effects of genetic manipulation in embryos and adults. Here, we describe strategies for executing these experiments and monitoring changes of cell state with gene expression microarrays in application to mouse embryology. Both statistical assessment and interpretation of data are discussed. We also present a protocol for performing microarray analysis on a small amount of embryonic materials.

1. Introduction Microarray technology is a high-throughput experimental tool to obtain the expression levels of essentially all genes by quantitating their transcripts (RNAs) amounts. Resultant expression profiles of cells and tissues provide snapshots of cell status, thereby uncover molecular signatures of various cell types and tissues, and explain embryo development by gene expression regulations. As numerous excellent reviews and technical guides in gene expression profiling technologies have already been published, adding another method article to this vast amount of literatures is hard to justify. Nevertheless, we have written this chapter with a hope that our hands-on experience in applying the technologies to mouse embryology may still be useful to the research community, as our lab has a long-standing interest in the global expression profiling (Ko, 1990, 2001, 2006) and has analyzed mouse embryos and cell cultures with more than 2000 microarrays for the past 10 years. However, our experiences are limited to cDNA clone-spotted nylon membrane arrays (Tanaka et al., 2000) and Agilent Technologies’ in situ synthesized 60-mer oligonucleotide glass slide microarrays. Our lab has indeed designed oligonucleotide sequences of Agilent’s first mouse microarray (‘‘mouse development array’’ (Carter et al., 2003), updated version (Carter et al., 2005)) in collaboration. Here, we describe the methods that our lab has been routinely using with the Agilent microarrays. Although there are platform-specific issues, we believe that most of the methods and

Gene Expression Profiling

513

considerations described here can be applied to other platforms. Affymetrix platform has been covered in detail in previous chapter in this volume (Hipp and Atala, 2006). Following the convention, gene sequences placed on microarrays are called ‘‘probes,’’ whereas RNA sequences labeled and hybridized to probes are called ‘‘targets’’ throughout this chapter.

2. Considerations for Methods of Gene Expression Profiling Most frequently performed expression profiling is for protein-coding RNAs, that is, transcripts from DNA sequences commonly called genes. A variety of methods has been developed and used, including traditional Northern blotting and quantitative RT-PCR, and state-of-the-art RNAseq (discussed below). As its running cost has been significantly reduced lately, microarray-based expression profiling can be a routine method of choice. However, there are some situations where the standard microarray technology does not seem to be best suited. Let us discuss these situations in the following sections.

2.1. Noncoding RNAs, microRNAs, and proteins Noncoding RNAs (ncRNAs), including both long ncRNAs (Wilusz et al., 2009) and microRNAs (miRNAs) (Olena and Patton, 2010) that regulate the expression of target mRNAs, are increasingly important but requires specialized techniques to quantitate for miRNAs (Thomson et al., 2007) and specialized microarray for long ncRNAs (Babak et al., 2005). However, as ncRNAs often affect the levels of coding RNAs, regular microarray-based expression profiling can still provide the overall status of cells and tissues. Similarly, protein profiling begins to be used as a way to examine the status of the cells; however, considering the changes of protein profiles also affect the global profile of coding RNAs, microarray-based profiling of RNAs often provides sufficient information about the status of cells and tissues.

2.2. Spatial resolution in complex tissues and organs One important issue to consider is the poor spatial resolution by microarray technology: gene expression patterns in complex tissues and organs, which consist of many different cell types, are difficult to assess. One way is to isolate specific cells in tissues/organs by microdissection (manually or lasercaptured) or FACS-sorting, and then carry out the microarray analysis. This usually requires the microarray analysis of small amount of materials (see Section 4). For example, primordial germ cells (PGC) are usually extracted

514

Alexei A. Sharov et al.

by dissociating the embryonic gonads into individual cells by trypsin, staining with germ cell-specific marker, and subsequent sorting (Abe et al., 1998; McCarrey et al., 1987). Another possible approach is gene expression profiling of individual cells of dissociated embryo, and then reconstructing cell type from gene expression. For example, it was possible to classify cells from the blastocyst into trophectoderm and ICM groups on the basis of their gene expression (Kurimoto et al., 2006). However, gene expression profiling of individual cells is challenging (see Section 4). Alternatively, a large number of genes selected after carrying out the microarray analysis can be examined by the whole mount in situ hybridizations (WISH). For example, WISH analyses have shown the localization of transcripts for nearly 100 genes in mouse blastocysts (Yoshikawa et al., 2006) and those for nearly 250 genes in mouse ES cell colonies (Carter et al., 2008).

2.3. Assessing absolute mRNA abundance with RNA-seq Microarray technology is reliable for comparison of the expression of the same gene in various cell types or various conditions. However, it is not best suited to assess absolute mRNA abundance and compare the expression levels of different genes in the same cell type, because the intensity of the signal can vary depending on oligonucleotide probes. For example, signal intensity in microarrays depends strongly on the location of the oligonucleotide probe within the transcript. Because the process of labeling starts from polyA tails, probes located far from the 30 -end of the transcript often show weaker signals than those located near the 30 -end. Other factors affecting signal intensity include differential efficiency of target amplification, nucleotide composition of probes, nonspecific hybridization, and cooperative hybridization (see Section 5.4). Recent advances in deep sequencing technologies (Mardis, 2008) have led to the emergence of the RNA-seq method, which yields reliable estimates of absolute mRNA abundance (Cloonan et al., 2008; Marioni et al., 2008; Mortazavi et al., 2008; Sultan et al., 2008). This method is based on sequencing of short cDNA fragments using the Illumina Genome Analyzer (Mortazavi et al., 2008), Applied Biosystems SOLiD technology, which is based on ‘‘sequencing by ligation’’ chemistry (Cloonan et al., 2008), or Roche 454 Life Science. Sequence tags are aligned back to the transcriptome to identify corresponding genes. As the RNA-seq becomes more affordable, it is often viewed as potentially a better method for gene expression profiling than microarrays (Oshlack and Wakefield, 2009; Wang et al., 2009). Furthermore, transcriptomes that can be measured by the RNA-seq are potentially broader than those measured by microarrays, as the detection is not limited to probes present on microarrays. However, several limitations of the RNA-seq method make it unlikely to replace microarrays in the near future. First, because the number

Gene Expression Profiling

515

of sequenced tags per gene is proportional to transcript length, short transcripts are represented by a small number of tags. As a result, the method has a ‘‘transcript-length bias,’’ which reduces statistical power to detect differential expression of short mRNAs compared with long ones (Oshlack and Wakefield, 2009). In contrast, microarrays have a uniform statistical power for both short and long transcripts. A similar problem may exist for lowexpressed genes, which are also represented by a small number of tags in RNA-seq. Second, reliable measurement of mRNA abundance in mammals requires large RNA-seq data sets ranging from 40 to 100 million tags (Cloonan et al., 2008; Mortazavi et al., 2008). As a result, the cost for RNA-seq experiment is currently 100-fold higher than that for microarrays. Although both technologies will become less expensive in the future, the cost ratio may not change much. Third, RNA-seq is substantially more labor-intensive and time-consuming. Finally, microarray data can be easily adjusted to represent the absolute abundance of mRNA: by doing RNA-seq and microarray analyses on the same RNA sample just once, adjustment coefficients can be estimated and then applied to all other microarray results.

3. Experimental Strategies 3.1. Check first existing data sets in the public database Before embarking on the microarray analysis, one should check whether the same or similar type of study has already been done or not. Vast amount of data have already been made available in the public database, for example, GEO (Barrett et al., 2009; http://www.ncbi.nlm.nih.gov/geo/) and ArrayExpress (Parkinson et al., 2009; http://www.ebi.ac.uk/microarray-as/ae/), so one may find necessary information without doing actual expression profiling. This is particularly relevant if a goal is to find the expression levels of genes of interest in some specific tissues. Global expression profiles of many tissues and organs can also be found in GNF database (Su et al., 2002; http://biogps. gnf.org/) and NIA database (GEO accession number GSE19806; http:// lgsun.grc.nia.nih.gov/exatlas/).

3.2. Analysis of embryonic materials without experimental manipulation The most common applications of microarray analysis are to obtain global gene expression profiles of embryonic tissues and cell cultures. Here are some typical questions to be addressed.  

Find a set of genes expressed in a specific cell type. Identify genes expressed commonly in multiple cell types.

516

Alexei A. Sharov et al.



Follow the time-course changes of gene expression patterns. For example, gene expression during preimplantation period has been studied by comparing RNA samples from unfertilized eggs, zygotes, two-cell embryos, fourcell embryos, eight-cell embryos, morulae, and blastocysts (Hamatani et al., 2004a; Wang et al., 2004; Zeng et al., 2004) (Fig. 25.1). Expected results from descriptive series include waves of gene activation at various time points of development (Cui et al., 2007; Hamatani et al., 2004a), or organ-specific profiles of gene expression (Frankenberg et al., 2007; Sherwood et al., 2007).  Demonstrate cell’s identity by showing similarities or differences among two or multiple cell types. It is worth mentioning that this task is not that simple as it appears to be. Similarities or differences can always be defined relatively. Only way to show how similar (or even identical) two cell types are to show that differences of global expression profiles are smaller than the variations among the same cell types from different samples or sources. For example, it is concluded that mouse embryonic stem (ES) cells and embryonic germ (EG) cells are indistinguishable in terms of gene expression profiles by showing that the differences between ES cells and EG cells are smaller than those between ES cells derived from different mouse strains or EG cells derived from different mouse strains (Sharova et al., 2007).

1.5 1.0 0.5 0.0 −0.5 −1.0 2 1 0

12,000

−1

10,000

−2 Unfertilized Fertilized

8000 6000 2-cell

4000

4-cell 8-cell

2000 Morula Blasto

Figure 25.1 An example of a bird’s-eye view of gene expression changes during mouse preimplantation development. Adapted from Ko (2006).

Gene Expression Profiling

517

3.3. Analysis of embryonic materials after experimental manipulation Microarray analysis of transgenic and gene-targeted mice has become a routine. In addition to the analyses of morphological and physiological defects, global gene expression profiling provides comprehensive pictures related to the effects of individual genes (Chen et al., 2007; Zhu et al., 2007). For example, gene-targeting often results in embryonic lethality but the cause of embryo death remains unknown. Gene expression profiling may elucidate gene regulatory problems, which eventually causes the death of embryo or adult mice (van Loo et al., 2007). Here are some typical questions to be addressed.     

Find regulatory pathways and/or networks affected by gene manipulations, such as overexpression or repression of gene expression. Find downstream target genes of transcription factors. Find downstream target genes of cell signaling. Examine the effects of environmental manipulation of cells on gene expression patterns. Find the effects of genetic manipulation in embryos and adults.

3.3.1. Common issues The main problem of aforementioned strategy is that results are not always interpretable in terms of cause–effect relationship. Observed changes in the phenotype and gene expression may be mediated by multiple intermediate steps and have little to do with the gene itself. Most of these changes may be compensatory responses of an organism to functional deficiency caused by gene suppression. This problem is most serious for housekeeping genes that are expressed constitutively, but tissue-specific and/or transiently expressed genes also suffer from the lack of causative-relationship due to functional compensation. 3.3.2. Possible solution: Use of normal-looking embryos at the early time points Obviously, one possible approach to this problem is to perform microarray analyses on embryos or embryonic tissues at earlier stage before the manifestation of phenotypic defects. Often, normal-looking transgenic or genetargeted embryos show the altered gene expression profiles, which provide great insights into the function of the gene of interest (e.g., Landry et al., 2008). Even in this case, finding the right control wild-type embryos may not be trivial, because genetic alteration may delay or accelerate the development. Thus, comparison with wild-type embryos collected at multiple time points may be required (Zhu et al., 2007). Cell culture system may be used as an alternative way to study the effects of gene manipulation. Often, it is easier to make ES cells with null alleles rather

518

Alexei A. Sharov et al.

than to generate nullizygous animals. Differentiation of mouse ES cells in various conditions (e.g., in the absence of leukemia inhibitory factor (LIF), in the presence of retinoic acids, or DMSO) mimics well certain events in the early embryogenesis. For example, the 3rd day of culture in LIF() conditions corresponds well to gastrulation stage in terms of gene expression patterns (Sharova et al., 2007). 3.3.3. Possible solution: Profiling the immediate effects of gene manipulation Alternatively, it is possible to design a controllable gene suppression or activation in the whole organism or in a specific organ. This is the most reliable experimental method that allows one to perform microarray analysis immediately after gene manipulation, and thereby, differentiate between direct effects of the gene and subsequent compensatory responses of the organism. For example, the expression of a gene can be altered via a transgene under the control of a promoter with a specific hormone- or tetracycline-responsive element (Bockamp et al., 2008; Lewandoski, 2001). An endogenous gene can be repressed by the conditional gene-targeting (both alleles) or via the RNAi technology. If a promoter of a transgene is organ-specific, then it is induced only in that organ, which makes it possible to avoid indirect effects mediated by gene activation in other organs. An alternative approach is to flank a transgene with LoxP sites, so that it can be easily excised when necessary. Gene manipulation can be used in cell culture system as well (Matsumoto et al., 2006; Nishiyama et al., 2009; Niwa et al., 2000). Manipulation of environmental factors (e.g., growth factors, signaling pathway inhibitors) can also provide tools to investigate the immediate effects of the manipulation on global expression profiles (Jung et al., 2007; Kitaya et al., 2007). For example, mesoderm development can be induced by TGFB factors (Kimelman, 2006; LaGamba et al., 2005), whereas neural differentiation can be induced by retinoic acid (Aiba et al., 2006; Wang et al., 2005; Williams et al., 2004). If a goal is to identify direct targets of a manipulated transcription factor, we recommend including early time points (e.g., 6, 12, and 24 h) after gene manipulation. These early time points can help to differentiate between direct and indirect effects in gene expression change (Masui et al., 2007). Later time points (e.g., 48 and 72 h) are also important, because activation of target genes usually progresses over time and is easier to detect at later time points. Nonmanipulated embryos (or cell culture) should be always used as additional controls, which are sampled at the same time points as manipulated organisms. Nonmanipulated control is important, because gene expression profiles often change over time even in control embryos or cell cultures. It is thus important to distinguish changes induced by manipulated gene or environmental factors from natural progression in gene expression.

Gene Expression Profiling

519

3.4. Number of replications required for the microarray analysis Planning the number of replications for gene expression profiling is a crucial step that determines the success of the project. Here by replications we mean true biological replications that are handled independently from a start to an end. Samples should thus be collected from different embryos or even from different mothers. In the case of cell culture, cells should be cultured independently. Although in the case of small amount of material, it is acceptable to pool samples from multiple embryos; in this case, the information on variability of gene expression between individuals will be lost. Thus, if possible it is better not to pool samples. Three main factors affect the required number of replications: (1) noise levels in microarray measurements, (2) expected difference in gene expression, and (3) the number of factors (e.g., cell types, time points, or organs) used in the experiment. Noise levels depend on microarray platforms and on the consistency of RNA processing from extraction to hybridization. Old type of nylon membrane arrays with spotted cDNA clones have shown high levels of noise (Tanaka et al., 2000). Currently, three microarray technologies are available, all of which are based on synthetic short oligonucleotide DNAs: (1) synthesis by photolithography for Affymetrix and NimbleGen; (2) synthesis by ink-jet printing for Agilent Technologies; and (3) beads with oligonucleotides for Illumina (Kawasaki, 2006). All of them generate highly reproducible results (Cheadle et al., 2007). Systematic studies by MicroArray Quality Control (MAQC) consortium have shown that these platforms provide comparable results (Shi et al., 2006) and the reliability of the results has been validated by other methods, such as qPCR (Canales et al., 2006). For Agilent microarrays, competitive hybridization of two RNA samples labeled with Cy3 and Cy5 dyes, respectively, increases the sensitivity of arrays and makes signal normalization more accurate (Hughes et al., 2001). If samples used for competitive hybridization represent experiment and control (e.g., gene-targeted embryos vs. wild-type embryos), then replications with dye swap are needed to eliminate possible effect of preferential hybridization with one of the dyes (Fig. 25.2A). However, it is better to use standard RNA sample (e.g., universal mouse reference, UMR) in all arrays so that they can be normalized on the basis of that common sample (Fig. 25.2B). A potential problem of using UMR is if UMR does not contain transcripts of all genes, this may lead to biased normalization (see Sections 4.1.2 and 5.4). If expected differences in gene expression are large (e.g., >100 genes with >3-fold change in expression), then three replications are usually sufficient for pair-wise comparison. But if differences in gene expression are small (e.g., <50 genes with > 1.5-fold change), then the number of replications should be increased. Statistical models are available for planning the number of replication on the basis of expected statistical power

520

Alexei A. Sharov et al.

A

Cy3labeling

Microarray

Cy5labeling

Cy5labeling

Cy3labeling

B Cy3labeling

Microarray

Cy5labeling

Universal reference RNAs

Figure 25.2 Design of two-color (dye) microarray experiments. (A) Experiment and control RNAs are hybridized to the same arrays with Dye Swap. (B) Universal reference RNAs are used as common controls for all experiments.

(Pan et al., 2002). The number of replications can be reduced in experiments that involve a large number of cell types and/or time points, because the error variance can be estimated from all cell types or time points simultaneously using ANOVA (Sharov et al., 2005b). However, two replications are still needed for reliable measurement of gene expression. Single replication may be acceptable for redundant cell/tissue types. For example, if three different types of controls are expected to have similar gene expression profiles, it may be possible to do just one replication of each. In general, it is recommended that at least one extra sample should be prepared as a possible substitute in case they fail during target preparation steps.

4. Expression Profiling of Small Amounts of RNAs Working with embryos requires the expression profiling of small amounts of RNAs. Various protocols for microarray analysis for cells as few as single cells have been reported (e.g., Kurimoto et al., 2006). However,

521

Gene Expression Profiling

these methods are usually labor-intensive and technically demanding. Due to their relatively low reproducibility, many replications are also required. A protocol described here has been routinely used in our laboratory to perform microarray analysis of oocytes and preimplantation embryos (Hamatani et al., 2004a,b). The protocol can be done relatively easily with a few extra days in target preparations and can be applied to 2 ng total RNAs reliably and reproducibly (Carter et al., 2003). A single oocyte, which contains 0.2–0.4 ng total RNAs (Nagy et al., 2002), could be used; however, it does not work routinely and the preparation of labeled targets have to be repeated many times until it passes QC described below. A pool of 10 oocytes (2 ng total RNAs) works routinely and produces the satisfactory results in terms of reproducibility and sensitivity. This protocol also works for microdissected tissues. Below is a step-bystep protocol.

4.1. Protocol 4.1.1. Cy3-labeled target preparation with two-round RNA amplification This protocol uses Agilent Low RNA Input Fluorescent Linear Amplification Kit (Cat. # 5184-3523) with the modification to allow two rounds of amplification to obtain enough quantity of labeled targets (Hopkins et al., 2003). We have made a slight modification to the protocol to obtain reproducible results from one to 10 oocytes or embryos. One of the key points is direct lysis of eggs/embryos/tissues collected in M2 medium in the primer annealing buffer without isolating total RNAs or polyAþ RNAs. Although use of single oocytes or preimplantation embryo can produce the expression profiles, use of 10 oocytes or embryos are strongly recommended as they produce high-quality and reproducible data consistently. 4.1.1.1. cDNA synthesis directly from embryos  Annealing reaction

Nuclease-free water 1–10 oocytes or embryos in M2 medium 5% NP-40 T7 promoter primer eightfold dilution Total volume

7.3 l 1.0 l 2.0 l 1.2 l 11.5 l

Incubate at 65  C for 10 min and place on ice for 5 min. 

cDNA reaction

522

Alexei A. Sharov et al.

Annealing reaction from above 5 first strand buffer 0.1 M DTT dNTP Mix (10 mM) MMLV RT RNase Out Total volume

11.5 l 4.0 l 2.0 l 1.0 l 1.0 l 0.5 l 20.0 l

Incubate at 40  C for 2 h, then inactivate at 65  C for 15 min, and place a tube on ice for 5 min. 4.1.1.2. Amplification—Round 1

cDNA reaction from above Nuclease-free water 4 transcription buffer 0.1 M DTT NTP Mix CTP 50% PEG RNase Out Inorganic pyrophosphatase T7 RNA polymerase Total volume

20.0 l 12.1 l 20.0 l 6.0 l 8.0 l 5.6 l 6.4 l 0.5 l 0.6 l 0.8 l 80.0 l

Incubate at 40  C for 4 h. Purify by Qiagen RNeasy Mini Spin Column, elute in 30 l of water, and dry in SpeedVac to a final volume of 10.5 l in water. 4.1.1.3. cDNA synthesis from cRNA  Annealing reaction—First strand synthesis

cRNA solution from above Random hexamers 10-fold dilution Total volume

10.5 l 1.0 l 11.5 l

Incubate at 65  C for 10 min and place on ice for 5 min. 

cDNA reaction—First strand synthesis Annealing reaction from above 5 first strand buffer 0.1 M DTT dNTP Mix (10 mM)

11.5 l 4.0 l 2.0 l 1.0 l

523

Gene Expression Profiling

1.0 l 0.5 l 20.0 l

MMLV RT RNase Out Total volume Incubate at 40  C for 2 h. 

Annealing reaction—Second strand synthesis 20.0 l 1.2 l 21.2 l

cDNA solution from above T7 promoter primer (no dilution) Total volume

Incubate at 65  C for 10 min and place on ice for 5 min. 

cDNA reaction—Second strand synthesis 21.2 l 5.8 l 4.0 l 4.0 l 2.0 l 2.0 l 1.0 l 40.0 l

Annealing reaction from above Nuclease-free water 5 first strand buffer 0.1 M DTT dNTP Mix (10 mM) MMLV RT RNase Out Total volume

Incubate at 40  C for 2 h. Inactivate at 65  C for 15 min and place on ice for 5 min. Dry in SpeedVac to a final volume of 35.3 l in water. 4.1.1.4. Amplification—Round 2 and Cy3 labeling

cDNA reaction from above CTP-Cy3 4 transcription buffer 0.1 M DTT NTP Mix 50% PEG RnNase Out Inorganic pyrophosphatase T7 RNA polymerase Total volume

35.3 l 2.4 l 20.0 l 6.0 l 8.0 l 6.4 l 0.5 l 0.6 l 0.8 l 80.0 l

524

Alexei A. Sharov et al.

Incubate at 40  C for 4 h. Purify by Qiagen RNeasy Mini Spin Column and elute in 60 l (2  30 l) water. 4.1.2. Cy5-labeled target preparation from universal reference RNAs Cy5-CTP-labeled reference target is prepared using the Agilent Low RNA Input Fluorescent Linear Amplification Kit (Cat. # 5184-3523) from the mixture of Stratagene UMR RNA and total RNAs from mouse ES cells cultured in the standard condition (LIFþ medium) at 2:1 ratio. 2.5 g mixed RNA is used for labeling in each tube. UMR is composed of total RNA from 11 mouse cell lines. We have added mouse ES cell RNAs as we find that genes specifically expressed in ES cells and early embryos are less represented in the UMR, which could cause some distortion of the expression data of ES-specific genes (see Section 3.4). 4.1.3. Purification of Cy3 and Cy5 targets Labeled targets are purified using the RNeasy Mini Kit (Qiagen, Cat. # 74104), and then quantified using a NanoDrop scanning spectrophotometer (NanoDrop Technologies). Purified samples can be stored at  80  C for long time. We have had experience in using labeled samples stored at 80  C for up to 8 years without any problems. 4.1.4. Quality control (QC) of the labeled targets It is important to check the quality of labeled targets before doing hybridizations. In the standard procedure with 2.5 g of total RNAs as a starting material and one round of amplification, usually 50 g of labeled targets will be produced and only 825 ng of them will be used for hybridization as indicated in the manufacture’s protocol. In the case of two-round amplification, a pool of 10 oocytes will produce  20 g of labeled targets in 30 l solution. However, this does not mean that they are of high quality. We run 1 l of the solution in NanoDrop and quantify the sample. Figure 25.3 shows an example of standard targets (>0.1 peak height; usually 0.12–0.17). Targets prepared by two round of amplification from one oocyte to five oocytes show peak heights lower than 0.1, which usually do not produce high-quality microarray data. So, we usually do not proceed to hybridization. Targets prepared by two-round amplification from 10 oocytes produce the peak height higher than 0.1. Four micrograms of targets will be used for hybridization. 4.1.5. Microarray hybridizations, washing, and scanning Target cRNAs are hybridized to the NIA Mouse 44K Microarray v3.0 (whole genome 60-mer oligonucleotide arrays; Mouse Development Microarray Kit by Agilent Technologies 015087) according to manufacturer’s

0.00

0.20

0.20

0.18

0.18

0.18

0.16

0.16

0.16

0.14

0.14

0.14

0.12 0.10 0.08 0.06

0.20

1 mm absorbance

1 mm absorbance

0.10

1 mm absorbance

0.20

0.12 0.10 0.08 0.06

0.00

0.04

0.02

0.02

0.02

0.00

0.00 300

350 400 450 500 550 Wavelength nm

600 650

700

750

-0.02 220 250

0.00 300

350

400 450 500 550 Wavelength nm

600 650

700

750

-0.02 220 250

0.20

0.18

0.18

0.16

0.16

0.16

0.14

0.14

0.14

0.12 0.10 0.08 0.06

1 mm absorbance

0.20

0.18

0.12 0.10 0.08 0.06

0.06 0.04

0.02

0.02

0.00 400 450 500 550 Wavelength nm

4 oocytes

600 650

700

750

-0.02 220 250 300

750

0.08

0.02

350

700

0.10

0.04

0.00

600 650

0.12

0.04

300

350 400 450 500 550 Wavelength nm

2 oocytes

0.20

-0.02 220 250

300

1 oocyte

1 mm absorbance

1 mm absorbance

0.06

0.04

Typical targets

0.10

0.08

0.04

-0.02 220 250

0.20

0.12 0.10

0.00 350

400 450 500 550 Wavelength nm

600 650

700

750

-0.02 220 250

5 oocytes

Figure 25.3 Quality control of labeled targets by NanoDrop.

300

350

400 450 500 550 Wavelength nm

600

10 oocytes

650

700 750

526

Alexei A. Sharov et al.

protocol (Two-Color Microarray-Based Gene Expression Analysis Protocol, Product # G4140-90050, Version 5.0.1). Slides are scanned with the Agilent DNA Microarray Scanner (model G2505-64120) at 100% and 10% PMT in both channels, with a scan resolution of 5 m.

5. QC of Microarray Results Microarray technology is highly complex in terms of the number of processing steps and factors that may alter the results. It is thus important to have robust QC tools for testing final microarray results. Low-quality results often remain unnoticed, because the software used for the analysis of scanned images often masks the problem via various kinds of normalizations. Normalization algorithms are designed to improve the quality of results within normal limits of variation; however, when applied to strongly distorted data, they simply hide the problem. Agilent Feature Extraction Software provides several indicators of microarray quality, which include spatial distribution of outliers, array uniformity, sensitivity, and spike-ins signal statistics (http://www.chem.agilent.com/Library/usermanuals/Public/ ReferenceGuide_050416.pdf). These indicators are useful for detecting technical problems within a single array, such as scratches, printing errors, and signal intensity trends across the array (i.e., spatial nonuniformity). However, many problems cannot be detected within a single array and require additional QC tests. In this section, we describe two such tests that we have been routinely using: Rank-Plot and correlation between replications.

5.1. Rank-Plot Rank-Plot can be generated in the following manner: 

Step 1: Extract value from two columns, ‘‘gBGSubSignal’’ and ‘‘rBGSubSignal,’’ from output files of the Agilent Feature Extraction Software.  Step 2: Transform these values in log10 scale. Substitute them with 0, if negative. Name these values as ‘‘GVal’’ and ‘‘RVal.’’  Step 3: Sort ‘‘GVal’’ and ‘‘RVal,’’ respectively, and index each value with its corresponding rank.  Step 4: If there are multiple microarray data from the past experiments, all these data should also be analyzed in the same manner. Then, make an ensemble average of ‘‘GVal’’ and ‘‘RVal’’ for all the past experiments. Name these values ‘‘GVal_mean’’ and ‘‘RVal_mean.’’ Sort ‘‘GVal_mean’’ and ‘‘RVal_mean,’’ respectively, and index each value with its corresponding rank.

Gene Expression Profiling

527



Step 5: Plot the followings in (X, Y) coordinate: (Rank, GVal), (Rank, GVal_mean), (Rank, RVal), (Rank, RVal_mean).  Plotting can be done by Excel or Gnuplot (free software). The Rank-Plot method appears sensitive in detecting various problems with microarrays, which include too low or too high amounts of targets in hybridization reaction, compromised washing buffers, degradation of Cy5 signal due to ozone or UV light, and scanner malfunctions. It is important to use nonnormalized data for the Rank-Plot, because dye normalization may compensate signal intensities and mask the problem. We recommend using background-subtracted signal intensities (‘‘gBGSubSignal’’ and ‘‘rBGSubSignal’’) for both green and red signals. Signals are log-transformed, sorted by increasing order (sorting is done independently for the green and red signals), and then plotted against the rank (Fig. 25.4A and B). It appears that Rank-Plots have very consistent shapes for all high-quality microarrays and do not depend on the sample tested. Although different cell types have different composition of expressed genes, genes are sorted based on their expression. Thus, the order of genes in Rank-Plots is specific to the cell type, but the shape of the graph remains the same. As a base line, we used averaged Rank-Plots for a set of known high-quality arrays (‘‘RVal_mean’’ and ‘‘GVal-mean’’). Then, deviations from the base line indicate potential problems. If the Rank-Plot matches well to the base line, then we assume that microarrays are of good quality. Of course, the Rank-Plot represents only the distribution of the signal in microarrays. Thus, some specific problems like scratches and misprints should be detected using other quality-control indicators, as provided by the Agilent Feature Extraction Software. Competitive hybridization of two targets labeled with Cy3 (green) and Cy5 (red) dyes yields best results when both have comparable intensity. Thus, ideally Rank-Plots for red and green signals should match each other. In our case, red intensity is slightly higher than green intensity due to historically selected RNA doses which we decided not to change for better compatibility with previous results (Fig. 25.4A and B). The difference in red and green Rank-Plots is within acceptable range (less than twofold). Downward shift of the Rank-Plot relative to the base line indicates low signal intensity. Low signals are often caused by low amount of labeled RNA used for hybridization (Fig. 25.4C and D). We have observed this problem when the initial RNA sample is small and requires two rounds of amplification. Although the quantities of all probes are measured based on NanoDrop reads, targets with two rounds of amplifications apparently contain a larger portion of ‘‘junk RNA.’’ As a result, the amount of real RNA appears diluted. The Rank-Plot can be used for adjusting RNA amounts. For example, the green line in Fig. 25.4C, which represents the test sample labeled with Cy3 dye, is shifted down by 0.3 in log10 scale relative to the base line. Because 100.3 ¼ 2, this shift can be compensated if

528

Alexei A. Sharov et al.

A

6

Cy5 Avg (old Cy5 Avg (new Cy3 Avg (old Cy3 Avg (new

0

5000

Cy3 Cy5 UMR) UMR) UMR) UMR)

10,000

15,000

5000

20,000

25,000

30,000

35,000

40,000

Cy3 Cy5 UMR) UMR) UMR) UMR)

10,000

15,000

25,000

30,000

35,000

40,000

2

1

1

0 45,000

0

5000

10,000

15,000

0

5000

15,000

20,000

25,000

30,000

35,000

40,000

25,000

30,000

35,000

40,000

0

5000

10,000

15,000

30,000

35,000

40,000

35,000

40,000

0 45,000

Cy3 Cy5 UMR) UMR) UMR) UMR)

5

4

2

1

0 45,000

0

5000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

F

0 45,000

6

Cy5 Avg (old Cy5 Avg (new Cy3 Avg (old Cy3 Avg (new

Cy3 Cy5 UMR) UMR) UMR) UMR)

5

4

3

3

2

2

1

1

0 45,000

0

5000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

H

0 45,000

6

Cy5 Avg (old Cy5 Avg (new Cy3 Avg (old Cy3 Avg (new

Cy3 Cy5 UMR) UMR) UMR) UMR)

5

4

3

3

2

2

1

1

0 45,000

0

5000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

J

0 45,000

6

Cy5 Avg (old Cy5 Avg (new Cy3 Avg (old Cy3 Avg (new

4

25,000

30,000

1

5

20,000

25,000

3

6 Cy3 Cy5 UMR) UMR) UMR) UMR)

20,000

2

4

20,000

15,000

6

5

I Cy5 Avg (old Cy5 Avg (new Cy3 Avg (old Cy3 Avg (new

10,000

Cy5 Avg (old Cy5 Avg (new Cy3 Avg (old Cy3 Avg (new

4

Cy3 Cy5 UMR) UMR) UMR) UMR)

10,000

5000

3

6

Cy5 Avg (old Cy5 Avg (new Cy3 Avg (old Cy3 Avg (new

0

D

5

G

4

3

6 Cy3 Cy5 UMR) UMR) UMR) UMR)

5

2

4

20,000

Cy3 Cy5 UMR) UMR) UMR) UMR)

3

5

E Cy5 Avg (old Cy5 Avg (new Cy3 Avg (old Cy3 Avg (new

Cy5 Avg (old Cy5 Avg (new Cy3 Avg (old Cy3 Avg (new

4

6

0

6

5

C Cy5 Avg (old Cy5 Avg (new Cy3 Avg (old Cy3 Avg (new

B

Cy3 Cy5 UMR) UMR) UMR) UMR)

5

4

3

3

2

2

1

1

0 45,000

0

5000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

Figure 25.4 Quality control of microarray data by Rank-Plots.

0 45,000

Gene Expression Profiling

529

the amount of targets used for hybridization is doubled. However, if the signal is extremely low as in Fig. 25.4D, increasing the amount of targets would not help in our case. Red dye (Cy5) can be bleached in the presence of ozone (Byerly et al., 2009; Fare et al., 2003), which may result in the downward shift of the Rank-Plot (red line in Fig. 25.4E and F). Although microarray facilities are kept ozone-free, this problem appears occasionally due to neglect or malfunction of the ozone filters. (We have set up bioBubble enclosure surrounding the bench space for the microarray work and the microarray scanner. Air intakes are handled by the Ozone-scrubber from the ScieGene.) Deviation of the Rank-Plot from the base line can be restricted to low-intensity signals (Fig. 25.4G–J). Upward deviation in both channels indicates nonspecific hybridization (Fig. 25.4G and H). After we have observed it in a number of arrays, we have done comprehensive troubleshooting and eventually identified the manufactures’ problem in washing buffer as the cause. After replacing the buffer, the Rank-Plot results have returned to the base line. Occasionally, we have observed abnormal RankPlots for individual arrays (e.g., Fig. 25.4I and J). Because it is difficult to find the cause in each case, we usually fix the problem by repeating the hybridization or using a different RNA sample for analysis.

5.2. Correlation between replications Correlation of signal intensity between biological replications is another useful QC method. It is better to use dye-normalized signals for this test, because these values usually show better matches. High correlation between log-transformed signals for independent biological replications indicates both: (1) high quality of microarray results and (2) reproducibility of the experiment. Gene expression in cultured cells is highly reproducible and the correlation of log-transformed signals between biological replications is mostly >0.995. The reproducibility of gene expression in tissues and organs of different animals is lower than in cultured cells, and the correlation between biological replications usually ranges from 0.95 to 1. Thresholds for acceptable correlation between replications should be selected based on the goals of the study.

5.3. Calibration of microarray signal intensities Certain experiments require accurate scaling of gene expression values so that the ratio of hybridization signal intensities is equal to the ratio of mRNA abundance across the entire range of expression levels. For example, we have used microarrays for detecting the rate of mRNA degradation in ES cells (Sharova et al., 2009). If arrays were not well calibrated, then the estimates of the rate of mRNA decay would be biased. We have tested the

530

Alexei A. Sharov et al.

Number of points in regression 2

Slope of log Cy3 vs. log RNA input

1.6

3

4

1.4 1.2 1.0 0.8 0.6 0.4 0.2 00 0

1

2 3 4 Log10 Cy3 signal at normal RNA input

5

6

Figure 25.5 Calibration of microarray signal intensities.

scaling of Agilent microarrays using a series of dilutions (5, 1, 0.2, and 0.04) of the same mRNA pool obtained from newborn mice of C57BL/ 6J strain and labeled with Cy3 dye. Samples have been hybridized together with the constant amount of UMR labeled with Cy5. Regression of logtransformed Cy3 signals versus log RNA input was close to 1 for most probes with log10 signal >1.5 (Fig. 25.5), which indicates proper scaling. The results confirmed accurate calibration of intensity signals across the entire range of values. Only 50 (0.13% of total) oligonucleotide probes with high signals (>4.8) has shown decreased slopes, indicating some degree of saturation effects (Fig. 25.5).

5.4. Cooperative hybridization issues The test of microarray calibration (see above) has revealed an unexpected fact: red signals for >1000 oligonucleotide probes are not constant despite the fact that the same amount of UMR labeled with Cy5 is used for all microarrays. Instead, red signals in these probes correlated positively with the amount of test RNA labeled with Cy3. This result shows that instead of competitive hybridization some oligonucleotide probes show cooperative hybridization. The dependency between Cy5 signal and test RNA input, measured by the slope of the linear regression, has been observed mostly for oligonucleotides that had green signals higher than red signals (Fig. 25.6). The mechanism of this effect is unknown, but we hypothesize that it stems from the chain hybridization when free ends of Cy3-labeled mRNA bound to probes on the microarray can hybridize with Cy5-labeled mRNA either nonspecifically or via repetitive sequences. Cooperative hybridization may cause underestimation of the difference in gene expression between samples

531

Gene Expression Profiling

1

Slope of log Cy5 signal vs. test RNA input

0.8 0.6 0.4 0.2 0 −0.2 −0.4 −0.6 −0.8 −3

−2

−1

0

1

2

3

4

Mean log (Cy5) – mean log (Cy3)

Figure 25.6

Cooperative hybridization issues.

if values are normalized based on the UMR signal. To avoid biased estimates, we do not use UMR for normalization for a small portion of probes, where we expect cooperative hybridizations.

6. Analysis and Interpretation of Microarray Data 6.1. Statistical analysis of microarray data The main purpose of statistical analysis of microarray data is to find genes that are differentially expressed between either stages of development or manipulated and intact embryos. The specific feature of microarray analysis is to simultaneously test a very large set of genes. Most whole genome microarrays include >40,000 oligonucleotide probes that match to ca. 30,000 nonredundant transcripts. If a statistical significance is evaluated using p-value (p ¼ 0.05 is traditionally used for testing individual hypotheses), 2000 genes (5% of 40,000) will be considered statistically significant even in the case of no real difference in expression (e.g., if we compare identical samples). Obviously, such a high number of false positives are not acceptable; thus, correction for multiple hypotheses testing is always required in microarray analysis. The most commonly used criterion for multiple hypotheses testing is false discovery rate (FDR), which is interpreted as the proportion of false positives among all genes that are considered significant. According to the theorem of Benjamini and Hochberg (1995), FDR is estimated:

532

Alexei A. Sharov et al.

  pr N FDRi ¼ min ri r

ð25:1Þ

where r is the rank of a gene ordered by increasing p-values, pr is the p-value for gene with rank r, and N is the total number of genes tested. The FDR value increases monotonously with increasing p-value. As follows from Eq. (25.1), FDR becomes more stringent as the number of significant genes decreases. Microarray data often include a small number of replications; for example, a typical pair-wise comparison with three replications leaves only four degrees of freedom for the error variance in ANOVA. In this case, the estimates of error variance are unstable and may appear very small for some genes by pure chance, and these genes are then erroneously treated as statistically significant. This problem can be fixed by adjusting the error variance on the basis of additional information coming from other genes. For example, the Bayesian method takes a weighted average of actual error variance and the average error variance for other genes with similar intensity (Baldi and Long, 2001). A simpler and more conservative approach is to take the maximum value from the actual error variance and the average error variance for other genes with similar intensity (Sharov et al., 2005b). Another method is to use a foldchange threshold as a criterion of statistical significance in addition to FDR. Several software packages are available for the analysis of microarrays, including SAM and Bioconductor (Gentleman et al., 2004; Tusher et al., 2001). In this chapter, we focus on the software that we have developed earlier, NIA Array Analysis (http://lgsun.grc.nia.nih.gov/ANOVA; Sharov et al., 2005b), as it combines algorithms, which we think are most important for microarray analysis (i.e., ANOVA, FDR, adjustment of error variance, PCA, gene clustering, and pattern matching), and offers a user-friendly web-based interface.

6.2. Finding a set of statistically significant genes The first step in using the NIA Array Analysis software (Sharov et al., 2005b) is to create an account which is password-protected. Then, the input file with data should be prepared using instructions available in the help page (http://lgsun.grc.nia.nih.gov/ANOVA/help.html#format). You can assemble it in Excel and then save it as a tab-delimited text file. Another option is to use ‘‘Arrayjoin’’ tool, which is designed for compiling an input file from multiple scanner files. If column headers are formatted properly, then the software can automatically identify tissue/cell types and replications used in the experiment. However, this information can also be added or modified after the file is uploaded. Data can be normalized during

Gene Expression Profiling

533

uploading step, if necessary. In addition, a file with array annotation (formatting rules are specified in http://lgsun.grc.nia.nih.gov/ANOVA/help. html#annotations) can be uploaded if it is not already available for all users. Software can be used for both one-dye arrays and two-dye arrays. In the latter case, we assume that each array is represented by two columns: (1) sample data and (2) control or reference data (e.g., UMR). If your input data are log-transformed, then select the appropriate log-transformation type (log10, log2, or loge) before loading the file. Outliers are removed based on a user-selected z-threshold level (default threshold z ¼ 8). Statistical analysis is based on the single-factor ANOVA of log-transformed data with optional adjustment of error variance and FDR criterion for selecting differentially expressed genes. Significant genes are displayed using scatter-plot (Fig. 25.7A and B), log-ratio plot (Fig. 25.7C and D), and tables. It is possible to modify criteria for gene significance (e.g., use p-values instead of FDR or use a fold-change threshold), and regenerate scatter-plot and log-ratio plot. Additional methods of analysis include hierarchical clustering, correlation matrix, estimation of error function (error vs. log intensity), and principal component analysis (PCA, see below). Hierarchical clustering of tissues/cell types is done using the average distance method. It is also possible to identify genes that are specific for each cluster. Data can be retrieved by (1) searching for a specific gene name; (2) pattern matching, that is, searching genes whose expression change matches a specific pattern; (3) browsing tables with differentially expressed genes; (4) clicking on individual genes in scatter-plot or log-ratio plot; and (5) downloading a full tab-delimited table of ANOVA results which can be then opened in Microsoft Excel for any kind of custom-defined analysis.

6.3. Principal component analysis PCA is a powerful method for determining patterns of gene expression change in large data sets, which may include multiple time points, tissues, or genotypes. The strongest pattern of gene expression is represented by the 1st principal component (PC1). A large set of genes show high correlation with this pattern, and correlation can be either positive or negative. The 2nd most important pattern is represented by PC2, which is orthogonal to PC1 and hence have no correlation with it. There are a smaller number of genes that correlate with PC2 compared to genes that correlate with PC1. Then the 3rd principal component (PC3) can be determined, which is orthogonal to both PC1 and PC2, and so on. PCA results are often shown as a scatterplot with principal components (PCs) as axis. This plot can be viewed as a projection of original data onto a subspace of PCs. Thus, PCA is a method for dimensionality reduction. As we have no ability to navigate in 50 dimensions if the data matrix includes 50 samples, it is possible to extract two or three PCs and project the data into this two (2D)- or three-dimensional

534

Alexei A. Sharov et al.

C 5.5 5 4.5 4 3.5 3 2.5 2 1.5 1

Control

Control

Dox− 1

Dox+ 1

2 3 4 Log-intensity

Dox−

0.5 0 –0.5

Dox+ 1.5 2 2.5 3 3.5 4 4.5 5 5.5 Log-intensity

5

B

D 5.5 5 4.5

Klf4 Klf4 Dox−

1.5

4

2.5 2

0.5 0

–0.5

1.5 1

Dox−

1

3.5 3

Log-ratio

Log-intensity

Log-ratio

Log-intensity

A

Dox+ 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 Log-intensity

–1 –1.5

Dox+ 1.5 2 2.5 3 3.5 4 4.5 5 5.5 Log-intensity

Figure 25.7 Scatter-plot (A, B) and log-ratio plot (C, D) for ES cells in the presence (Dox) or absence (Doxþ) of overexpression of transcription factor (control and Klf4). Data from Nishiyama et al. (2009).

(3D) space which can be visualized. For example, analysis of gene expression change of differentiating mouse ES cell have revealed three cell lineages, primitive ectoderm, trophoblast, and extraembryonic endoderm, as cell lineage trajectories in 3D space (Fig. 25.8). PCA is done using the singular value decomposition (SVD) method that generates eigenvectors both for rows and columns of the log-transformed data matrix (Chapman et al., 2002; Sharov et al., 2005b). The advantage of SVD method is that it can project both rows and columns of the data matrix on the same coordinates represented by PCs (biplot); thus, the user can visually explore associations between genes and tissues or time points. The NIA Array Analysis software generates 2D and 3D (based on Virtual Reality Modeling Language (VRML)) biplots. All biplots (including 3D) are interactive; each gene is hyperlinked with its annotation and histogram that

535

Gene Expression Profiling

m er od

P4 Z2

al

Trophoblast

nd ni ce br yo m

N5 N6

Z1

ur

tra e

N4

ne —

Ex

N3

rm de to ec

G2 G3 G4 G5

Z0

N2 P0

ve iti

F3 G1 F4 F5

E1 N1

im

F2

N0

PC2

Pr

F0 F1 G0

Z3 TS Z4 Z5 PL

NS PC1

PC3

DC

Figure 25.8 3D PCA plots showing the locations of mouse ES cells during differentiation based on the microarray data. Adapted from Aiba et al. (2009).

shows details of expression pattern. For each PC, we identify two clusters of genes that are positively and negatively correlated with this PC. The degree of gene expression change within a specific PC is measured by the slope of regression of log-transformed gene expression versus the corresponding eigenvector multiplied by the range of values within the eigenvector. It is possible to save PCA results from one experiment and then use them for the analysis of another experiment as reference coordinate system. This method works best if both experiments used the same microarray platform, that is, all probe IDs are matching. Partial matching of probe IDs is also acceptable (e.g., if one microarray is an extension of the earlier array version). Use ‘‘same array platform’’ checkbox if both experiments were done with the same array platform; in this case, the software will not attempt to renormalize data sets. If array platforms are different, then PCA can be based on common gene symbols, GenBank accession number, or other identifiers. After PCA is generated, click the button ‘‘Save for import.’’ When you analyze the second data set, then saved files will appear in the select box (PCA menu). Saved files remain available for 1 day on the NIA Array Analysis.

6.4. Functional annotation: Gene ontology, pathways, and transcription factor binding sites Getting information on differentially expressed genes is only the first step of microarray analysis. The next step is to do functional annotation of these genes, which can help to interpret gene expression changes. The most common method of functional annotation is the use of Gene Ontology (GO) database, which specifies gene function based on literature and expert

536

Alexei A. Sharov et al.

opinions (http://www.geneontology.org; Ashburner et al., 2000). However, statistical analysis is needed to determine which GO terms are overrepresented within a gene list (e.g., among genes that are downregulated in a gene-targeted embryo). This analysis can be done using a variety of software tools, including GenMAPP (www.genmapp.org; Dahlquist et al., 2002), GSEA (http://www.broad.mit.edu/gsea; Kim and Volsky, 2005; Subramanian et al., 2005), GOminer (http://discover.nci.nih.gov/gominer; Zeeberg et al., 2003), or NIA Mouse Gene Index (http://lgsun.grc.nia.nih. gov/geneindex/mm8; Sharov et al., 2005a; Sharova et al., 2007). A list of gene symbols generated by the NIA Array Analysis software can be uploaded into the NIA Mouse Gene Index by selecting an option ‘‘View a list of selected genes/transcripts.’’ Then, this list can be tested for overrepresentation of GO terms by clicking the ‘‘GO-annotation’’ button. Statistical significance is determined using two criteria: FDR and overrepresentation ratio with adjustable thresholds. The result can be viewed interactively in the web browser or can be downloaded as a tab-delimited text (see ‘‘Results’’ hyperlink at the top of the web page showing the results of analysis). In addition, the NIA Mouse Gene Index can generate a similar table with overrepresented protein domains and identify clusters of neighboring genes in the genome. Overrepresentation of genes that belong to a specific signaling pathway have certain transcription factor binding sites (TFBS) in their promoters, or clustered in the genome can be analyzed by the GSEA software (Kim and Volsky, 2005; Subramanian et al., 2005). Recent chromatin-immunoprecipitation experiments have yielded genome-wide data on the location of TFBS as well as chromatin modification patterns in the mouse genome (Loh et al., 2006; Mikkelsen et al., 2007). Combined analysis of TFBS for a specific transcription factor (TF) and microarray data after overexpressing or repressing the same TF in cells and embryos can help to identify a set of genes that are direct downstream of the TF (see the methods in Nishiyama et al., 2009; Sharov et al., 2008).

7. Submitting the Data to the Public Database Expression profiling does not end with the analysis of the data. Most journals require that data on microarray experiments are submitted to public databases (e.g., GEO (Barrett et al., 2009; http://www.ncbi.nlm.nih.gov/ geo/) or ArrayExpress (Parkinson et al., 2009; http://www.ebi.ac.uk/microarray-as/ae/)) in the format compliant with the Minimum Information About a Microarray Experiment (MIAME) (Brazma et al., 2001). To be MIAME compliant, which dramatically increases the usability of the microarray data in the future and benefits the research community, the detailed information about all the experimental conditions and sample preparations are

Gene Expression Profiling

537

required. The best strategy is to collect the detailed information, while one is still doing experiments, with the data submission in mind. One important issue is gene nomenclature. Although most journals encourage authors to use the standard gene nomenclature, authors often use the gene names/gene symbols that they have coined or they like. However, we strongly recommend that authors follow the standard nomenclature that is set by the International Committee on Standardized Genetic Nomenclature for Mice (http://www.informatics.jax.org/nomen/). If it is a new gene with no name assigned yet, the authors are encouraged to contact the nomenclature committee to discuss the appropriate gene symbols.

ACKNOWLEDGMENTS The authors thank past and current members of the Ko lab for contributing the work described in this chapter. The authors’ laboratory had a CRADA arrangement with Agilent Technologies; however, the authors have no personal financial interest in Agilent Technologies. The work was entirely supported by the Intramural Research Program of the NIH, National Institute on Aging.

REFERENCES Abe, K., Ko, M. S., and MacGregor, G. R. (1998). A systematic molecular genetic approach to study mammalian germline development. Int. J. Dev. Biol. 42, 1051–1065. Aiba, K., Sharov, A. A., Carter, M. G., Foroni, C., Vescovi, A. L., and Ko, M. S. (2006). Defining a developmental path to neural fate by global expression profiling of mouse embryonic stem cells and adult neural stem/progenitor cells. Stem Cells 24, 889–895. Aiba, K., Nedorezov, T., Piao, Y., Nishiyama, A., Matoba, R., Sharova, L. V., Sharov, A. A., Yamanaka, S., Niwa, H., and Ko, M. S. (2009). Defining developmental potency and cell lineage trajectories by expression profiling of differentiating mouse embryonic stem cells. DNA Res. 16, 73–80. Ashburner, M., Ball, C. A., Blake, J. A., Botstein, D., Butler, H., Cherry, J. M., Davis, A. P., Dolinski, K., Dwight, S. S., Eppig, J. T., et al. (2000). Gene ontology: Tool for the unification of biology. The Gene Ontology Consortium. Nat. Genet. 25, 25–29. Babak, T., Blencowe, B. J., and Hughes, T. R. (2005). A systematic search for new mammalian noncoding RNAs indicates little conserved intergenic transcription. BMC Genomics 6, 104. Baldi, P., and Long, A. D. (2001). A Bayesian framework for the analysis of microarray expression data: Regularized t-test and statistical inferences of gene changes. Bioinformatics 17, 509–519. Barrett, T., Troup, D. B., Wilhite, S. E., Ledoux, P., Rudnev, D., Evangelista, C., Kim, I. F., Soboleva, A., Tomashevsky, M., Marshall, K. A., et al. (2009). NCBI GEO: Archive for high-throughput functional genomic data. Nucleic Acids Res. 37, D885–D890. Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate—a practical and powerful approach to multiple testing. J. R. Stat. Soc. B 57, 289–300. Bockamp, E., Sprengel, R., Eshkind, L., Lehmann, T., Braun, J. M., Emmrich, F., and Hengstler, J. G. (2008). Conditional transgenic mouse models: From the basics to genome-wide sets of knockouts and current studies of tissue regeneration. Regen. Med. 3, 217–235.

538

Alexei A. Sharov et al.

Brazma, A., Hingamp, P., Quackenbush, J., Sherlock, G., Spellman, P., Stoeckert, C., Aach, J., Ansorge, W., Ball, C. A., Causton, H. C., et al. (2001). Minimum information about a microarray experiment (MIAME)—Toward standards for microarray data. Nat. Genet. 29, 365–371. Byerly, S., Sundin, K., Raja, R., Stanchfield, J., Bejjani, B. A., and Shaffer, L. G. (2009). Effects of ozone exposure during microarray posthybridization washes and scanning. J. Mol. Diagn. 11, 590–597. Canales, R. D., Luo, Y., Willey, J. C., Austermiller, B., Barbacioru, C. C., Boysen, C., Hunkapiller, K., Jensen, R. V., Knight, C. R., Lee, K. Y., et al. (2006). Evaluation of DNA microarray results with quantitative gene expression platforms. Nat. Biotechnol. 24, 1115–1122. Carter, M. G., Hamatani, T., Sharov, A. A., Carmack, C. E., Qian, Y., Aiba, K., Ko, N. T., Dudekula, D. B., Brzoska, P. M., Hwang, S. S., et al. (2003). In situ-synthesized novel microarray optimized for mouse stem cell and early developmental expression profiling. Genome Res. 13, 1011–1021. Carter, M. G., Sharov, A. A., VanBuren, V., Dudekula, D. B., Carmack, C. E., Nelson, C., and Ko, M. S. (2005). Transcript copy number estimation using a mouse whole-genome oligonucleotide microarray. Genome Biol. 6, R61. Carter, M. G., Stagg, C. A., Falco, G., Yoshikawa, T., Bassey, U. C., Aiba, K., Sharova, L. V., Shaik, N., and Ko, M. S. (2008). An in situ hybridization-based screen for heterogeneously expressed genes in mouse ES cells. Gene Expr. Patterns 8, 181–198. Chapman, S., Schenk, P., Kazan, K., and Manners, J. (2002). Using biplots to interpret gene expression patterns in plants. Bioinformatics 18, 202–204. Cheadle, C., Becker, K. G., Cho-Chung, Y. S., Nesterova, M., Watkins, T., Wood, III, W., Prabhu, V., and Barnes, K. C. (2007). A rapid method for microarray cross platform comparisons using gene expression signatures. Mol. Cell. Probes 21, 35–46. Chen, Y., Haviernik, P., Bunting, K. D., and Yang, Y. C. (2007). Cited2 is required for normal hematopoiesis in the murine fetal liver. Blood 110, 2889–2898. Cloonan, N., Forrest, A. R., Kolle, G., Gardiner, B. B., Faulkner, G. J., Brown, M. K., Taylor, D. F., Steptoe, A. L., Wani, S., Bethel, G., et al. (2008). Stem cell transcriptome profiling via massive-scale mRNA sequencing. Nat. Methods 5, 613–619. Cui, X. S., Li, X. Y., Shen, X. H., Bae, Y. J., Kang, J. J., and Kim, N. H. (2007). Transcription profile in mouse four-cell, morula, and blastocyst: Genes implicated in compaction and blastocoel formation. Mol. Reprod. Dev. 74, 133–143. Dahlquist, K. D., Salomonis, N., Vranizan, K., Lawlor, S. C., and Conklin, B. R. (2002). GenMAPP, a new tool for viewing and analyzing microarray data on biological pathways. Nat. Genet. 31, 19–20. Fare, T. L., Coffey, E. M., Dai, H., He, Y. D., Kessler, D. A., Kilian, K. A., Koch, J. E., LeProust, E., Marton, M. J., Meyer, M. R., et al. (2003). Effects of atmospheric ozone on microarray data quality. Anal. Chem. 75, 4672–4675. Frankenberg, S., Smith, L., Greenfield, A., and Zernicka-Goetz, M. (2007). Novel gene expression patterns along the proximo-distal axis of the mouse embryo before gastrulation. BMC Dev. Biol. 7, 8. Gentleman, R. C., Carey, V. J., Bates, D. M., Bolstad, B., Dettling, M., Dudoit, S., Ellis, B., Gautier, L., Ge, Y., Gentry, J., et al. (2004). Bioconductor: Open software development for computational biology and bioinformatics. Genome Biol. 5, R80. Hamatani, T., Carter, M. G., Sharov, A. A., and Ko, M. S. (2004a). Dynamics of global gene expression changes during mouse preimplantation development. Dev. Cell 6, 117–131. Hamatani, T., Falco, G., Carter, M. G., Akutsu, H., Stagg, C. A., Sharov, A. A., Dudekula, D. B., VanBuren, V., and Ko, M. S. (2004b). Age-associated alteration of gene expression patterns in mouse oocytes. Hum. Mol. Genet. 13, 2263–2278.

Gene Expression Profiling

539

Hipp, J., and Atala, A. (2006). GeneChips in stem cell research. Methods Enzymol. 420, 162–224. Hopkins, C., Sadelova, S., Fournier, T., Ilsley, D., and Park, M. (2003). In ‘‘Application of Agilent’s 60-mer Oligonucleotide Microarrays for Gene Expression Analysis of Small RNA Quantities Derived from Laser Capture Microdissection’’. Agilent Technologies, Palo Alto, CA. Hughes, T. R., Mao, M., Jones, A. R., Burchard, J., Marton, M. J., Shannon, K. W., Lefkowitz, S. M., Ziman, M., Schelter, J. M., Meyer, M. R., et al. (2001). Expression profiling using microarrays fabricated by an ink-jet oligonucleotide synthesizer. Nat. Biotechnol. 19, 342–347. Jung, H. J., Shim, J. S., Suh, Y. G., Kim, Y. M., Ono, M., and Kwon, H. J. (2007). Potent inhibition of in vivo angiogenesis and tumor growth by a novel cyclooxygenase-2 inhibitor, enoic acanthoic acid. Cancer Sci. 98, 1943–1948. Kawasaki, E. S. (2006). The end of the microarray Tower of Babel: Will universal standards lead the way? J. Biomol. Tech. 17, 200–206. Kim, S. Y., and Volsky, D. J. (2005). PAGE: Parametric analysis of gene set enrichment. BMC Bioinformatics 6, 144. Kimelman, D. (2006). Mesoderm induction: From caps to chips. Nat. Rev. Genet. 7, 360–372. Kitaya, K., Yasuo, T., Yamaguchi, T., Fushiki, S., and Honjo, H. (2007). Genes regulated by interferon-gamma in human uterine microvascular endothelial cells. Int. J. Mol. Med. 20, 689–697. Ko, M. S. (1990). An ‘equalized cDNA library’ by the reassociation of short double-stranded cDNAs. Nucleic Acids Res. 18, 5705–5711. Ko, M. S. (2001). Embryogenomics: Developmental biology meets genomics. Trends Biotechnol. 19, 511–518. Ko, M. S. (2006). Expression profiling of the mouse early embryo: Reflections and perspectives. Dev. Dyn. 235, 2437–2448. Kurimoto, K., Yabuta, Y., Ohinata, Y., Ono, Y., Uno, K. D., Yamada, R. G., Ueda, H. R., and Saitou, M. (2006). An improved single-cell cDNA amplification method for efficient high-density oligonucleotide microarray analysis. Nucleic Acids Res. 34, e42. LaGamba, D., Nawshad, A., and Hay, E. D. (2005). Microarray analysis of gene expression during epithelial–mesenchymal transformation. Dev. Dyn. 234, 132–142. Landry, J., Sharov, A. A., Piao, Y., Sharova, L. V., Xiao, H., Southon, E., Matta, J., Tessarollo, L., Zhang, Y. E., Ko, M. S., et al. (2008). Essential role of chromatin remodeling protein Bptf in early mouse embryos and embryonic stem cells. PLoS Genet. 4, e1000241. Lewandoski, M. (2001). Conditional control of gene expression in the mouse. Nat. Rev. Genet. 2, 743–755. Loh, Y. H., Wu, Q., Chew, J. L., Vega, V. B., Zhang, W., Chen, X., Bourque, G., George, J., Leong, B., Liu, J., et al. (2006). The Oct4 and Nanog transcription network regulates pluripotency in mouse embryonic stem cells. Nat. Genet. 38, 431–440. Mardis, E. R. (2008). Next-generation DNA sequencing methods. Annu. Rev. Genomics Hum. Genet. 9, 387–402. Marioni, J. C., Mason, C. E., Mane, S. M., Stephens, M., and Gilad, Y. (2008). RNA-seq: An assessment of technical reproducibility and comparison with gene expression arrays. Genome Res. 18, 1509–1517. Masui, S., Nakatake, Y., Toyooka, Y., Shimosato, D., Yagi, R., Takahashi, K., Okochi, H., Okuda, A., Matoba, R., Sharov, A. A., et al. (2007). Pluripotency governed by Sox2 via regulation of Oct3/4 expression in mouse embryonic stem cells. Nat. Cell Biol. 9, 625–635.

540

Alexei A. Sharov et al.

Matsumoto, N., Kubo, A., Liu, H., Akita, K., Laub, F., Ramirez, F., Keller, G., and Friedman, S. L. (2006). Developmental regulation of yolk sac hematopoiesis by Kruppel-like factor 6. Blood 107, 1357–1365. McCarrey, J. R., Hsu, K. C., Eddy, E. M., Klevecz, R. R., and Bolen, J. L. (1987). Isolation of viable mouse primordial germ cells by antibody-directed flow sorting. J. Exp. Zool. 242, 107–111. Mikkelsen, T. S., Ku, M., Jaffe, D. B., Issac, B., Lieberman, E., Giannoukos, G., Alvarez, P., Brockman, W., Kim, T. K., Koche, R. P., et al. (2007). Genome-wide maps of chromatin state in pluripotent and lineage-committed cells. Nature 448, 553–560. Mortazavi, A., Williams, B. A., McCue, K., Schaeffer, L., and Wold, B. (2008). Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat. Methods 5, 621–628. Nagy, A., Gertsenstein, M., Vintersten, K., and Behringer, R. R. (2002). Manipulating the Mouse Embryo: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Nishiyama, A., Xin, L., Sharov, A. A., Thomas, M., Mowrer, G., Meyers, E., Piao, Y., Mehta, S., Yee, S., Nakatake, Y., et al. (2009). Uncovering early response of gene regulatory networks in ESCs by systematic induction of transcription factors. Cell Stem Cell 5, 420–433. Niwa, H., Miyazaki, J., and Smith, A. G. (2000). Quantitative expression of Oct-3/4 defines differentiation, dedifferentiation or self-renewal of ES cells. Nat. Genet. 24, 372–376. Olena, A. F., and Patton, J. G. (2010). Genomic organization of microRNAs. J. Cell. Physiol. 222, 540–545. Oshlack, A., and Wakefield, M. J. (2009). Transcript length bias in RNA-seq data confounds systems biology. Biol. Direct 4, 14. Pan, W., Lin, J., and Le, C. T. (2002). How many replicates of arrays are required to detect gene expression changes in microarray experiments? A mixture model approach. Genome Biol. 3, research0022. Parkinson, H., Kapushesky, M., Kolesnikov, N., Rustici, G., Shojatalab, M., Abeygunawardena, N., Berube, H., Dylag, M., Emam, I., Farne, A., et al. (2009). ArrayExpress update—From an archive of functional genomics experiments to the atlas of gene expression. Nucleic Acids Res. 37, D868–D872. Sharov, A. A., Dudekula, D. B., and Ko, M. S. (2005a). Genome-wide assembly and analysis of alternative transcripts in mouse. Genome Res. 15, 748–754. Sharov, A. A., Dudekula, D. B., and Ko, M. S. (2005b). A web-based tool for principal component and significance analysis of microarray data. Bioinformatics 21, 2548–2549. Sharov, A. A., Masui, S., Sharova, L. V., Piao, Y., Aiba, K., Matoba, R., Xin, L., Niwa, H., and Ko, M. S. (2008). Identification of Pou5f1, Sox2, and Nanog downstream target genes with statistical confidence by applying a novel algorithm to time course microarray and genome-wide chromatin immunoprecipitation data. BMC Genomics 9, 269. Sharova, L. V., Sharov, A. A., Piao, Y., Shaik, N., Sullivan, T., Stewart, C. L., Hogan, B. L., and Ko, M. S. (2007). Global gene expression profiling reveals similarities and differences among mouse pluripotent stem cells of different origins and strains. Dev. Biol. 307, 446–459. Sharova, L. V., Sharov, A. A., Nedorezov, T., Piao, Y., Shaik, N., and Ko, M. S. (2009). Database for mRNA half-life of 19 977 genes obtained by DNA microarray analysis of pluripotent and differentiating mouse embryonic stem cells. DNA Res. 16, 45–58. Sherwood, R. I., Jitianu, C., Cleaver, O., Shaywitz, D. A., Lamenzo, J. O., Chen, A. E., Golub, T. R., and Melton, D. A. (2007). Prospective isolation and global gene expression analysis of definitive and visceral endoderm. Dev. Biol. 304, 541–555.

Gene Expression Profiling

541

Shi, L., Reid, L. H., Jones, W. D., Shippy, R., Warrington, J. A., Baker, S. C., Collins, P. J., de Longueville, F., Kawasaki, E. S., Lee, K. Y., et al. (2006). The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene expression measurements. Nat. Biotechnol. 24, 1151–1161. Su, A. I., Cooke, M. P., Ching, K. A., Hakak, Y., Walker, J. R., Wiltshire, T., Orth, A. P., Vega, R. G., Sapinoso, L. M., Moqrich, A., et al. (2002). Large-scale analysis of the human and mouse transcriptomes. Proc. Natl. Acad. Sci. USA 99, 4465–4470. Subramanian, A., Tamayo, P., Mootha, V. K., Mukherjee, S., Ebert, B. L., Gillette, M. A., Paulovich, A., Pomeroy, S. L., Golub, T. R., Lander, E. S., et al. (2005). Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. USA 102, 15545–15550. Sultan, M., Schulz, M. H., Richard, H., Magen, A., Klingenhoff, A., Scherf, M., Seifert, M., Borodina, T., Soldatov, A., Parkhomchuk, D., et al. (2008). A global view of gene activity and alternative splicing by deep sequencing of the human transcriptome. Science 321, 956–960. Tanaka, T. S., Jaradat, S. A., Lim, M. K., Kargul, G. J., Wang, X., Grahovac, M. J., Pantano, S., Sano, Y., Piao, Y., Nagaraja, R., et al. (2000). Genome-wide expression profiling of mid-gestation placenta and embryo using a 15,000 mouse developmental cDNA microarray. Proc. Natl. Acad. Sci. USA 97, 9127–9132. Thomson, J. M., Parker, J. S., and Hammond, S. M. (2007). Microarray analysis of miRNA gene expression. Methods Enzymol. 427, 107–122. Tusher, V. G., Tibshirani, R., and Chu, G. (2001). Significance analysis of microarrays applied to the ionizing radiation response. Proc. Natl. Acad. Sci. USA 98, 5116–5121. van Loo, P. F., Mahtab, E. A., Wisse, L. J., Hou, J., Grosveld, F., Suske, G., Philipsen, S., and Gittenberger-de Groot, A. C. (2007). Transcription factor Sp3 knockout mice display serious cardiac malformations. Mol. Cell. Biol. 27, 8571–8582. Wang, Q. T., Piotrowska, K., Ciemerych, M. A., Milenkovic, L., Scott, M. P., Davis, R. W., and Zernicka-Goetz, M. (2004). A genome-wide study of gene activity reveals developmental signaling pathways in the preimplantation mouse embryo. Dev. Cell 6, 133–144. Wang, L., Mear, J. P., Kuan, C. Y., and Colbert, M. C. (2005). Retinoic acid induces CDK inhibitors and growth arrest specific (Gas) genes in neural crest cells. Dev. Growth Differ. 47, 119–130. Wang, Z., Gerstein, M., and Snyder, M. (2009). RNA-Seq: A revolutionary tool for transcriptomics. Nat. Rev. Genet. 10, 57–63. Williams, S. S., Mear, J. P., Liang, H. C., Potter, S. S., Aronow, B. J., and Colbert, M. C. (2004). Large-scale reprogramming of cranial neural crest gene expression by retinoic acid exposure. Physiol. Genomics 19, 184–197. Wilusz, J. E., Sunwoo, H., and Spector, D. L. (2009). Long noncoding RNAs: Functional surprises from the RNA world. Genes Dev. 23, 1494–1504. Yoshikawa, T., Piao, Y., Zhong, J., Matoba, R., Carter, M. G., Wang, Y., Goldberg, I., and Ko, M. S. (2006). High-throughput screen for genes predominantly expressed in the ICM of mouse blastocysts by whole mount in situ hybridization. Gene Expr. Patterns 6, 213–224. Zeeberg, B. R., Feng, W., Wang, G., Wang, M. D., Fojo, A. T., Sunshine, M., Narasimhan, S., Kane, D. W., Reinhold, W. C., Lababidi, S., et al. (2003). GoMiner: A resource for biological interpretation of genomic and proteomic data. Genome Biol. 4, R28. Zeng, F., Baldwin, D. A., and Schultz, R. M. (2004). Transcript profiling during preimplantation mouse development. Dev. Biol. 272, 483–496. Zhu, H., Cabrera, R. M., Wlodarczyk, B. J., Bozinov, D., Wang, D., Schwartz, R. J., and Finnell, R. H. (2007). Differentially expressed genes in embryonic cardiac tissues of mice lacking Folr1 gene activity. BMC Dev. Biol. 7, 128.

Author Index

A Aach, J., 476–477, 536 Abbas, A. I., 198 Abbott, A., 544 Abe, K., 304, 310, 514 Abeygunawardena, N., 515, 536 Abremski, K., 111, 168 Abreu-Goodger, C., 343 Abreu, S. L., 484 Abuin, A., 223, 247, 250 Acevedo-Arozena, A., 331 Achacoso, P., 255 Ackerman, S. L., 486 Adamantidis, A., 197–198 Adams, D. J., 233, 235, 260 Adams, H. A., 478 Adams, N. C., 127, 417 Adamson, S. L., 416 Aebischer, P., 446 Affar el, B., 352 Affourtit, J., 488, 490–491, 493, 495–497, 499 Agami, R., 353, 416 Aguet, M., 116 Ahringer, J., 369 Aiba, K., 512, 514, 518, 521, 535–536 Aifantis, I., 416 Aigner, B., 334 Airan, R., 197–198 Aizawa, S., 130, 249 Akagi, K., 102, 165, 186, 254 Akil, H., 496 Akita, K., 518 Akizuki, M., 251 Akutsu, H., 459, 521 Alavizadeh, A., 223 Albert, H., 111–112 Albright, T. D., 198 Alcorn, H. L., 303, 308, 333 Aldea, M., 127 Al Emran, O., 436, 445 Alenina, N., 444 Alexander, G. M., 198 Alexander, W. S., 304, 310, 336 Allak, B. A., 118 Allen, J. A., 198, 262 Aller, M. I., 434 Altschmied, J., 112–113, 246 Alvarez, P., 536 Alvarez-Saavedra, M., 440

Amaya, E., 272 Anastassiadis, K., 109–110, 112–119, 125–127, 129–130, 190, 196, 250, 430 Anderson, D. J., 186, 198, 205 Anderson, K. V., 303, 308, 310, 331, 333 Andreas, S., 116 Andrews, B. J., 147, 156 Anger, M., 487 Angrand, P.-O., 127, 129, 147, 190, 250 Angr, P.-O., 110, 114–115 Ang, Y. S., 353, 355 Ansorge, W., 476–477, 536 Antonchuk, J. L., 304, 310 Antunes, C., 116 An, W., 55 Anzai, M., 198 Aoki, A., 38 Araki, K., 249, 251 Araki, M., 247, 251 Aravanis, A. M., 197–198 Aravin, A. A., 487 Arber, S., 191 Are´chaga, J. M., 31, 33 Arenkiel, B. R., 198 Arkell, R., 304, 310 Armbruster, B. N., 198 Arnold, H. H., 247 Aronovich, E. L., 64 Aronow, B. J., 518 Artzt, K., 306 Asada, H., 459 Ashburner, M., 536 Ashe, A., 304, 310 Ashery-Padan, R., 113, 190 Ashique, A. M., 333 Atala, A., 513 Athey, B., 496 Auerbach, B. A., 248, 272 Auerbach, W., 127, 196, 417 Austermiller, B., 519 Austin, C. P., 185 Austin, S., 156 Avarbock, M. R., 18–19, 25–27, 29–31, 33–34 Avery, B. J., 248 Awatramani, R. B., 114, 159–160, 164, 184–186, 189–192, 195–196 Axton, R., 250 Ayala, R., 113, 190, 196 Ayres, E. K., 147

543

544

Author Index B

Babak, T., 513 Baba, S., 18–19 Baba, T., 26 Babineau, D., 147 Babitzke, P., 127 Bach, M. E., 436, 439–441 Bachvarova, R. F., 483–485, 488 Badea, T. C., 116 Bader, M., 444 Baer, A., 113, 251, 372 Bae, Y. J., 516 Bahlo, M., 339 Bai, C. B., 171, 196 Bajorek, E., 458 Baker, P. J., 504–505 Baker, S. C., 519 Balciunas, D., 58 Balderes, D., 147 Baldi, P., 532 Baldwin, D. A., 460, 463, 478, 516 Ball, C. A., 476–477, 536 Ballou, L. U., 5 Balordi, F., 174 Baltimore, D., 4, 9 Banchaabouchi, M. A., 335 Bangham, J. W., 299, 308, 332 Bank, A., 255 Banrezes, B., 460 Banta, H., 4 Barbacid, M., 102 Barbacioru, C. C., 476–477, 495–497, 500–501, 506, 519 Barbara, M., 486, 488 Barchas, J. D., 459, 486, 488 Barco, A., 430 Bardeesy, N., 417 Barlow, D. P., 19 Barnes, K. C., 519 Barnett, S., 115, 118 Baron, U., 196, 433–434, 446 Barrett, T., 515, 536 Barsh, G. S., 303 Bartolomei, M. S., 461, 478 Barton, M. D., 444 Bartsch, D., 445–446 Bass, B. L., 369 Bassey, U. C., 495, 514 Bast, T., 434 Bates, D. M., 532 Battiste, J., 189, 191 Batzer, M. A., 93 Baugh, L. R., 486 Baumgartel, K., 433 Bausen, M., 434 Baus, J., 61, 94 Bautch, V. L., 287 Baxter, L. L., 304, 310, 336

Baylink, D. J., 304, 310 Bazopoulou, D., 54 Beachy, P. A., 331 Beam, T., 258 Beard, C., 191, 443, 446–447 Beatty, L. G., 147 Becher, B., 186 Becker, K. G., 519 Becker, L., 331 Beckstead, W. A., 342 Beddington, R. S., 248, 253, 486 Behringer, R. R., 9–11, 145, 185–186, 303, 310, 503, 521 Beier, D. R., 115, 303, 308, 316–317, 329, 331, 333, 342 Beijersbergen, R. L., 358 Bejar, R., 434 Bejjani, B. A., 529 Belin, D., 485 Bellen, H. J., 84 Belloc, E., 485 Belloni, E., 331 Belteki, G., 374 BeltrandelRio, H., 250 Belur, L. R., 64 Bender, G., 432 Benes, V., 113, 127 Benjamini, Y., 531 Benjamin, L. E., 486 Benvenisty, N., 288 Berens, C., 431 Berger, S., 445–446 Bergstrom, R. A., 221, 223 Berhardt, K., 109, 430 Berkhout, B., 433 Bernards, R., 353 Berninger, P., 461 Berns, A., 92, 94, 116, 118, 186, 233, 260, 262 Bernstein, A., 251–252 Bernstein, E., 352–353, 355, 369, 416 Bernstein, J. G., 198 Berta, P., 331 Berube, H., 515, 536 Besschetnova, T. Y., 333 Bethel, G., 514–515 Betz, U. A., 114 Beutler, B., 331, 335 Beuzard, Y., 331 Beverloo, H. B., 118 Beverly, L. J., 441–442 Bex, A., 116 Bialek, P., 341 Bickmore, W. A., 248 Biggs, P., 127 Bignell, G. R., 92 Bindels, E., 116 Bird, A. W., 127 Birmingham, A., 358, 370

545

Author Index

Bishop, J. M., 352, 436 Bixler, L. S., 19 Bjork, B. C., 341–342 Black, S., 458 Blaess, S., 167 Blair, K., 444 Blake, J. A., 247, 486–488, 491, 495, 536 Blash, S., 27 Blencowe, B. J., 513 Blewitt, M. E., 304, 310 Blomer, U., 4, 6 Bockamp, E., 116, 518 Bodeau, J., 476–477, 495–497, 500–501, 506 Bode, J., 111, 113, 251, 372–373 Bode, V. C., 331–332 Bodo, S., 491 Boeke, J. D., 126, 224 Bolen, J. L., 514 Boles, M. K., 308, 310, 324, 344 Bolle, I., 331 Bolstad, B. M., 493, 532 Bolton, A. D., 341 Bonaldo, P., 251 Bonin, A. L., 430–431, 446 Bonnerot, C., 169 Boocock, M. R., 186 Boreen, S. M., 484 Bornstein, P., 19 Borodina, T., 514 Borowski, C., 416 Bos, J. L., 102 Botstein, D., 536 Boucher, M., 127 Boucher, R. C., 218 Bourque, G., 536 Boutros, M., 369 Bowers, J., 478 Box, N., 303, 310 Boyd, A. D., 496 Boyden, E. S., 198 Boysen, C., 519 Bozinov, D., 517 Bradley, A., 91, 94, 96, 100, 113–115, 126, 217, 219–220, 223–224, 227–228, 230, 233, 238, 252, 254–255, 258, 262, 273, 288, 300, 303, 310 Brady, G., 486 Brambrink, T., 443 Branda, C. S., 110, 114, 116, 126, 184–186 Brantjes, H., 416 Braun, J. M., 518 Braun, T., 113 Brazma, A., 476–477, 536 Bremel, R. D., 5 Brennan, J., 248 Brickman, J. M., 250, 253, 271, 288 Brielmeier, M., 331 Brigman, K. K., 218

Brinster, C. J., 19 Brinster, R. L., 18–19, 25–27, 29–31, 33–34 Broadie, K., 197 Brocard, J., 115 Brockman, J. M., 495–497, 506 Brockman, W., 536 Brodsky, F. M., 352 Broll, I., 186 Bronchain, O. J., 272 Bronson, R. T., 198, 303, 308, 331, 417 Brooker, D., 300, 308, 331, 333 Brookes, A. J., 233, 259, 290 Brower, P. T., 484 Brown, A., 245, 341 Brown, E. J., 4, 9, 417 Brown, E. L., 486 Browning, V. L., 221 Brown, M. K., 514–515 Brown, S. D., 307, 330–331, 335 Brummelkamp, T. R., 353, 371, 388, 416 Bruneau, B. G., 416 Bruning, J. C., 416–417, 419, 446 Brunk, B. P., 478 Bruxner, T. J., 304, 310 Brzoska, P. M., 512, 521 Buac, K., 304, 310 Bubunenko, M., 127 Bucan, M., 223 Buchholz, F., 110, 112–115, 117, 126–127, 130, 136, 147, 157, 190, 196, 250, 352 Buch, T., 186 Buehr, M., 444 Buhl, D. L., 197 Bujard, H., 116, 376–377, 416, 429–435, 437–438, 445–446 Bullock, W. O., 300 Bult, C. J., 247, 486–488, 491, 495 Bultman, S. J., 461 Bunney, W. E., 496 Bunting, K. D., 517 Burchard, J., 416, 519 Burgess, D. J., 445 Burnett, M. B., 223, 247 Burns, J. C., 5 Busch, D. H., 304, 310 Bushell, W., 127 Bushman, F. D., 11 Butler, H., 536 Butler, S. L., 11 Butt, S. J., 189, 191 Buxton, E. C., 249–250 Byerly, S., 529 Bygrave, A., 19 C Cabrera, R. M., 517 Cadinanos, J., 94, 96, 100, 238, 258

546 Califano, A., 474 Callaway, E. M., 197–198 Calle´ja, C., 112–113 Calzada-Wack, J., 331 Cambridge, S., 434 Campbell, D., 303 Campbell, R., 245 Camp, J. R., 189, 192 Canales, R. D., 519 Canham, M., 286 Caoile, C., 458 Cao, S., 252 Capecchi, M. R., 115, 118, 127, 130, 218 Carey, B. W., 443, 447 Carey, M. R., 184, 191–192, 196–198 Carey, V. J., 447, 532 Carlson, C. M., 61, 92, 226, 254 Carmack, C. E., 512, 521 Carmell, M. A., 372 Carpenter, A. E., 359 Carpenter, D. A., 299, 332, 337 Carpinelli, M. R., 304, 310 Carroll, P., 287 Carter, M. G., 478, 487, 495, 512, 514, 516, 518, 521 Cary, L. C., 59, 224, 254 Casallo, G., 247 Casanova, E., 114, 116 Caspary, T., 303, 308, 313, 333 Casper, K. B., 172 Cassady, J. P., 443 Caudy, A. A., 352, 416 Causton, H. C., 476–477, 536 Cavalcoli, J., 221 Cawthorne, C., 419 Celikel, T., 440 Cepko, C. L., 185 Certa, U., 486 Cetin, A., 434 Chabanis, S., 110, 129 Chambers, I., 113, 127, 219, 250, 279, 286 Chambon, P., 112–113, 115–116, 157, 246 Chan, A. W., 5 Chang, A., 335 Chang, K., 445 Chang, L. J., 30 Chang, P. J., 247 Chang, W. C., 197–198 Chan, Y. M., 458 Chao, S., 220 Chapman, S., 534 Chapman, V., 337 Chatterjee, A., 221 Chavala, S., 19 Chavez, S., 19 Cheadle, C., 519 Cheloufi, S., 487 Chen, A. E., 516

Author Index

Chen, C. Y., 38, 289, 352 Chen, D. F., 205 Chen, F. H., 27, 486–488, 491, 495 Cheng, A. W., 443 Cheng, L., 19 Cheng, P. H., 4 Cheng, Y., 461 Chen, J., 353, 437 Chen, Q., 306 Chen, W. V., 129, 247, 250, 252–253, 255, 259, 263, 276, 283, 286 Chen, X., 536 Chen, Y. T., 110, 221, 252, 316, 517 Chen, Z., 253 Chernomorski, R., 127 Cherry, J. M., 536 Chesebro, A. L., 333 Chew, J. L., 536 Chiang, C., 331 Chiba, H., 486–487 Chiba, J., 249 Chikama, T., 437 Ching, K. A., 515 Chin, L., 438 Cho-Chung, Y. S., 519 Choi, J. K., 478 Choi, V. M., 444 Chowdhury, K., 251 Christoph, T., 372 Chuang, P. T., 247 Chudakov, D. M., 159 Chu, G., 532 Chuma, S., 30, 39 Chung, K. H., 371 Chu, T., 444 Ciemerych, M. A., 478, 487–491, 493, 516 Claeysen, S., 197–198 Clapoff, S., 245 Clark, A. T., 303, 310 Clark, K. J., 67 Claudiani, P., 246 Cleary, M. A., 416 Cleaver, O., 516 Clegg, K. B., 483–485 Clements, S., 330 Clevers, H., 416 Cloonan, N., 500, 514–515 Clouthier, D. E., 18, 30 Coates, C. J., 224 Cobellis, G., 246–247, 273 Coffey, A. J., 233 Coffey, E. M., 529 Coghill, E. L., 316, 330, 335 Cohn, J. B., 273 Colbert, M. C., 518 Colby, D., 219, 250 Cole, T. B., 419 Collier, L. S., 66, 73, 92–96, 98, 102–103, 226, 254

547

Author Index

Collins, F. S., 247, 284, 325 Collins, P. J., 519 Comoglio, P. M., 252 Conklin, B. R., 197–198, 247, 249, 536 Conklin, D. S., 352–353, 416 Conner, D. A., 220 Conrad, C., 358 Cooke, M. P., 515 Cook, M. C., 331, 335 Cook, M. N., 184, 191–192, 196–198 Cooley, L., 54, 93 Cooper, G. M., 485 Cope, L., 493 Copeland, N. G., 115, 126–127, 186, 244, 254 Corbeau, P., 4 Cordaux, R., 93 Corden, J. L., 331 Cordes, S. P., 273, 303, 333, 337 Cordon-Cardo, C., 445 Cormier, C., 353, 355 Cornett, C. V., 299, 332 Corrales, J. D., 167 Corrin, P. D., 247, 250, 252–253, 255 Corsaro, B. G., 224, 254 Costantini, F., 186 Costantino, N., 127 Cost, G. J., 444 Coude, M., 331 Coumoul, X., 374, 416 Court, D. L., 115, 126–127, 186 Covarrubias, L., 244 Cowan, A., 417 Cowling, R., 443 Cox, A. V., 247 Cox, M. M., 111 Cox, R. D., 330–331 Cox, T., 127, 227–228, 247, 249, 254–255, 258 Crabbe, J. C., 335 Creemers, L. B., 30 Croker, B., 4–5 Cross, S. H., 343 Cross, S. K., 504 Cruz, P., 246–247 Cui, X. S., 444, 516 Cui, Y., 486 Cui, Z., 61, 80 Cullen, B. R., 371, 375 Cullen, D., 417 Cummins, A., 185 Cuzin, F. Czauderna, F., 376 D Dahlquist, K. D., 536 Dai, H., 529 Dai, M., 496 Dains, K., 221 Dale, E. C., 111–112

Dalke, C., 331 Dalma-Weiszhausz, D. D., 458–459, 463, 474 Dalmay, T., 343 Dani, C., 279 Danielian, P. S., 116 Daniels, G. R., 94 Dann, C. T., 371 Das, A. T., 433 Da Silva, V., 331 Datsenko, K. A., 127 Datta, S., 127 Daub, C. O., 500 Dausman, J. A., 443 Davey, R. E., 253 Davidson, B. L., 353 Davidson, N., 198 Davis, A. P., 305, 536 Davis, M., 333 Davison, C., 330 Davison, I. G., 198 Davis, R. W., 147, 478, 516 Davis, S., 486 Dayn, Y., 4, 9 Dean, D., 38 De Angelis, M. H., 334, 343 Dear, N., 303, 333 DeGennaro, L., 444 de Graaf, C. A., 304, 310, 336 DeGregori, J. V., 246, 249 Deininger, P. L., 94 Deisseroth, K., 197–198 De Las Rivas, J., 418 de Lecea, L., 197–198 De Leon, V., 484–485 Delic, S., 389 DeLoia, J. A., 486 de Longueville, F., 519 Delrow, J., 247, 250, 252–253, 255 DeMarini, D. M., 92 Dement, W. C., 459, 486, 488 Deneris, E. S., 184, 186, 189–192, 195 Deng, C. X., 416 Deng, K. Y., 246, 250–251 Deng, Y., 506 Dennis, G., Jr., 475 de Nooij, J. C., 191 den Ouden, K., 30 Denys, M., 458 De Palma, M., 4 Depaulis, A., 434 de Ridder, J., 101 de Rooij, D. G., 18, 20, 26, 30 DeRuiter, S. L., 353 Desai, R., 443 De Sauvage, F. J., 333 Desimone, R., 198 Desrosiers, R. C., 4 Destefano, D., 19

548 Destrempes, M. M., 27 Dettling, M., 532 Deuschle, U., 446 Deutsch, U., 438 Devon, R. S., 80, 233, 259, 290 de Vries, I., 354 De Vries, W. N., 481, 486–488, 491, 495, 501, 504 Devroe, E., 371 de Wit, T., 59, 62, 103 Deybach, J. C., 331 De-Zolt, S., 112–113, 246, 248–249, 259–260 Dhanasekaran, N., 488 Dhara, S. K., 288 Diamond, I., 436 Dickins, R. A., 376–377, 445 Dickson, M., 458 Dilioglou, S., 129, 253 Dill, K., 248 Ding, L., 354 Ding, S., 59, 62, 92, 94, 224, 255, 258 Dinnyes, A., 491 Dionne, K. M., 223, 247 Dionne, N., 220 DiPietrantonio, H. J., 190 Di Rago, L., 304, 310 Disteche, C. M., 19 Di Tizio, T., 127 Dive, C., 419 Dobrinski, I., 34 Dobson, A. T., 461 Doerflinger, N., 112–113 Doetschman, T. C., 287 Doi, H., 487 Dolinski, K., 536 Dolka, I., 248 Dominguez, M. G., 417 Donahue, S. L., 252 Donello, J. E., 5 Donnay, I., 488, 490 Donovan, P. J., 19 Dore, A. T., 417 Doring, A., 438 Doss, M. X., 354 Doughty, M., 185 Dove, W. F., 331, 335 Dowel, A., 279 Dressel, R., 19 Driver, S. E., 352 Duboule, D., 115 Dudai, Y., 439 Dudek, B. C., 335 Dudekula, D. B., 495, 512, 520–521, 532, 536 Dudoit, S., 532 Duerschke, K., 113–114, 117, 126, 130, 190, 196, 250 Duffy, T., 335 Dufva, I. H., 488

Author Index

Dufva, M., 488 Dull, T., 4–6 Dumont, D. J., 438 Dumuis, A., 197–198 Dunbar, M. E., 437 Duncan, F. E., 457, 461 Dunham, I., 233 Dunican, D. J., 358 Duniec, K., 444 Dunlop, J. W., 444 Dupuy, A. J., 53, 60, 62, 64, 66, 73, 92–96, 98, 102, 226, 254 Du, X., 331, 335 Du, Y. C., 441–442 Dwight, S. S., 536 Dwyer, N. D., 341 Dykxhoorn, D. M., 388 Dylag, M., 515, 536 Dymecki, S. M., 110, 113–114, 116, 126, 157, 159–160, 164, 183–186, 189–192, 194–198, 204 Dym, M.., 19 E Ebbert, A., 248 Ebersole, T. A., 306 Ebert, B. L., 536 Eberwine, J. H., 459, 486, 488 Eckermann, K., 440 Economides, A. N., 127 Eddy, E. M., 514 Edenhofer, F., 113, 117, 126, 130, 250 Edwards, D., 303 Edwards, J., 250 Efrat, S., 436, 445 Eggan, K., 85, 443 Eggenschwiler, J. T., 303, 308, 333 Ehlers, M. D., 198 Ehling, U. H., 299, 331 Ehrhardt, N., 331 Ehrlich, L., 110, 112 Eisenhaure, T. M., 359 Ekker, M., 333 Elbashir, S. M., 352, 369, 388, 416 Ellard, F. M., 4 Elledge, S. J., 147, 354, 445 Elliott, E. C., 304, 310, 336 Ellis, B., 532 Ellis, H. M., 127 Elvert, R., 331 Emam, I., 515, 536 Emelyanov, A., 67 Emmrich, F., 518 Endoh-Yamagami, S., 333 Engelke, D. R., 353 Engel, W., 19 Ennifar, E., 110

549

Author Index

Eppig, J. J., 460, 463, 488, 490–491, 493, 495–497, 499–501, 506 Eppig, J. T., 247, 536 Epp, T., 247 Ericson, J., 333 Eriksson, E., 376 Ermakov, A., 333 Ernst, S., 127 Esau, L., 417 Escaravage, J. M., 417 Escobar, J., 458 Escoffre, J.-M., 38 Eshkind, L., 116, 518 Espinosa, J. S., 175, 191 Esserman, L. J., 486 Evangelista, C., 515, 536 Evangelista, M., 333 Evans, H. H., 92 Evans, J. A., 38 Evans, M. J., 218–219 Evsikov, A. V., 485–488, 491, 495–497, 501, 506 Ewerling, S., 4 F Factor, S., 445 Falciatori, I., 19 Falco, G., 495, 514, 521 Fancher, K. S., 486–488, 491, 495 Fan, H., 4–5 Farago, A. F., 184, 186, 189–192, 195–196 Fare, T. L., 529 Farley, F. W., 156, 196 Farne, A., 515, 536 Faulkner, G. J., 500, 514–515 Faust, C., 303 Faust, N., 116 Favor, J., 331, 337 Federman, S., 486 Fedoriw, A., 461 Fehsenfeld, S., 116 Feil, R., 115–116, 157, 175 Feinberg, E. H., 303, 308, 333 Feldman, J. L., 198 Feldman, M., 246, 250–251 Felsher, D. W., 436, 441 Feng, W., 536 Feng, Y. Q., 373–374 Fernandez, L., 303, 308 Fernez-Bueno, C., 19 Feschotte, C., 72 Fieck, A., 300 Fields-Berry, S. C., 154, 185 Fiering, S., 191 Figurski, D. H., 147 Filutowicz, M., 137 Finch, R. A., 250 Fine, M., 274 Finnell, R. H., 223, 247, 517

Fire, A., 352, 369 Fischer, A., 445 Fischer, S. E., 60, 62, 73, 76, 93 Fishell, G., 174–175, 189, 191 Fisher, E., 331 Fishman, G. I., 437, 445 Flajnik, M. F., 54 Flaswinkel, H., 300, 308, 331 Flemr, M., 461 Floss, T., 112–113, 246–249, 259–260, 272, 279 Flowers, D., 458 Foissac, S., 485 Fojo, A. T., 536 Follenzi, A., 4 Fong, G. H., 417 Foo, S. Y., 359 Fordis, C. M., 39 Foreman, R., 443 Forest, M., 331 Forni, P. E., 167 Foroni, C., 518 Forrest, A. R., 514–515 Forrester, L. M., 250–251, 253, 287 Forrester, W. C., 352 Forsayeth, J. R., 197–198 Forster, T., 358 Foudi, A., 447 Fournier, T., 521 Frankenberg, S., 516 Franzesi, G. T., 198 Franz, G., 58 Franz, T. J., 331 Fraser, M. J., 224, 254 Fraser, M. J., Jr., 258 Fraser, S. E., 185 Frazier, J. P., 247, 250, 252–253, 255 Frendewey, D., 127, 417 Frenz, T., 186 Freundlieb, S., 432, 434, 446 Freyvert, Y., 444 Friddle, C. J., 250 Fridmacher, V., 441 Friedel, C., 246, 249 Friedel, R. H., 102, 243, 247, 272, 275, 279 Friedman, H., 446 Friedman, S. L., 518 Friedrich, G. A., 246, 248–249, 253, 255, 272–273, 283 Fuchs, H., 300, 304, 308, 310, 331, 343 Fuchtbauer, A., 486 Fuchtbauer, E. M., 247, 486 Fu, D., 443 Fu, J., 113–114, 117, 125–126, 130, 190, 196, 250 Fuller, S., 303 Furukawa, T., 38 Furuta, Y., 249, 303, 310 Fusco-DeMane, D., 436, 445 Fushiki, S., 518

550

Author Index G

Gaasterland, T., 476–477 Gage, F. H., 6 Gaiano, N., 272 Gailus-Durner, V., 331 Gal, A. B., 491 Galantino-Homer, H., 27 Gale, K., 335 Gale, N. W., 19, 127, 417 Galileo, D. S., 185 Gallay, P., 4 Gallicano, G. I., 19 Galloway, J., 113, 190, 196 Gambarotta, G., 252 Gao, Q., 443 Gao, W., 475 Garcia, E. L., 445 Garcia-Garcia, M. J., 303, 308, 333 Gardiner, B. B., 514–515 Gaudenz, K., 331 Gautier, L., 493, 532 Gautier, P., 343 Gavin, W. G., 27 Gawrys, L., 444 Gay, F., 352 Gebauer, F., 485 Gebuhr, T. C., 461 Gehring, W. J., 84, 272 Geisterfer-Lowrance, A. A., 220 Gellen, B., 446 Geng, Y., 434 Gentile, A., 252 Gentleman, R. C., 532 Gentry, J., 532 George, J., 536 Georgel, P., 331, 335 Gerber, A. S., 233 Gerber, D., 205 Gerfen, C. R., 185 Gerlich, D. W., 358 Gerstein, M., 477–478, 514 Gertsenstein, M., 9–11, 113, 185–186, 191, 443, 503, 521 Geurts, A. M., 60–62, 73, 84, 86, 94, 102, 444 Ge, W., 416 Ge, Y., 532 Ghosh, K., 110 Ghyselinck, N. B., 112–113, 246 Giannoukos, G., 536 Gibney, G., 304, 310 Gibson, M., 486, 488 Gilad, Y., 514 Gillette, M. A., 536 Gimenez, E., 438 Giovannini, M., 186 Giraldo, P., 5 Girard, A., 487 Giritharan, G., 461

Gish, G., 416 Gittenberger-de Groot, A. C., 517 Gizang, E., 484 Glaser, S., 109–110, 112–114, 116, 118–119, 126–127, 129, 430 Glazier, A. M., 343 Glenisson, P., 476–477 Glenister, P., 330–331 Godinho, S. I., 331 Goebel, M., 224, 254 Goff, S. P., 4, 218, 255 Goga, A., 352 Goggolidou, P., 333 Gogos, J. A., 437 Goldberg, I., 514 Golden, J. A., 154 Goldhamer, D. J., 189, 192 Goldstein, S., 39 Golestaneh, N., 19 Golini, E., 335 Golub, T. R., 516, 536 Gondo, Y., 244, 316, 343 Gong, S., 185 Goodnow, C. C., 331, 335 Good, P. D., 353 Goodrich, L. V., 248, 253 Gordon, J. W., 63 Gorlich, D., 445–446 Gosgnach, S., 198 Gossen, M., 376–377, 416, 429–434, 437, 445–446 Gossler, A., 245–246, 248, 272–273 Gotoh, Y., 38, 49 Goulding, M., 198 Graber, J. H., 481, 486, 495–497, 499–501, 506 Gradinaru, V., 197–198 Grafham, D., 308, 344 Gragerov, A., 248 Gragerova, G., 248, 446 Grahovac, M. J., 486–487, 512, 519 Grasser, F. A., 96 Graybiel, A. M., 198 Gray, G. E., 185 Gray, I. C., 300, 308, 331 Greene, K. S., 246, 250–251 Greenfield, A., 516 Gregg, K., 486 Greig, K. T., 304, 310, 336 Grenier, J. K., 359 Grignani, F., 255 Grimm, D., 370, 372 Grimwood, J., 458 Grindley, N. D., 110 Groen, A., 116 Gromoll, J., 19 Gronostajski, R. M., 147 Gross, C., 441 Grosveld, F. G., 19, 517

551

Author Index

Grotkopp, D., 244 Gruber, M., 417 Grueneberg, D. A., 359 Grundner-Culemann, E., 331 Grunweller, A., 369 Gruss, P., 113, 190, 249, 251, 275, 277 Guan, C., 102 Guan, K., 19 Guenet, J. L., 331 Guettier, J. M., 197–198 Gu, H., 114, 186, 191, 193, 368 Guigo, R., 485 Guinn, G. M., 299, 332 Gu, M., 486, 488 Gunther, E. J., 437 Guo, C., 186 Guo, F., 110 Guo, G., 220, 223–224, 230, 238, 252 Guo, Y., 436 Gupta, A., 250 H Haac, B. E., 504 Hacker, T., 333 Hackett, P. B., 224, 254 Hada, T., 459 Haddad, B. R., 19 Hadjantonakis, A. K., 113 Hagan, J., 300, 308, 331 Haigis, K. M., 102 Haines, B. B., 353 Hakak, Y., 515 Hakansson, J., 488 Hall, D. J., 226 Halliday, A., 185 Hall, J., 444 Hamalainen, R., 443 Hamatani, T., 459, 478, 487, 512, 516, 521 Hambardzumyan, D., 441–442 Hameyer, D., 116 Hamilton, C. M., 127 Hamlet, M. R., 58 Hammelbacher, S., 331 Hammer, R. E., 18, 30 Hammond, S. M., 512, 519 Hampl, A., 495–497, 506 Hancock, J. F., 102 Hancock, J. M., 307 Handler, A. M., 224, 258 Han, M., 224, 255, 258 Hanna, J., 443, 447 Hannenhalli, S., 461 Hannon, G. J., 352–353, 416, 445, 461, 486–487, 499 Hansen, G. M., 223, 247, 250 Hansen, J., 112–113, 246–249, 259–260, 283 Hansen, M. S., 11 Hant, P., 247

Han, X., 198 Haqq, C. M., 486 Harbaugh, C. R., 185 Harbers, K., 244 Harborth, J., 352, 416 Hardisty-Hughes, R. E., 335 Hardouin, N., 251 Harper, C. A., 247 Harrell, R. A., 258 Harris, E., 186 Harrow, J., 308, 344 Hart, A. W., 262, 343 Hartford, S. A., 335 Hartl, D. L., 233 Hartley, K. O., 272 Hartmann, J., 198 Haruna, K., 247, 251 Harwood, B. N., 504 Hasan, M. T., 433 Hasegawa, A., 486 Hasenfuss, G., 19 Hashimoto-Gotoh, T., 136 Hasuwa, H., 371, 389 Hatten, M. E., 185 Haussecker, D., 370 Haviernik, P., 517 Hawkins, R. D., 436, 439–441 Hawrylycz, M. J., 186, 191, 193 Hayashi, K., 286 Hayashi, S., 173 Haycraft, C. J., 333 Hay, E. D., 518 Heffner, S., 300, 308, 331 Hegemann, B., 127 Heimann, C., 116 Heintz, N., 127, 185–186 Heissmeyer, V., 416 Helinski, D. R., 137 He, M., 333 Hemann, M. T., 371 Hemley, S. J., 304, 310 Hemmers, S., 186 Henderson, N., 147 Hengstler, J. G., 518 Heninger, A. K., 354 Henkelman, R. M., 416 Hennek, T., 116 Hennessy, K., 258 Henschel, A., 370 Henseling, U., 245 Hentges, K. E., 303, 308, 310, 344 He´rault, Y., 115 Herb, J., 434 He´riche´, J. K., 127 Herlitze, S., 198 Hermann, T., 113, 127, 129 Hernando, E., 445 Herold, M. J., 372, 376, 446

552 Herron, B. J., 303, 308, 331, 333 Herzenberg, L. A., 191 Hess, J., 436 Hess, M. W., 18 He, Y. D., 529 Hickey, J. F., 417 Hicks, G. G., 247, 249, 283 Higgs, D. R., 113 Hill, A. A., 486 Hillen, W., 431–433 Hilton, A. A., 304, 310, 336 Hilton, D. J., 304, 310, 336, 339 Hingamp, P., 476–477, 536 Hinkle, G., 359 Hippenmeyer, S., 165, 171 Hipp, J., 513 Hirabayashi, M., 30 Hirashima, M., 252, 275, 287 Hirschi, K. K., 310 Hirst, E., 333 Hitotsumachi, S., 299, 316 Hitz, C., 373uˆ375, 389, 392, 405 Hjerling-Leffler, J., 189, 191 Hochberg, Y., 531 Hochedlinger, K., 191, 443, 446–447 Hock, H., 447 Hodges, E., 487 Hoebe, K., 331, 335 Hoess, R. H., 111, 156 Hofmann, A., 4 Hofschneider, P. H., 38 Hogan, B. L.M., 19, 536 Hogan, C., 380, 382 Holbrook, A. E., 486, 495–497, 501, 506 Hollatz, M., 112–113, 246, 248 Holstege, F. C., 416 Holter, S. M., 335 Homan, E. J., 5 Hong, E. J., 4, 9 Hong, N., 341 Honjo, H., 518 Honjo, T., 30, 287 Honoramooz, A., 27 Hood, K., 434 Hope, T. J., 5 Hopkins, C., 521 Horie, K., 60–62, 66, 71, 73–74, 76–78, 93, 220, 231, 238, 248, 446 Hormuzdi, S., 19 Horn, C., 78, 249, 259–260, 290 Horner, E., 303 Horner, J. W., 445 Hornig, N., 500 Horwitz, G. D., 198 Hosack, D. A., 475 Hostick, U., 198 Hough, T., 300, 308, 331 Hou, J., 517

Author Index

Hou, L., 304, 310, 336 Howard, B. H., 39 Hrabe de Angelis, M. H., 300, 308, 331, 337 Hsiao, E. C., 197–198 Hsieh, C. L., 444 Hsu, K. C., 514 Huang, C. C., 247, 444 Huang da, W., 475 Huangfu, D., 303, 308, 333 Huang, Q., 445 Huang, Y. Y., 224, 227, 230, 238, 252, 254, 258, 262, 273, 288, 436, 439–441 Huang, Z., 39, 352 Huarte, J., 485 Hua, Z. L., 116 Hu¨bner, M. R., 114 Hubner, N., 354 Hu, C. J., 417 Hu, E., 197–198 Huesken, D., 370 Hu, G., 354 Hughes, J., 113 Hughes, T. R., 513, 519 Hugill, A., 303, 330 Hung Wong, W., 493 Hunkapiller, J., 251 Hunkapiller, K., 519 Hunsicker, P. R., 299, 308, 332 Hunter, C. P., 486 Hunter, J., 330 Hunter, N. L., 157, 173, 190 Hurtley, S. M., 248, 253 Hu, S., 113, 117, 126, 130, 250 Hutchins, J. R., 127 Hutchison, K. W., 244, 499–501 Hvalby, O., 434 Hwang, S. S., 512, 521 Hwang, S. Y., 485–486, 491, 500 Hyland, C. D., 304, 310, 336 Hyman, A. A., 127 Hyvo¨nen, M. E., 18 I Iida, R., 252 Ikawa, M., 4, 9, 18–19, 34 Ikawa, Y., 130, 249 Ikeda, R., 77 Ilsley, D., 521 Imaizumi, T., 247, 251 Imamoto, A., 191 Incao, A., 304, 310, 336 Indra, A. K., 157, 175 Ingraham, R. H., 226 Inoue, K., 18–20, 25–27, 29–30, 33–34 Irgang, M., 259–260 Irizarry, R. A., 493 Irvine, K. M., 500

553

Author Index

Iscove, N. N., 486, 488 Ishida, Y., 76, 250, 252–253, 275–277, 284 Ishino, F., 19, 33 Issac, B., 536 Itohara, S., 196 Ivanyi, E., 113 Ivics, Z., 60, 72–73, 80, 92, 94, 103, 126, 224, 254–255 Iwama, A., 38, 49 Iwano, T., 19, 33 Iyer, V., 127 Izadyar, F., 30 Izsva´k, Z., 60–61, 103, 126, 224, 254–255 J Jackson, A. L., 370 Jackson, D., 127 Jackson, I. J., 343 Jacobson, J. W., 92 Jacque, J. M., 371 Jaenisch, R., 4–5, 19, 116, 191, 218, 223, 244, 443, 446–447 Jaffe, D. B., 536 Jahner, D., 5, 244 Jaisser, F., 446 Jalife, J., 445 James, D., 19 James, T., 353, 355 Janczewski, W. A., 198 Jaradat, S. A., 487, 512, 519 Jarosz, M., 478 Jasin, M., 372 Jay, P., 331 Jechlinger, M., 441–442 Jenkins, N. A., 115, 126–127, 186, 244, 254 Jenkins, S. S., 444 Jensen, P., 184, 186, 189–192, 195 Jensen, R. V., 519 Jensen, V., 434 Jerecic, J., 433, 437, 446 Jessee, J., 486 Jessell, T. M., 186, 191 Jia, L., 444 Jiang, J., 19 Ji, H., 92 Jitianu, C., 516 Ji, W., 486 Johansson, T., 186 Johnson, A., 484 Johnson, D., 486 Johnson, J. E., 38–40, 43, 189, 191 Johnson, K. R., 486 Johnson, R. L., 303, 310 Johnson, R. S., 417 Johns, S. J., 247 John, S. W.M., 221 Joliffe, I. T., 475

Jones, A. R., 186, 191, 193, 519 Jones, D. R., 478 Jones, E. G., 496 Jones, N., 416 Jones, W. D., 519 Jonkers, J., 116, 118, 233, 260 Jouault, H., 331 Joyner, A. L., 153, 159, 171, 185–186, 196, 245–246, 248, 251, 272 Ju, H., 437 Jung, H. J., 518 Jung, M., 300, 308, 331 Junqueira, M., 354 Justice, M. J., 92, 297, 300, 303, 305–308, 310, 316–317, 320, 331, 337–338, 341 K Kaczmarek, L., 444 Kadin, J. A., 247 Kaelin, W. G., Jr., 417 Kaji, K., 65, 443 Kakoki, M., 171 Kalaydjiev, S., 304, 310 Kallnik, M., 335 Kaloff, C., 247 Kamdar, S., 488, 490–491, 493, 495–497, 499 Kaminski, J. M., 224 Kamnasaran, D., 288 Kamoun, P., 331 Kanaar, R., 118 Kanatsu-Shinohara, M., 17–20, 25–27, 29–31, 33–34 Kandel, E. R., 205, 436, 439–441 Kaneda, M., 486–487 Kane, D. W., 536 Kanellopoulou, C., 416 Kang, J. J., 516 Kanisicak, O., 189, 192 Kanki, H., 196 Kaplan, G., 484 Kapushesky, M., 515, 536 Karasawa, M., 252 Karaskova, J., 220 Karayannis, T., 189, 191 Kargul, G. J., 486–487, 512, 519 Karlen, M., 333 Karsenty, G., 341 Kasahara, M., 54 Kasarskis, A., 303, 308, 331 Kasper, J., 113, 190, 196 Kato, M., 30 Kato, T., 18–19 Katz, L. C., 198 Kaufman, M. H., 45, 218–219 Kaul, A., 113 Kauppi, M., 304, 310, 336 Kauselmann, G., 116

554 Kawaichi, M., 250, 252, 275–276, 284 Kawakami, K., 56, 58, 272 Kawamoto, M., 247 Kawasaki, E. S., 519 Kawauchi, D., 39–40, 43 Kay, M. A., 353, 370 Kazan, K., 534 Kazazian, H. H., Jr., 55 Kazuki, Y., 18–19, 27, 33–34 Keane, J., 197 Keith, B., 417 Kelleher, D., 358 Kellendonk, C., 115 Keller, A., 197 Keller, G., 518 Keller, R. E., 185 Kelly, E. M., 298–299, 308, 332 Kelly, M., 4–6, 331 Kelly, O. G., 248 Keng, V. W., 58, 62–63, 66, 73–74, 76, 80, 99, 101, 103 Kennedy, C. J., 358 Kentner, D., 445–446 Kentros, C., 198 Kern, B., 341 Kern, H., 116, 416 Kerrison, J. B., 438 Kerr, W. G., 191 Kessler, B., 4 Kessler, D. A., 529 Kholodilov, N., 440 Khvorova, A., 369 Kile, B. T., 303–304, 310, 336, 339 Kilian, K. A., 529 Kim, D. Y., 333 Kimelman, D., 518 Kim, I. F., 515, 536 Kim, J. C., 19, 114, 157, 159–160, 164, 183–184, 186, 189, 191–192, 194, 196–198, 204, 354, 434 Kimmel, R. A., 173 Kim, N. H., 516 Kim, S. Y., 359, 445, 478, 536 Kim, T. K., 536 Kimura, S., 249 Kimura, Y., 500 Kim, Y. M., 518 King, A. M., 226 King, D., 335 King, L. A., 247 Kinoshita, T., 220, 231, 238 Kinsella, T., 255 Kiskinis, E., 443 Kistner, A., 433, 437, 446 Kitabatake, Y., 198 Kitada, K., 76 Kita, K., 18–19 Kitaya, K., 518

Author Index

Kitchen, J. R., 486 Kittler, R., 127, 352, 354 Klaften, M., 304, 310, 334 Klaver, B., 433 Klein, C. A., 441 Klein, H., 230 Kleinhammer, A., 387 Kleinman, M. E., 370 Klein, M. E., 198 Kleinridders, A., 416–417, 419, 446 Klejman, A., 444 Klemm, M., 223 Klempt, M., 304, 310, 334 Klevecz, R. R., 514 Klingenhoff, A., 514 Kloepfer, A. M., 359 Klopstock, T., 334 Klur, S., 486 Knight, C. R., 519 Knipscheer, P., 262 Knostman, K. A., 438 Knowles, B. B., 481, 485–488, 491, 495–497, 500–501, 506 Koche, R. P., 536 Koch, J. E., 529 Koch, L., 369 Kodama, H., 287 Kohara, Y., 486–487 Kohler, S. W., 300 Koike, T., 38 Kokkinaki, M., 19 Kokubu, C., 71, 73–74, 80, 82–83, 86–87 Kokubu, Y., 219 Kolb, A. F., 110, 159 Kolber, B. J., 440 Kolesnikov, N., 515, 536 Kolle, G., 514–515 Koller, B. H., 218 Kominami, R., 249 Ko, M. S.H., 459, 478, 486–487, 511–512, 514, 516–518, 520–521, 532, 535–536 Kondoh, G., 220, 231, 238 Kondrychyn, I., 58 Kong, J., 224, 227–228, 233, 235, 255 Konopka, W., 444 Kono, T., 486–487 Ko, N. T., 512, 521 Kool, J., 92, 94, 233, 260 Koop, K. E., 186, 191 Kornhauser, J., 335 Korn, R., 245 Koschmieder, S., 437 Koshibu, K., 433 Kossack, N., 19 Kostas, S. A., 352 Ko¨ster, M., 113 Kostov, G., 496 Kothary, R., 84, 245, 252

555

Author Index

Kotlikoff, M. I., 246, 250–251 Kotnik, K., 372, 376, 444 Kouno, M., 220, 231, 238 Koutsourakis, M., 127 Ko, Y., 478 Kramer, F., 446 Kramer, I., 191 Kramer, M. G., 418 Kranz, A., 109, 113–114, 116–119, 190, 196, 430 Kreppner, W., 186, 191 Krestel, H. E., 434 Kretz, P. L., 300 Krimpenfort, P., 116 Krizhanovsky, V., 445 Krugers, H., 434 Kuan, C. Y., 518 Kubista, M., 488 Kubo, A., 518 Kubota, H., 18–19, 25–26 Kubota, Y., 18–19, 33 Kugler, S., 434 Kuhnert, F., 287 Ku¨hn, R., 114–116, 307, 387–388, 416 Kuji, N., 459 Kulik, J., 444 Ku, M., 536 Kumar, P., 445 Kunath, T., 288, 371, 389, 416 Kuramochi-Miyagawa, S., 486–487 Kurimoto, K., 514, 520 Kushner, S. R., 127 Kuter-Luks, B., 416–417, 419, 446 Ku¨ter-Luks, S., 116 Kwak, S. P., 444 Kwon, H. J., 518 Kyweriga, M., 198 L Lababidi, S., 536 Labosky, P. A., 19 Lad, H. V., 335 LaGamba, D., 518 Lago, G., 247 Lai, L., 310 Lakso, M., 18 Lallemand, Y., 165 Lalouette, A., 331 Lamenzo, J. O., 516 Lam, S., 444 Landel, C. P., 444 Lander, E. S., 536 Landry, J., 517 Landsberg, R. L., 190 Lane, H. C., 475 Lanfrancone, L., 255 Langford, C., 343 Lang, J. E., 486 Lan, Q., 247, 253

Lao, K., 476–477, 495–497, 500–501, 506 Largaespada, D. A., 88, 92–93, 226, 254 Larkins, C. E., 333 Larralde, O., 253 Larson, D. M., 304, 310 Lassmann, T., 500 Latham, K. E., 461, 486–488, 491, 495 Latour, A. M., 218 Laub, F., 518 Laux, T., 165 Lawinger, P., 255, 283 Lawlor, S. C., 536 Lawson, K. A., 154 Lechner, H. A., 198 Le, C. T., 520 Lederman, B., 223 Leder, P., 76, 250, 253, 275, 277, 284 Le Douarin, N., 185 Ledoux, P., 515, 536 Lee, C., 476–477, 495–497, 500–501, 506 Lee, D.-F, 351 Lee, E. C., 111–112, 127, 186, 331 Lee, G., 110–111 Lee, H. G., 440 Lee, J., 18–20, 25, 27, 30, 33–34, 246, 250–251 Lee, J. D., 303, 308, 333 Lee, J. H., 19 Lee, J. S., 196 Lee, K. Y., 519 Lee, M., 248 Leenders, H., 437 Lee, N. S., 388 Lee, P., 445 Lee, R. E., 247 Lee, S. K., 219, 445 Lefebvre, C., 474 Lefebvre, L., 220 Lefkowitz, S. M., 519 Legue´, E., 153, 164, 177 Lehmann, T., 518 Leighton, P. A., 248, 253, 287 Lein, E. S., 186, 191, 193, 198 Lemberger, T., 116 Lemberg, H., 436, 445 LeMeur, M., 115 Lemischka, I. R., 351, 353, 355 Lempicki, R. A., 475 Lendeckel, W., 352, 416 Lengeling, A., 223 Lengner, C. J., 443 Leng, Y., 354 Leong, B., 536 LeProust, E., 529 Lequarre, A. S., 488, 490 Lerchner, W., 198 Ler, E. S., 72 Lerner, C., 416 LeRoith, D., 113–114, 190, 196 Leroy-Viard, K., 331

556 Lesaffre, B., 252 Lester, H. A., 198 Lewandoski, M., 186, 196, 518 Lewin, H. A., 478 Lewis, D. L., 370–371 Lewis, M. A., 343 Liang, H. C., 518 Liang, Q., 59, 94, 115, 224, 227–228, 233, 255 Liang, Y., 446 Libby, B., 221 Libby, B. J., 223 Li, C. H., 247, 249, 341, 416, 493 Licht, T., 440 Lickert, H., 416 Li, C. Y., 247 Li, E., 218 Lieberman, E., 536 Lieberman, J., 388 Liem, K. F., Jr., 333 Li, F., 353, 355 Li, G., 224, 255, 258 Li, L., 371 Lilleberg, S. L., 249 Li, M., 19, 278–279 Li, M. A., 126, 217, 224 Lim, M. K., 486–487, 512, 519 Lim, T., 3 Lim, W. K., 474 Lin, C., 227–228, 254–255, 258 Lin, C. C., 30, 444 Lin, C.-S., 186 Lindahl, M., 18 Lindpaintner, K., 486 Lin, J., 520 Lin, Q., 252, 275–276 Linsley, P. S., 370, 416 Lin, T. Y., 335 Lin, X., 370 Li, P., 444 L’Italien, L., 247 Litingtung, Y., 331 Liu, A., 333 Liu, B., 303, 308, 310, 344 Liu, G., 94 Liu, H., 518 Liu, J., 536 Liu, L., 64 Liu, P., 227–228, 233, 252, 254–255, 258, 443 Liu, S., 247, 303, 308, 331 Liu, X., 19 Liu, Y. J., 444 Livesey, F. J., 486 Livet, J., 160–161, 175 Livigni, A., 288 Li, X., 198, 258 Li, X. M., 247, 249 Li, X. Y., 516 Lloyd, K. C., 115

Author Index

Lobanenkov, V., 461 Lobe, C. G., 159, 186, 191 Lobo, N., 59 Lodato, M. A., 443 Loftus, S. K., 304, 310, 336 Logie, C., 112, 115 Lohler, J., 244 Loh, Y. H., 536 Lois, C., 4, 9 Lomeli, H., 186, 191 Long, A. D., 358, 532 Longo, L., 19 Loonstra, A., 116, 118, 167–168 Lopez, J. M., 485 Lorens, J., 255 Loring, J., 19 Lounas, V., 110, 112 Lovel, K. L., 116, 118–119 Lowe, S. W., 445 Low, R., 445–446 Lowrey, P., 335 Lubbert, H., 433, 437, 446 Lubitz, S., 116, 118–119 Luche, H., 161 Lu, D., 227–228, 254–255, 258 Luna-Crespo, F., 113 Lund, A. H., 233, 260 Luo, B., 359 Luo, G., 60, 72, 84, 94, 255 Luo, J., 27 Luo, L., 175, 191, 197 Luo, Y., 519 Luquet, S., 419 Lutz, M., 249 Lu, W., 303, 308, 331 Lu, X., 248, 253 M Mabuchi, I., 130 MacGregor, G. R., 514 Machold, R. P., 175, 189, 191 Mack, V., 441 Maddux, S. C., 299, 308, 332 Madisen, L., 186, 191, 193, 248, 446 Maeda, Y., 220, 231, 238 Magbanua, M. J., 486 Magen, A., 514 Mager, J., 478 Magnuson, T., 221, 303, 461 Mahtab, E. A., 517 Maier, L. S., 19 Mai, J. J., 114, 184, 186, 189–192 Maika, S. D., 18, 30 Mainguy, G., 252, 287 Maione, F., 246 Ma, J., 461 Majors, J., 185

557

Author Index

Makrigiorgos, G. M., 486 Maleszewski, M., 444 Malik, R., 461 Mallet, J., 444 Malouf, N. N., 218 Malumbres, M., 102 Mamo, S., 491 Mandillo, S., 335 Mane, S. M., 514 Mangues, R., 102 Manners, J., 534 Manning, D. K., 341 Manova, K., 303, 308, 331 Mansour, S. L., 130 Mansuy, I. M., 433, 436, 441 Mantamadiotis, T., 116 Mao, M., 519 Mao, Q., 353 Ma, P., 460 Maquat, L. E., 275 Marchandise, J., 488, 490 Marchuk, D. A., 303, 308 Mardis, E. R., 514 Marin de Evsikova, C., 485–486, 491 Marioni, J. C., 514 Markesich, D. C., 223, 247 Markoulaki, S., 443, 447 Markowitz, D., 255 Mar, L., 333 Marques-Kranc, F., 113 Marschall, S., 300, 308, 331 Marshall, K. A., 515, 536 Martin, D. I., 446 Martinez de Velasco, J., 186 Martin, G. R., 196 Martin, J. R., 197, 331 Martin, P. R., 495 Marton, M. J., 519, 529 Martynov, V. I., 159 Maruyama, T., 459 Marzio, G., 433 Mascrez, B., 115 Mason, C. E., 514 Masui, S., 38, 49, 518, 536 Masuya, H., 343 Masuyama, N., 38, 49 Matera, I., 304, 310, 336 Ma´te´s, L., 61, 78, 94, 126, 224 Matoba, R., 495, 514, 518, 535–536 Matsuda, E., 252, 275, 286 Matsui, Y., 19 Matsukura, S., 376 Matsumoto, N., 518 Matta, J., 517 Matteoli, M., 197 Mauer, J., 416 Maxson, R. E., 444 Maxwell, A., 308, 310, 344

Mayer, U., 334 Mayford, M., 205, 434, 436, 439–441 May, S. R., 333 McAleer, M. A., 233 McAuliffe, F., 416 McBride, J. L., 371 McCaffrey, A. P., 353, 371 McCarrey, J. R., 514 McCarthy, K. D., 197–198 McClintock, B., 54, 92 McCue, K., 501, 514–515 McCulloch, E. A., 29 McDermott, J., 111, 117, 126, 130 McDonald, J., 331, 335 McDonald, J. A., 437 McDonald, J. D., 303, 308, 317, 331–332 McEachern, M. J., 137 McGrew, M. J., 4 McHugh, T. J., 197 McInerney, P., 478 McJunkin, K., 445 McKie, L., 343 McKinney, S., 198 McLaughlin, J., 485, 491, 500 McLay, R., 444 McMahon, A. P., 116, 173, 335, 374–375 McManus, M. T., 353 McMath, H., 303 McNamara, J. O., 198 Mear, J. P., 518 Medico, E., 252 Meek, S., 444 Megee, S., 27 Mehrad, B., 436 Mehrian-Shai, R., 444 Mehta, S., 518, 534, 536 Meir, Y. J., 224 Meister, G., 352 Meistrich, M. L., 20, 26 Melchner, H., 112–113 Melican, D. T., 27 Mello, C. C., 352 Melton, D. A., 516 Meltzer, L., 197–198 Melvin, D., 227–228, 254–255, 258 Mencarelli, A., 255 Mendez, R., 485 Meneses, J., 19 Meng, E. C., 247 Meng, F., 496 Meng, X., 18, 30, 444 Mercadier, J. J., 446 Mercer, E. H., 186, 205 Merino, G., 147 Merrick, W. C., 253 Metcalf, D., 304, 310, 336 Metzger, D., 115–116, 157 Meuse, L., 353

558 Meuwissen, R., 116 Meyer, B. I., 249, 275, 277 Meyer, J. E., 110 Meyer, M. R., 519, 529 Meyers, E. N., 196, 518, 534, 536 Meyer, W. K., 446 Meziane, H., 335 Michael, I. P., 443 Michael, S. K., 116 Michalon, A., 433 Michel, D., 334 Mifsud, S. L., 304, 310 Miki, H., 18–20, 25–27, 29, 33–34 Mikkelsen, T. S., 536 Mikkers, H., 233, 260, 262 Mileikovsky, M., 443 Milenkovic, L., 478, 487–491, 493, 516 Miller, D., 337 Miller, J. C., 444 Miller, W., 486 Millet, A., 446 Mills, A. A., 310, 445 Milos, P. M., 478 Minamishima, Y. A., 417 Mintz, B., 4, 244 Mishina, M., 237 Miskey, C., 67 Mitchell, K. J., 248, 253, 283, 308, 344 Miura, K., 251 Miyada, C. G., 458–459, 463, 474 Miyado, K., 459 Miyagi, S., 38, 49 Miyazaki, J., 165, 191, 518 Miyoshi, G., 189, 191 Miyoshi, H., 6 Mizusawa, Y., 459 Mizutani, K., 38–40, 43, 49 Mizutani, Y., 38 Mocanu, M. M., 440 Mochimaru, Y., 459 Model, P., 127 Modelski, M., 27 Modi, C., 486, 488 Moens, C., 113 Moe, S., 248 Moffat, J., 359 Mohan, S., 304, 310 Mohseni, P., 443 Moisyadi, S., 224 Monaghan, P.-A., 115 Montagutelli, X., 331 Montecucco, C., 197 Montesinos, M. L., 252 Montgomery, K., 341 Montgomery, M. K., 352 Montini, E., 64 Montoliu, L., 5 Moore-Jarrett, T., 252

Author Index

Moore, T., 353, 355 Mootha, V. K., 536 Moqrich, A., 515 Moran, J. L., 115, 303, 308, 341 Moreno-Pelayo, M. A., 343 Morgan, B. J., 475 Morgan, J. E., 343 Mori, M. A., 444 Morimoto, T., 29 Moritz, S., 288 Moriyama, Y., 38 Morley, G., 445 Morris, J. H., 247 Morrison, G. M., 288 Mortazavi, A., 501, 514–515 Mortensen, R. M., 220 Moser, A. R., 331, 337 Moskowitz, I., 310 Moslehi, J., 417 Moss, J. E., 248, 253 Mountford, P., 279 Mowrer, G., 518, 534, 536 Moy, S. S., 198 Muccino, D., 116 Muenke, M., 331 Mujica, A., 127 Mukherjee, S., 536 Muller, G., 432 Mu¨ller, W., 114 Mulligan, J. T., 147 Muncan, V., 416 Munroe, R. J., 221 Murakami, F., 40 Muramatsu, M., 38, 49 Muramatsu, T., 38 Murchison, E. P., 461, 486–487, 499 Murcia, N. S., 333 Muroyama, Y., 43 Murphy, A. J., 19, 127 Murphy, J. M., 304, 310, 336 Muta, M., 247, 251 Muyrers, J. P., 115, 126–127, 136, 186 Muzumdar, M. D., 191 Myers, R. M., 496 Myslinski, E., 373, 375 N Naf, D., 223 Nagano, M., 19, 27, 29, 33 Nagarajan, V., 506 Nagaraja, R., 487, 512, 519 Nagaya, M., 443 Nagy, A., 9–11, 87, 113, 126, 185–186, 191, 220, 224, 251, 396, 503, 521 Naiche, L. A., 118, 167 Nakagata, N., 247, 251 Nakamura, E., 167, 173

559

Author Index

Nakamura, H., 247, 251, 303, 310 Nakamura, T., 19 Nakanishi, S., 198 Nakano, T., 287, 486–487 Nakashiba, T., 197, 440 Nakatake, Y., 518, 534, 536 Nakatsuji, N., 30, 38–40, 43, 49 Nakauchi, H., 38, 49 Naldini, L., 4 Narasimhan, S., 536 Naruse, I., 237 Nashmi, R., 198 Nathans, J., 116 Naumann, R., 113–115, 117, 190, 196 Navas, C., 248 Nawshad, A., 518 Nayernia, K., 19 Nedorezov, T., 535–536 Nelson, C., 512 Nelson, H., 223 Nelson, M. L., 434 Nelson, P., 369 Nesterova, M., 519 Neuhauser, A., 299 Neuhauser-Klaus, A., 299, 331, 337 Neuhaus, H., 113, 245 Neumann, E., 38 Nevian, T., 434 Ng, L. L., 186, 191, 193 Nguyen Dinh Cat, A., 446 Nguyen, H. N., 19 Nguyen, M. T., 171, 173 Nicholas, C., 19 Nicholls, R. E., 440 Nichols, C. D., 198 Nichols, J., 219, 279 Nicola, N. A., 304, 310, 336 Nicolas, J. F., 164, 169, 177 Nicolelis, M. A., 198 Niehaves, S., 416–417, 419, 446 Niemann, H., 197 Nie, T., 440 Ning, Z., 227–228, 254–255, 258 Nishijima, I., 310 Nishimoto, M., 38, 49 Nishiyama, A., 518, 534–536 Nissenson, R. A., 197–198 Niswander, L. A., 303, 310, 333 Nitzsche, A., 127, 354 Niwa, H., 38, 49, 76, 165, 191, 249, 275, 518, 535–536 Niwa, O., 18–19 Nobis, P., 244 Noda, T., 58, 343 Nolan, G. P., 255 Nolan, P. M., 300, 308, 331, 335 Nolte, J., 19 Nonneman, R. J., 198

Noppinger, P. R., 249, 259–260 Nord, A. S., 247, 249, 272, 278, 280, 286 Nordman, E., 476–477, 495–497, 500–501, 506 Nordmann, Y., 331 Norman, A. R., 251 Novak, A., 159–160, 186 Noveroske, J. K., 300 Nuber, S., 440 Nybakken, K., 335 O Oatley, J. M., 30 Obata, Y., 486–487 Oberdoerffer, P., 372–373, 375, 416 OBrien, M. J., 460, 463 O’Brien, W., 303, 310 Ocbina, P. J., 333 Ochman, H., 233 Ogawa, T., 18–19, 31, 33–34 Ogonuki, N., 18–20, 25–27, 29–30, 33–34 Ogura, A., 18–20, 25–27, 29–30, 33–34 Oh, B., 485–486, 491, 500 Ohbo, K., 18–19, 116, 118–119 Ohinata, Y., 514, 520 Ohishi, M., 247 Ohkawa, J., 375–376 Ohmori, Y., 38 Ohmura, M., 18–19, 33 Oh, S. W., 186, 191, 193 Oike, Y., 247, 251 Ojala, E., 446 Okabe, M., 377–378 Okada, Y., 87 O’Kane, C. J., 84, 197, 272 Okochi, H., 518 Okuda, A., 38, 49, 518 Okumura, J., 38 Olena, A. F., 513 Olson, E. N., 341 Om, J., 221 Ono, M., 518 Ono, Y., 514, 520 Orkin, S. H., 354 Ornitz, D. M., 341 Orth, A. P., 515 Orwig, K. E., 19, 26, 30, 33–34, 444 Osborne, L. R., 247 O’Shaughnessy, P. J., 504–505 Oshimura, M., 18–19, 27, 33–34 Oshlack, A., 514–515 Osipovich, A. B., 250–252, 275–276, 283 Osten, P., 434 Ostermeier, A., 219 Ostertag, E. M., 55 Ouagazzal, A. M., 335 Overton, S. A., 27 Oving, I., 416

560

Author Index

Owczarek, D., 444 Ow, D. W., 111–112 Owens, G. C., 185 Ozil, J. P., 460 Ozsolak, F., 478 P Paddison, P. J., 352–353, 355, 371, 388, 399, 416 Paeper, C., 416 Pakhomov, A. A., 159 Palais, G., 446 Pallai, P. V., 226 Palmenberg, A. C., 226 Palmiter, R. D., 186, 191, 193, 419 Panek-Huet, N., 446 Pang, Z. P., 219 Pan, H., 460–461, 463, 486, 499 Pantano, S., 487, 512, 519 Pant, D., 19 Pan, W., 520 Papaioannou, V. E., 118, 167 Pargent, W., 300, 308, 331 Parker, A., 335 Parker, J. S., 512, 519 Park, F., 13 Parkhomchuk, D., 514 Parkinson, H., 515, 536 Parkinson, N., 330 Park, J. W., 486 Park, M., 521 Parks, G. D., 226 Parvinen, M., 18 Passman, R. S., 437 Paszkowski-Rogacz, M., 354 Patel, B., 488 Patsch, C., 113, 117, 126, 130, 250 Patton, J. G., 513 Paul, C. P., 353, 371 Paulovich, A., 536 Paulson, H. L., 353 Paulson, J. B., 440 Pausch, M. H., 198 Pavan, W. J., 304, 310, 336 Pavlova, M., 248 Pavlova, M. N., 446 Pawlak, M., 247, 249 Pawling, J., 113 Pawson, T., 416 Paynton, B. V., 483–485 Pease, S., 4, 9 Peaston, A. E., 481, 486–488, 491, 495, 501 Pei, Y., 197–198 Pelczar, P., 224 Pelicci, P. G., 255 Pelletier, L., 127 Pellow, J. W., 274 Peluso, I., 246

Peng, S., 371 Pera, M. F., 444 Perrimon, N., 335 Person, C., 249 Peschle, C., 255 Petersen, C. P., 353 Petersen, G., 54 Peters, G., 92 Peters, H., 303, 308, 331 Peters, J., 300, 308, 331 Peterson, A. S., 245, 333, 446 Petit, A. C., 164 Pettitt, S. J., 115, 217 Pfeifer, A., 3–6, 9 Phamluong, K., 333 Pham, T. T., 353 Philipsen, S., 517 Phipps, E. L., 299, 308, 332 Piao, Y., 487, 495, 511–512, 514, 517–519, 534–536 Piccart, M. J., 458 Pichel, J. G., 18 Piedrahida, A., 111 Pieles, G., 333 Pietsch, J., 272 Piggott, J., 250 Piipari, M., 343 Piko, L., 483–485 Pinson, K. I., 248, 253 Pinto, L., 335 Piotrowska, K., 478, 487–491, 493, 516 Piqani, B., 359 Pique, M., 485 Pirity, M., 113 Pirooznia, M., 506 Pitot, H. C., 331 Plasterk, R. H., 224, 254 Plath, K., 191, 446 Platt, A. R., 478 Pleasance, E. D., 92, 101 Plessy, C., 500 Plum, L., 369, 416 Podsypanina, K., 441–442 Poeschla, E., 4 Polfi, P. P., 19 Pomeroy, S. L., 536 Pon Fong, M. T., 416 Popova, E., 444 Porteous, D. J., 233, 259, 290 Portet, T., 38 Poser, I., 127 Pot, C., 437 Potter, P, 331 Potter, S. S., 518 Poueymirou, W. T., 127, 417 Powles-Glover, N., 333 Pozniakovski, A., 127 Pozniakovsky, A., 127

561

Author Index

Prabhu, V., 519 Preis, J. I., 304, 310 Premsrirut, P. K., 445 Prieto, J., 418 Prigge, J. R., 115, 118 Pritham, E. J., 72 Probst, F. J., 297, 308, 344 Prochiantz, A., 252 Proctor, B. M., 437 Prosser, H. M., 115 Proteau, G. A., 147 Provost, G. S., 300 Psaltis, G., 444 Putman, D. L., 300 Q Qian, X., 198 Qian, Y., 495, 512, 521 Qiu, H., 333 Quackenbush, J., 476–477, 536 Quint, E., 343 Qureshi, S. J., 100 Qu, S., 246, 250–251 Quwailid, M. M., 303 R Raatikainen-Ahokas, A., 18 Radcliffe, P. A., 4 Radford, E. E., 486–488, 491, 495, 504 Radomska, H. S., 436 Rad, R., 91, 224 Rahman, A., 250 Rairdan, X. Y., 115 Raja, R., 529 Rajendran, L., 370 Rajewsky, K., 114–116, 118, 196, 416, 438, 445–446 Rakeman, A. S., 303, 308, 333 Rambhatla, L., 488 Ramer, S. W., 147 Ramirez, F., 518 Ramirez-Solis, R., 233, 250 Ramsay, R. G., 304, 310 Ramsden, M., 440 Rance, R., 233, 260 Rao, A., 416 Rao, C., 303, 308, 331 Rassoulzadegan, M., 115 Rathkolb, B., 300, 304, 308, 310, 331, 334 Ratnakumar, K., 353, 355 Rauvala, H., 18 Ravimohan, S., 226, 254 Rayburn, H., 246, 249 Raymer, G. D., 299 Raymond, C. S., 118, 157, 190, 238, 250, 415 Ray, R. S., 183, 190 Reboredo, M., 418

Reddy, S., 246, 249 Redshaw, N., 343 Regehr, W. G., 184, 191–192, 196–198 Reichardt, H. M., 446 Reid, L. H., 519 Reid, S. W., 405 Reid, T., 247 Reifenberger, J. G., 478 Reijo-Pera, R. A., 19 Reinhold, W. C., 536 Reis, A., 300, 308, 331 Reiter, J. F., 251, 333 Rempel, R., 483, 485 Resnick, J. L., 19 Reynolds, A., 370 Rhode, A., 248 Rice, P. A., 110 Richard, H., 514 Richardson, J. E., 247 Richardson, M., 113 Richter, J. D., 485 Richter, L. J., 223, 247, 250 Richter, T., 300, 308, 331 Rientjes, J. J., 136 Riganelli, D., 255 Rijswijk, L., 116 Rinaudo, P. F., 461 Rinchik, E. M., 337 Ringrose, L. H., 110, 112, 114, 129, 190 Risley, M. D., 308, 344 Ritchie, W. A., 4 Rivas, C., 304, 310, 336 Rivkin, E., 333 Robb, L., 116, 118–119 Roberts, A. P., 54 Roberts, A. W., 304, 310 Roberts, D. M., 287 Robertson, E. J., 218–219 Robertson, M., 219, 279 Robinson, M. P., 38 Robson, J. E., 333 Rockman, H. A., 303, 308 Rode, A., 116 Rodriguez, C. I., 113–114, 184–186, 189–190, 192, 196 Rodriguez-Zas, S. L., 478 Roes, J., 114 Roessler, E., 331 Rogan, S. C., 197–198 Rogers, D., 300, 308, 331 Roguev, A., 127 Rohde, A. D., 446 Rojas, J., 127 Ro¨llig, D., 116, 118–119 Rols, M.-P., 38 Roma, G., 246–247, 279, 286 Romeijn, L., 262 Romito, A., 283

562

Author Index

Rone, J. D., 19 Rosen, B., 127 Rosen, E., 224, 254 Rossant, J., 113, 245–248, 252, 284, 416 Rossi, F., 114, 190 Rossi, J. J., 352–353 Ross-Macdonald, P., 54 Rossmeissl, P. J., 111 Roth, B. L., 197–198 Roth, P., 248, 446 Roth, S., 438 Rothstein, J. L., 486 Rotolo, T., 116 Rouyer-Fessard, P., 331 Rowitch, D. H., 116 Roy, M., 197–198 Rubenstein, J. L., 333 Rubinson, D. A., 371, 389 Rubio-Aliaga, I., 304, 310 Rucker, E. B., 111 Rudnev, D., 515, 536 Ruiz, P., 112–113, 246–249 Ruley, H. E., 246–247, 249–252, 275–276 Russell, L. B., 298, 320 Russell, W. L., 298–299, 308, 316, 332 Rustici, G., 515, 536 Ryan, M. D., 226 Ryder, E. F., 185 Ryu, B. Y., 19, 26–27, 30 S Saba, R., 38–40, 43 Sachidanandam, R., 487 Sadelova, S., 521 Sadowski, P. D., 147 Sagai, T., 343 Sainio, K., 18 Saito, F., 247, 251 Saitoh, H., 38 Saito, I., 110–111 Saito, T., 37–40, 43, 45, 49 Saitou, M., 514, 520 Sajjadi, F., 218 Sakaki, Y., 486–487 Sakurai, T., 39 Salamonsen, L. A. 463 Salinger, A. P., 303, 306, 310, 317, 337–338 Salisbury, J., 499–501 Salles, F. J., 485 Salminen, M., 249, 275, 277 Salomonis, N., 536 Sam, M., 251 Sanbo, M., 237 Sandig, V., 186 Sands, A. T., 223, 247, 249 Sanes, J. R., 154, 185 Sanges, R., 247

Sano, Y., 486–487, 512, 519 Santacruz, K., 440 Santoro, M., 30 Santos, N., 251 Santoyo-Lopez, J., 358 Sapinoso, L. M., 515 Sarao, R., 438 Sardiello, M., 246–247 Sariola, H., 30 Sarkis, C., 5 Sarov, M., 127, 136 Sasaguri, H., 372 Sasaki, E., 4 Sass, L. E., 478 Sato, T., 171 Sato, Y., 58 Sauer, B., 111, 117, 126, 130, 147, 156, 191 Sauvain, M. O., 446 Savakis, C., 58 Sawada, H., 33 Sawinski, J., 434 Sawitzke, J. A., 127 Scales, S. J., 333 Schaefer-Ridder, M., 38 Schaeffer, L., 501, 514–515 Schafer, X. L., 353, 355 Schaft, J., 87, 113, 127, 129, 156, 190, 381 Schaniel, C., 351, 353, 355 Scheel, J. R., 288 Schellander, K., 478 Schelter, J. M., 519 Schenk, P., 534 Scherer, S. W., 331 Scherf, M., 514 Scherr, D. S., 19 Scherz, P. J., 248, 253, 333 Schiavo, G., 197 Schimenti, J. C., 221, 223, 335 Schimenti, K. J., 221, 335 Schimpf, B. A., 446 Schindler, J. W., 447 Schlake, T., 111, 373 Schmidt, E. E., 115, 118 Schmidt-Supprian, M., 118 Schmitt, S., 116 Schmutz, J., 458 Schneider, E., 221 Schneider, M. B., 197–198 Schneider, S., 127 Schnieke, A., 244 Schnu¨tgen, F., 85, 112–113, 246–250, 255, 259–260, 272, 278–279, 283 Scho¨nig, K., 116, 368, 429, 431, 434–435, 438, 445–446 Schoor, M., 116, 245 Schrock, M., 341 Schroder, K., 500

Author Index

Schultz, R. M., 457, 460–461, 463, 478, 484, 486–487, 499, 516 Schulze, E. N., 444 Schulz, H., 354 Schulz, M. H., 514 Schu¨tz, G., 115 Schwartzberg, P. L., 218 Schwartz, R. J., 517 Schwarz, D. S., 369 Schwenk, F., 114–116, 118–119, 196, 367, 371–372, 416–417, 419, 434, 438, 445–446 Scott, B. B., 4 Scott, J. H., 486 Scott, M. M., 184, 186, 189–192, 195 Scott, M. P., 478, 516 Searles, L. L., 54, 56 Seberg, O., 54 Sedlmeier, R., 334 Seeburg, P. H., 114, 434 Seed, B., 127 Seel, M., 19 Seibler, J., 116, 118–119, 367–368, 371–374, 376–378, 389, 392, 396, 412–413, 416–417, 419, 444, 446 Seidman, J. G., 220 Seifert, M., 514 Seiler, S., 445 Seisenberger, C., 112–113, 246–248, 259–260 Sekiguchi, M., 136 Sekimoto, T., 247, 251 Selfors, L. M., 358 Semb, H., 488 Senecoff, J. F., 111 Seng, T. W., 226 Seol, A. D., 251 Sequerra, R., 113, 190, 196 Serguera, C., 5 Severin, J., 127 Sgaier, S. K., 171 Shaffer, L. G., 529 Shah, J. V., 333 Shaik, N., 514, 536 Shakya, R., 436 Shanks, E., 358 Shannon, K. W., 519 Shao, X. M., 198 Sharov, A. A., 478, 487, 495, 511–512, 516–518, 520–521, 532, 534–536 Sharova, L. V., 514, 517, 535–536 Sharpe, J. A., 113 Sharp, P. A., 353 Shaunak, S., 4 Shaw-Smith, C., 113–114 Shaywitz, D. A., 516 Shedlovsky, A., 331 Shefi, S., 19 Shen, C., 184, 191–192, 196–198 Shendure, J., 92

563 Sheng, Y., 444 Shen, X. H., 516 Shen, Y., 333 Sherlock, G., 476–477, 536 Sherman, A., 4 Sherman, B. T., 475 Sherratt, D. J., 186 Sherwood, R. I., 516 Shevchenko, A., 354 Shibata, A., 38 Shi, E. G., 247, 249 Shigeoka, T., 250, 252, 275–276, 284 Shigetani, Y., 130 Shi, L., 519 Shimada, A., 4 Shima, N., 335 Shimizu, E., 441 Shimizu, K., 343 Shim, J. S., 518 Shimosato, D., 518 Shimshek, D. R., 114, 116, 434 Shin, M. K., 415 Shinohara, E. T., 64 Shinohara, T., 17–20, 25–27, 29–31, 33–34 Shippy, R., 519 Shiroishi, T., 343 Shi, Y., 352 Shmelkov, S. V., 19 Shojatalab, M., 515, 536 Shoji, M., 459 Shook, N. A., 189, 192 Short, J. M., 300 Shui, B., 246, 250–251 Shukla, V., 416 Shumacher, A., 303 Shuttleworth, J. J., 444 Shyu, A. B., 289 Siddiqui, A., 476–477, 495–497, 500–501, 506 Siepka, S. M., 341 Sigrist, M., 191 Siguier, P., 54 Silva, J., 287, 444 Silver, D. L., 304, 310, 336 Silver, J. D., 339 Silver, P. A., 371 Simon, I., 244 Simon, M. C., 417 Singer, B., 299 Singh, A., 250–251, 275–276 Singh, P., 495–497, 506 Singla, V., 251, 285 Sinzelle, L., 67 Siolas, D., 416 Skarnes, W. C., 115, 125, 127, 245–249, 251, 253, 272–275, 278, 280 Skowronski, J., 486 Slabicki, M. M., 352, 354 Slimko, E. M., 198

564 Slobodskaya, O., 446 Smallwood, P. M., 116 Smerdou, C., 418 Smith, A., 219, 279, 287, 444 Smith, A. A., 113, 127 Smith, A. G., 518 Smith, A. J.H., 113, 127 Smithies, O., 83, 218 Smith, L., 516 Smith, M. C., 110 Smith, Q., 358 Smith, R., 221 Smye, S. W., 38 Snouwaert, J. N., 218 Snyder, M., 477–478, 514 Soboleva, A., 515, 536 Soejima, Y., 247, 251 Soewarto, D., 300, 304, 308, 310, 331, 337 Soininen, R., 245 Soldatov, A., 514 Solloway, M. J., 333 Solter, D., 485–488, 491, 495–497, 500–501, 506 Song, H., 444 Soohoo, C., 352 Sorge, J. A., 300 Soriano, P., 114, 118, 129, 157, 171, 184, 186, 189–192, 196, 238, 243, 246–250, 252–253, 255, 259, 263, 272–273, 283, 416 Sosic, D., 341 Sotiriou, C., 458 Sousa, V. H., 189, 191 Southern, E. M., 12 Southey, B. R., 478 Southon, E., 517 Spector, D. L., 513 Speed, T. P., 496 Spellman, P., 476–477, 536 Spergel, D. J., 114 Spirig, D. H., 433 Spottswood, M., 147 Sprengel, R., 114, 116, 518 Spurr, N., 300, 308, 331 Squire, J. A., 220 Srinivasan, S., 197–198 Srinivas, S., 171, 186 Srivastava, A. K., 304, 310 Stadtfeld, M., 443 Staerk, J., 443 Stagg, C. A., 495, 514, 521 Stahlberg, A., 488 Stalker, J., 224, 227–228, 233, 235, 255 Stanchfield, J., 529 Stanford, W. L., 247, 249, 252–253, 271, 273, 280, 287 Stangegaard, M., 488 Stark, W. M., 186 Starr, T. K., 66, 99 Steele, M. H., 299

Author Index

Steffen, D., 92 Steffen, L. S., 196 Steine, E., 443 Stein, P., 460–461, 486–487, 499 Stelzner, K. F., 299 Ste Marie, L., 419 Stephen, D., 196 Stephens, M., 514 Steptoe, A. L., 500, 514–515 Sternlicht, A. L., 484 Stern, P., 198, 374 Steuber-Buchberger, P., 374, 389 Stevens, J. L., 333 Stevenson, L., 251 Stevens, S., 417 Stewart, A. F., 109–110, 112–119, 125–127, 129–130, 136, 147, 186, 190, 196, 250, 430 Stewart, C. L., 536 Stewart, M. D., 145 Stewart, T. A., 244 Stockton, D. W., 303, 310 Stoeckert, C., 476–477, 536 Stottmann, R. W., 329, 333, 342 Stoykova, A., 251 Stratford, I. J., 419 Stratton, M. R., 92, 103 Strauss, M., 186 Stremmel, W., 113, 127, 129 Streu, S., 416 Strickland, S., 485 Strivens, M., 300, 308, 331 Stryke, D., 247, 283 Stuhlmann, H., 252, 287 Stupka, E., 246–247 Su, A. I., 515 Subkhankulova, T., 486 Subramanian, A., 536 Subramani, S., 218 Suck, D., 110 Suda, T., 18–19 Sudhof, T. C., 219 Suemizu, H., 4 Sugiura, K., 488, 490–491, 493, 495–497, 499 Suh, H., 443 Su, H. H., 191, 220, 223, 230, 252 Suh, Y. G., 518 Sui, G., 352, 371 Sukhwani, M., 444 Sullivan, T., 536 Sultan, M., 514 Sundin, K., 529 Sung, H. K., 443 Sung, P., 230 Sun, H., 113–114, 190, 196 Sunkin, S. M., 186, 191, 193 Sun, L., 299 Sun, Q., 246, 250–251 Sunshine, M., 536

565

Author Index

Sun, T., 486 Sunwoo, H., 513 Sunyaev, S., 342 Surani, M. A., 476–477, 486–487, 495–497, 500–501, 506 Suske, G., 517 Suster, M. L., 58, 64 Sutherland, H. G., 275 Su, Y. Q., 488, 490–491, 493, 495–497, 499 Suzuki, H., 196 Svoboda, K., 197 Svoboda, P., 461 Sweeney, S. T., 197 Swift, S., 255 Swindell, R., 419 Swing, D. A., 186 Szulc, J., 446 Szymczak, A. L., 129, 253 T Taira, K., 375–376 Takahashi, J. S., 335, 341 Takahashi, K., 65, 219, 443, 518 Takahashi, M., 6, 38 Takano, M., 459 Takayanagi, K., 249 Takebayashi, H., 167 Takeda, J., 71, 220, 224, 231, 238 Takeda, K., 417 Takeda, N., 130, 249 Takehashi, M., 18–19, 27, 29–31, 33–34 27 Takeuchi, J. K., 416 Takeuchi, T., 172 Talbi, S., 461 Tamayo, P., 536 Tam, O. H., 487 Tamura, M., 39, 343 Tamura, Y., 18–19 Tan, A., 198 Tanabe, Y., 186 Tanaka, T. S., 253, 286–288, 486–487, 512, 519 Tan, E. M., 198 Tang, F., 476–477, 495–497, 500–501, 506 Taniguchi, H., 40 Taniguchi, M., 237 Taniguchi, S., 249 Tanimoto, E. Y., 458–459, 463, 474 Taniuchi, I., 191 Taniwaki, T., 247, 251, 283 Tan, W., 198 Tarantino, L. M., 223 Tate, P., 248, 275 Tavernarakis, N., 54 Taylor, D. F., 514–515 Taylor, D. S., 115, 118 Tegelenbosch, R. A., 20, 26 Teissie, J., 38

Teng, P. I., 438 Teranishi, Y., 198 Tessarollo, L., 186, 405, 517 Tessier-Lavigne, M., 247–249, 253 Testa, G., 113, 115, 126–127, 129 Teucher, M., 125 Theis, M., 352 Thellmann, M., 433 Theunissen, J. W., 333 Thibault, S. T., 94 Thiesen, H. J., 446 Thomas, G. P., 226 Thomas, K. R., 130, 218 Thomas, M. K., 127, 436, 518, 534, 536 Thomason, L. C., 127 Thompson, D. M., 303, 310 Thompson, H. M., 299 Thompson, J. F., 478 Thomson, J. G., 111 Thomson, J. M., 512, 519 Thomson, V. J., 147 Thornton, C., 331 Thorpe, H. M., 110 Threat, T. A., 486 Tibshirani, R., 532 Tichelaar, J. W., 436–437 Ticknor, C., 223 Tilesi, F., 370 Till, J. E., 29 Timmer, J. R., 303, 308, 333 Tiscornia, G., 371 To, C., 247, 271, 277, 283 Todiras, M., 444 Tokunaga, T., 249 Tomashevsky, M., 515, 536 Tom, C., 205 Tonegawa, S., 197, 205 Tong, C., 444 Torres, M., 251 Torres, R., 19 Toth, S., 460 Totoki, Y., 486–487 Toy, K., 486 Toyoda, A., 486–487 Toyoda, M., 459 Toyoda, Y., 127 Toyokuni, S., 18–20, 25–27, 30, 33–34 Toyooka, Y., 286, 518 Toyoshima, M., 18–19 Tran, P. V., 333, 341 Traverso, J. M., 488, 490 Tripoli, G., 246 Trombly, M. I., 220, 223, 230, 252 Tronche, F., 115, 167, 446 Trono, D., 5, 446 Troup, D. B., 515, 536 Tsakiridis, A., 250, 253, 271, 273, 275–277, 279–280, 286–288, 290

566

Author Index

Tsao, N., 247 Tseng, H., 461 Tsien, J. Z., 205 Tsirka, S. E., 485 Tsui, L. C., 331 Tsujita, M., 157, 172 Tucci, V., 335 Tuch, B. B., 476–477, 495–497, 500–501, 506 Turbe-Doan, A., 333 Turck, F., 335 Turek, P. J., 19 Turner, D. L., 353 Tuschl, T., 352, 369, 371, 416 Tusher, V. G., 532 Tutois, S., 331 Tweedie, S., 219, 248 Tzouanacou, E., 250, 253 U Ueda, H. R., 514, 520 Ullmer, C., 433, 437, 446 Umezawa, A., 459 Uno, K. D., 514, 520 Uprichard, S. L., 369 Urasaki, A., 58 Ure, J., 444 Uren, A. G., 65, 92, 100, 102, 233, 260 Urlaub, G., (1986), Urlinger, S., 433 Utikal, J., 443 Utoh, A., 247, 251 V Vaidya, D., 445 Valancius, V., 83 Valencik, M. L., 437 Valenzuela, D. M., 127, 284 Vallis, K. A., 287 Valtieri, M., 255 van Beek, M. E.A. B., 20, 26 Van Buren, D., 447 VanBuren, V., 495, 512, 521 van den Brandt, J., 372, 446 van der Hoeven, F., 113, 127, 129, 190 Van der Valk, M., 186 van der Weyden, L., 113–114, 233, 260 van de Wetering, M., 376, 396–397, 416 van Dongen, S., 343 van Drunen, E., 118 van Duyne, G. D., 110 Van Gelder, R. N., 459, 486, 488 Vanin, E. F., 129, 253 van Leenen, D., 416 van Lohuizen, M., 233, 260 van Loo, P. F., 517 van Pelt, A. M., 30 Van Sloun, P., 112–113, 246–247

van Trigt, L., 198 Varmus, H., 441–442 Vassalli, A., 485 Vassalli, J. D., 485 Vassiliou, G., 91 Vauti, F., 247 Vega, R. G., 515 Vega, V. B., 536 Vegeto, E., 157 Ventura, A., 373, 375 Verhoef, K., 433 Verma, I. M., 4, 6, 9 Verrotti, A. C., 485 Vescovi, A. L., 518 Vetter, D., 147 Vickaryous, N. K., 304, 310 Vickers, T. A., 369 Vierbuchen, T., 219 Vignali, D. A., 129, 253 Vignali, K. M., 129, 253 Vijaykumar, R., 197–198 Vincent, A., 444 Vink, E., 262 Vink, M., 433 Vintersten, K., 9–11, 113, 115, 127, 185–186, 503, 521 Vinuesa, C. G., 331, 335 Viosca, J., 440 Vitale-Cross, L., 436 Vitaterna, M., 335 Vitaterna, M. H., 341 Vizor, L., 300, 303, 308, 331 Vogt, T. F., 415 Vokes, S. A., 335 Volovitch, M., 252 Volsky, D. J., 536 Von Both, I., 416 von Melchner, H., 112, 246–250, 259–260 von Zastrow, M. E., 459, 486, 488 Vooijs, M., 118, 186 Vranizan, K., 536 W Wada, N., 198 Wade, R., 110, 112 Wagner, E. F., 244 Wagner, J., 116 Wagner, S., 19, 304, 310, 334 Wahlsten, D., 335 Wakana, S., 343 Wakasugi, S., 249 Wakefield, M. J., 514–515 Wakenight, P., 251 Waki, K., 500 Walker, E., 285 Walker, J. R., 515 Wallace, H. A., 113 Walls, J. R., 416

Author Index

Walsh, C., 185 Wang, B., 333 Wang, C., 113–114, 190, 196 Wang, D., 517 Wang, G., 486, 536 Wang, H. G., 224, 254 Wang, J., 136 Wang, K., 474 Wang, L. P., 198, 436, 439–441, 518 Wang, M. D., 536 Wang, P., 496 Wang, Q. T., 478, 487–491, 493, 516 Wang, R. H., 416 Wang, W., 60, 63, 77, 94, 96, 220, 223–224, 227–228, 230, 233, 238, 252, 254–255, 258, 262, 273, 288, 443 Wang, X., 220, 223, 227–228, 230, 252, 254–255, 258, 436, 445, 476–477, 486–487, 495–497, 500–501, 506, 512, 519 Wang, Y., 38, 129, 253, 476–477, 495–497, 500–501, 506, 514 Wang, Z., 477–478, 514 Wani, S., 500, 514–515 Wan, L. B., 461 Wanner, B. L., 127 Warrington, J. A., 458–459, 463, 474, 519 Washbourne, R., 300, 308, 331 Washburn, B. K., 127 Wasungu, L., 38 Watanabe, D., 198 Watanabe, H., 40 Watanabe, S., 237 Watanabe, T., 186, 486–487 Waterston, R. H., 72 Watkins-Chow, D. E., 304, 310, 336 Watkins, T., 519 Watson, D. J., 27 Watson, S. J., 496 Watt, A., 251 Watts, T. M., 253 Wawrzyniak, M., 444 Weatherbee, S. D., 333 Weber, J. N., 4 Weber, J. S., 300, 306, 316, 337–338 Weber, K., 352, 416 Weedon, J., 303 Wehr, M., 198, 440 Weible, A. P., 198 Weidenfeld, I., 445–446 Weidlich, S., 113–114, 117, 126, 130, 190, 196, 250 Weigl, D., 127 Weinberg, R. A., 441 Weir, G., 443 Weitzman, M. D., 4 Wells, C., 331 Wells, S., 303, 331 Welman, A., 419

567 Welstead, G. G., 443 Wendling, O., 112–113 Wen, X., 333 Weppert, M., 4 Wernig, M., 219, 443 Wess, J., 197–198 West, A. P., 127 West, J., 437 West, K., 343 West, M. F., 485 Westphal, H., 18, 331 Wetts, R. F., 185 Wheeler, J., 458 White, C. A., 463 Whitehead, K. A., 370 Whitehill, E., 300, 308, 331, 333 Whitelaw, C. B., 4 Whitelaw, E., 304, 310, 446 Whiteson, K. L., 110 Wierzbicki, A., 111 Wiesner, S. M., 438 Wigglesworth, K., 460, 463, 488, 490–491, 493, 495–497, 499–501 Wilhite, S. E., 515, 536 Wilkinson, B. M., 308, 310, 344 Willey, J. C., 519 William, C. M., 186 Williams, B. A., 501, 514–515 Williams, C. J., 461 Williams, J., 116 Williams, M. P., 486 Williams, S. S., 518 Willson, T. A., 304, 310 Wilming, L. G., 308, 344 Wilson, C. H., 233, 260 Wilson, D., 333 Wilson, J., 246, 250–251 Wilson, L. A., 223, 335 Wilson, M. H., 94 Wilson, V., 250, 253 Wiltshire, T., 515 Wilusz, J. E., 513 Winegarden, N., 486, 488 Winer, I., 353 Wisse, L. J., 517 Wiznerowicz, M., 446 Wlodarczyk, B. J., 517 Wold, B., 501, 514–515 Wolf, E., 304, 310, 334 Wolfe, J. H., 27 Wolf, F., 19 Woltjen, K., 65, 443 Wong-Staal, F., 4 Wood, A., 444 Wood, III,W. H., 487, 519 Woodman, P., 303 Woodroofe, C., 110, 129 Woodward, L. P., 308, 344

568

Author Index

Wood, W. G., 113 Woo, Y., 488, 490–491, 493, 495–497, 499 Workman, C. J., 129, 253 Wosczyna, M. N., 189, 192 Woychik, R. P., 221, 305 Wrana, J. L., 416 Wrobel, D., 358 Wuensch, K., 300, 308, 331 Wu, H., 19, 198, 341 Wu, N., 444 Wu, Q., 115, 127, 536 Wurst, W., 112–113, 246–249, 251–252, 259–260, 284, 287, 307, 387 Wu, S. C., 62, 67, 115, 127, 224, 443 Wutz, A., 191, 446 Wu, W., 198 Wu, X., 224, 255, 258 Wu, Y., 113–114, 157, 190, 196 Wyler, M. R., 303, 308, 333 X Xia, H., 353, 371 Xian, X., 488 Xiao, C., 198 Xiao, H., 517 Xia, P. Y., 444 Xia, X. G., 372, 444 Xie, W., 436 Xin, H. B., 246, 250–251 Xin, L., 518, 534, 536 Xuan, Z., 461, 486, 499 Xu, N., 289, 476–477, 495–497, 500–501, 506 Xu, Q., 354 Xu, S., 352 Xu, T., 224, 255, 258 Xu, W., 485 Y Yabuta, Y., 514, 520 Yada, Y., 343 Yagi, R., 518 Yagi, T., 130, 237, 249 Yakar, S., 113–114, 190, 196 Yalcin, A., 352, 416 Yamada, M., 459 Yamada, R. G., 514, 520 Yamagata, K., 26 Yamaguchi, T., 518 Yamaguchi, Y., 198 Yamamoto, A., 441 Yamamoto, M., 189, 192, 198 Yamamoto, S., 189, 192 Yamamoto, T., 130 Yamamura, K., 191, 249, 251 Yamanaka, S., 65, 219, 252, 443, 535 Yamazaki, T., 26 Yancopoulos, G. D., 19, 417

Yan, F., 197–198 Yang, D., 352 Yang, J., 444, 458, 475 Yang, M., 444 Yang, P., 198 Yang, S. H., 4 Yang, W., 186 Yang, X. W., 127, 341, 359 Yang, Y. C., 127, 517 Yant, S. R., 60, 64 Yan, Y., 444 Yao, F., 376 Yasenchak, J., 127 Yasuda, R., 434 Yasuo, T., 518 Yee, A., 247 Yee, D., 221 Yee, S., 518, 534, 536 Ye, L., 437 Yenofsky, R. L., 274 Ye, X., 116, 436 Ying, G., 115, 127 Ying, Q. L., 444 Yoder, B. K., 333 Yokoyama, M., 198 Yong, K. E., 331 Yool, A., 459, 486, 488 Yoo, O. J., 478 Yoshida, K., 130 Yoshida, M., 249 Yoshida, S., 20 Yoshikawa, T., 514 Yoshimoto, M., 18–19 Yoshimura, Y., 459 Yoshimuta, J., 251 Yoshinobu, K., 251 Young, J. Z., 197 Youngman, P. J., 93 Young, S. G., 247, 249 Young, S. R., 333 You, Y., 223 Yu, D., 127, 186 Yue, J., 444 Yu, H. M., 436–437 Yu, J. Y., 65, 333, 353, 371, 374–375 Yu, K., 341 Yu, M., 247 Yumura, Y., 33 Yusa, K., 65, 77, 217, 220, 224, 231, 238 Yu, U., 478 Yu, Y., 113, 115, 246, 250–251, 303, 310 Z Zachgo, J., 245 Zagoraiou, L., 59, 62 Zakhartchenko, V., 4 Zambrowicz, B. P., 165, 191, 223, 247, 249–250, 275, 284

569

Author Index

Zamparini, A. L., 288 Zandstra, P. W., 253 Zarbalis, K., 333 Zariwala, H. A., 186, 191, 193 Zavolan, M., 461 Zayed, H., 61 Zeeberg, B. R., 536 Zeilhofer, H. U., 186 Zeiser, S., 304, 310 Zeitler, B., 444 Zender, L., 445 Zeng, F., 460–461, 463, 478, 486–487, 516 Zeng, H., 186, 191, 193, 248, 446 Zeng, W., 27 Zeng, Y., 371, 375 Zennou, V., 5 Zernicka-Goetz, M., 478, 516 Zervas, M., 159, 185–186 Zevnik, B., 116, 252, 279 Zhang, F., 19, 197–198 Zhang, M. Q., 461, 486, 499 Zhang, W., 536 Zhang, Y. E., 113, 115, 126–127, 129, 136, 186, 517 Zhang, Z., 324, 486 Zheng, B., 300

Zheng, Q. Y., 221 Zhong, J., 514 Zhou, H. H., 198, 444 Zhou, M., 440 Zhou, W., 486 Zhou, X., 433 Zhuang, Y., 224, 255, 258 Zhu, F., 235 Zhu, H., 517 Zhu, J., 171 Zhu, L., 415 Zhu, P., 434 Zhu, Q., 223, 247 Zhu, W., 460 Zijlstra, M., 218 Ziman, M., 519 Zimmermann, F., 433, 437, 446 Zimmermann, J. W., 18, 30–31 Zimmermann, K., 3 Zinyk, D., 186 Zohn, I. E., 303, 310 Zong, H., 161–162, 165, 175, 177, 191 Zsebo, K., 19 Zufferey, R., 4–6 Zwingman, T. A., 186, 191, 193

Subject Index

A Allele design applications of, 112–113 and in silico work, recombineering exon selection, 127–129 lacZ-neo stop cassette, 129 30 loxP site, 129–130 p15A-pTK-DTA-ampR, subcloning vector, 130–131 scheme for, 127–128 Antibody staining, genetic fate mapping, 169–170 Autosomal ENU mutations dominant, 300–301 recessive, 300–301, 303 B Bacterial artificial chromosome (BAC), recombineering, 132 Balancer chromosomes, 302–303 Behavioral phenotypes, 335–336 Bifunctional activating/inactivating transposons, 96–97 Blmtet allele, inducible, 239 Busulfan, 30 C Calcium-calmodulin-dependent kinase IIa (CamKIIa), 439–441 Cancer genetics, transposable elements (TEs), 65–67 research, tetracycline-controlled transcription, 439, 441–443 Cancer gene discovery. See DNA transposons, cancer gene discovery cDNA FGO library genes, 498–499 gene expression profiling, 466–472 generation of, 422 oligo(dT)-primed synthesis, 488 synthesis and probe position, RNA relative quantification, 490 and RNA extraction, 504 Cell culture and gene targeting, 85 Cell-function mapping, 196–198 Cellular phenotypes, 335

Central nervous system (CNS), in vivo electroporation, 39–40 Chicken b-actin chimeric (CAG) promoter, 76–77 Chromosomal engineering, site-specific recombination, 84–87 Clonal analysis experiments, genetic fate mapping ligand, optimal dose of, 176–177 noninduced labeling, 176 and reporter lines, 175 Conditional cre/loxP sh RNA, RMCE vectors generation, 406 Conditional gene knockdown, 445–446 cre-inducible, 405–409 Cre/loxP sh RNA, RMCE vectors generation, 406 dox-inducible, 409–412 Conditional mutagenesis, 113 and Cre zoo, 114–117 Conditional RNAi inducible approaches, 375–376 tissue-specific approaches, 373–375 Conditional tetR/O shRNA vectors cloning, 409–411 generation, 410 pretesting of, 411 stable transfection of, 411–412 testing of, 412 Constitutive gene knockdown, shRNA RMCE method, 402–405 RMCE vectors generation, 400 transgenic mice generation, 405 vectors cloning, 399–402 vectors transient transfection, 402 Cre recombinase conditional knockdown, 405–409 for shRNA expression, 395–396 Cre toxicity, 167–168 Cre zoo and conditional mutagenesis combined spatial-temporal regulation, 116–117 deleters, 115 spatial regulation, 115 temporal regulation, 115–116 cRNA fragmentation and assessment, 472–473 labelling, 472 quality control assessment, 473–474 Cut-and-paste transposons, 57

571

572

Subject Index D

Diaminobenzidine (DAB), 169–170 Digoxigenin-labeled riboprobes synthesis, 200–201 DNA transposons cancer gene discovery bifunctional activating/inactivating transposons, 96–97 gene-activating transposons, 95–96 gene-trapping transposons, 96 idiosyncrasies of, 102–103 inducible transposase expression, 100 insertional mutagenesis, 92–93 insertional mutagens, comparison of, 93 mapping of, 100 mice, generation of, 96 putative target genes validation, 101–102 recurrent integration sites, statistical mining of, 101 selection of, 93–94 tissue-specific mutagenesis, 98–99 whole-body (constitutive) screens, 98 cut-and-paste transposons, 57 genome-wide mutagenesis, forward genetic screens, 224–225 Minos, 58–59 piggyBac (PB), 59–60 Sleeping Beauty (SB), 60–61 Tol2, 56, 58 Dominant screens, 219 Doxycycline (Dox), 419 conditional knockdown, 409–412 for shRNA expression, 396–397 Dre recombinase, 117 E Ectopic tumor formation, Tet-control, 442 Electroporation. See In vivo electroporation, mouse embryos Embryonic stem (ES) cells 3D PCA plots, 535 electroporation, PiggyBac transposon, 258–259 induction trapping and phenotypic screens, 251–252 infection and clone picking, retroviral vectors, 256–257 scatter-plot and log-ratio plot in, 534 Embryonic stem cells (ESC), self-renewal and pluripotency lentivirus-based system design and cloning, shRNA, 360–361 infection, 362–363 lentiviral particles generation, 361–362 pLKO.puro and pLKO.pig vectors, 359–360 quantification, knockdown efficiency, 364

vs. siRNA approach, 358 maintenance, 353–354 siRNA-mediated method automated image scanning, 357–358 GFP expression detection, 355 image acquisition process, 357 quantitative image analysis, 357–358 screening parameters, 354 transfection, 355–356 Embryo production, tetraploid complementation, 87 Endogenous driver gene vs. recombinase, in situ hybridization, 199–204 Enhancer detection, SB knock-in approach, 84 ENU. See N-ethyl-N-nitrosourea (ENU) mutagenesis Exon selection, recombineering, 127–129 Exo utero electroporation microinjection and, 48 repositioning, 48–49 uterine wall, incision of, 48 F False discovery rate (FDR), microarray data, 531–532 FGOs. See Full-grown oocytes (FGOs) Flp recombinase, 114, 190 Fluorescent microscope, SB transgenic approach, 78 Forward genetic screens, ENU mutagenesis screens. See also Genome-wide forward genetic screens, mouse ES cells design, 315–316 execution breeding, 319–320 establishing, independent lines, 322 materials, 317 mutagenesis, 316–317 procedure, 317–319 screening, 321 gene identification causality proving tests, 325–326 cryopreservation, 326 mapping, 323–324 sequencing, 324–325 Full-grown oocytes (FGOs), 496–499, 506 Functional genomics. See Sleeping Beauty (SB) transposons G b-Galactosidase (bgal) activity, X-gal staining, 168–169, 205–207, 263, 334 Gene-activating transposons, 95–96 Gene expression profiling, oocytes and preimplantation embryos experimental design considerations for, 463 linear two-step amplification, 462

Subject Index

MIAME, elements of, 476–477 microarray data analysis methods, 474–475 RNA isolation and cRNA target preparation cRNA fragmentation and gel electrophoresis, 472–473 double-stranded cDNA recovery, 469–472 equipment for, 465 reagents for, 466–467 total RNA and double-stranded cDNA synthesis, 466–468 RNA sequencing (RNA-seq), 476–478 validation methods, 476 Gene knockdown ESC self-renewal and pluripotency lentivirus-based system, 358–364 siRNA-mediated method, 354–358 RNA interference conditional methods, 373–376, 405–412 constitutive methods, shRNA, 399–405 cre recombinase-inducible, 395–396 dox-inducible, 396–397 materials, 398–399 mice generation, 392–393 protocols, 376–382 short hairpin RNA, 370–371 short interfering RNAs, 369–370 shRNA, 390–392, 394–395 transgenic shRNA, 371–373 vectors, genomic integration of, 392–393 tetracycline-inducible shRNA mice, in vivo analysis of cDNA generation, 422 induced shRNA expression determination, 423–426 in vivo doxcycline induction shRNA, 419 mouse tissues, harvesting and storage, 420 RT-PCR expression analysis, 422–423 tissue homogenization and RNA purification, 420–422 Gene ontology (GO) database, 535–536 Gene targeting loxP and FRT sites, 146–147 PGK-neo cassette, 146 Gene therapy, transposable elements (TEs), 64 Genetic fate mapping, SSRs clonal analysis experiments ligand, optimal dose of, 176–177 noninduced labeling, 176 and reporter lines, 175 Cre and Flp recombinase, 156 Cre toxicity, 167–168 GIFM experiments initial population, 174–175 labeling kinetics, 173 ligand administration, 171–173 ligands toxicity, 174 promoters selection and reporter allele, 170–171

573 methods clonal analysis, 163–164 cumulative, 163 GIFM studies, 163 intersectional, 164 modifications to, 156–159 mouse breeding, 168 mutant analysis and interchromosomal recombination, 177–178 intrachromosomal recombination, 177 principles, 154–156 promoters selection expression, 166–167 reporter allele, 164–166 protocol for antibody staining, 169–170 b-galactosidase activity, X-gal staining, 168–169 recombination kinetics of, 167 reporter alleles interchromosomal recombination-MADM, 161–162 intrachromosomal recombination, 159–161 Genetic inducible fate mapping (GIFM) initial population, 174–175 labeling kinetics, 173 ligand administration preparation of, 172 routes of, 172 volumes injected, 172–173 ligands toxicity, 174 promoters selection and reporter allele, 170–171 Gene trap mutagenesis gene discovery and annotation, 285–287 hypothesis-driven screens, 287–288 ordering and handling, 261–263 protocols colony picking, 289 ES cell transfection and expansion, 288–289 splinkerette, 290–291 targeted trapping, 289–290 random integration, mutations generation, 278–279 resources of, ES cell clones, 281–282 high-throughput gene targeting resources, 284 IGTC, 280 Soriano gene trap resource, 283 tools, RMCE vectors, 285 retroviral vectors ES cells infection and clone picking, 256–257 generation of, 255 virus supernatant production, 255–256 splinkerette PCR, insertion sites identification, 259–262 strategies for

574 Gene trap mutagenesis (cont.) alleles, types of, 250–251 induction trapping and phenotypic screens, 251–252 modifications, site-specific recombinases, 250–251 promoter and polyA, 246, 248–250 resources, 247 targeted trapping SA-type vectors, 279 SD-type vectors, 279–280 transposon vectors ES cells electroporation, PiggyBac transposon, 258–259 plasmids generation, 258 vectors for expression-dependent and-independent approaches, 274 NMD, 275–276 plasmid and retroviruses, 254 PolyA trap vectors, 276–277 reporter and selection genes, 253–254 splice acceptor vectors (SA), 273–275 splicing elements, 253 transposon, 254–255 Gene-trapping transposons, 96 Genome-wide forward genetic screens, mouse ES cells BLM-deficient cells, recessive screens homozygote mutants generation, mitotic recombination, 229–231 homozygous mutant library generation, 231–232 tet-off cassette, 231 dominant screens, 219 experimental steps, 220–221 genome-wide mutagenesis insertional mutagens, designs for, 225–226 methods and protocols for, 226–229 mutagen selection, 220–225 mutant validation genes, 235–236 genotype–phenotype causality, 236–238 splinkerette-PCR method, 233–235 recessive screens, 219–220 Genome-wide germline mutagenesis. See Transgenic approach, Sleeping Beauty transposon Genome-wide mutagenesis, forward genetic screens insertional mutagens, designs for, 225–226 methods and protocols for library-complexity assessment, 229 plasmid harboring transposon, transfection of, 227–228 preintegrated transposon, intragenomic mobilization of, 228–229 mutagen selection

Subject Index

chemical agents, 221, 223 comparison of, 222 DNA transposons, 224–225 physical agents, 223 retroviral vectors, 223 Germline mutagenesis, 61–63 Germline stem (GS) cells drug selection, 28 establishment and maintenance of adult testes and, 26 basal medium, 21–22 cell culture, initiation of, 23 cell culture medium, 20 culture appearances, 24 feeder-free culture and suspension culture, 25 multipotent GS (mGS) cells, 26–27 problems in, 25 testis cells, dissociation of, 20, 23 gene transduction, 28 homologous recombination, 29 GFP. See Green fluorescent protein (GFP) Glial cell line-derived neurotrophic factor (GDNF), 18 GO. See Gene ontology (GO) database Green fluorescent protein (GFP), 77–78, 161, 178, 205, 207, 355 I Idiosyncrasies, DNA transposons, 102–103 Induced pluripotent stem cells (iPSC) technology, secondary, 443 Inducible knockdown (iKD) models, 416–418 Induction gene trapping, 251–252 Insertional mutagens cancer gene discovery, 93 genome-wide mutagenesis, 225–226 In silico, DNA cloning program, 131–132 In situ hybridization, recombinase vs. endogenous driver gene digoxigenin-labeled riboprobes synthesis, 200–201 postnatal mice, tissue preparation, 200 prenatal mice, tissue preparation, 199–200 RNA detection on, 202–204 Interchromosomal recombination, 161–162, 177–178 International Gene Trap Consortium (IGTC), 280–281 Intersectional genetic fate mapping cell-function mapping neuromodulation, 198 neuromodulators, 197 pharmaco (chemico)-genetic approaches, 198 cell-subtype selectivity, dual-recombinase Cre and Flp, 190

575

Subject Index

target indicator transgenes, 190–193 in situ hybridization, recombinase vs. endogenous driver gene digoxigenin-labeled riboprobes synthesis, 200–201 postnatal mice, tissue preparation, 200 prenatal mice, tissue preparation, 199–200 RNA detection on, 202–204 in utero, tracer molecule delivery single-vs. dual-recombinase, 187 transgenesis approaches, 186 mouse lines (3) and transgene construction (1), 196 progenitor cells and descendants b-gal, X-gal detection of, 205–207 lineage tracer molecule detection, immunofluorescence, 207–208 reporter molecule, 195 spatiotemporal profile comparison, recombinase expression, 204–205 subtractive and lineage tracer molecules, codetection of, 208 and subtractive populations, 193–195 Intrachromosomal recombination, 159–161, 177 In utero electroporation of embryo, 46 localized transfection of, 46 microinjection, 45 peritoneum, suture of, 47 pulling of, 43–44 repositioning of, 47 In utero, tracer molecule delivery single-vs. dual-recombinase, 187 transgenesis approaches, 186 In vivo doxcycline, shRNA induction, 419 In vivo electroporation, mouse embryos in CNS, 39–40 equipments materials, 40, 42 operating board, 42 standard setup for, 41 tools for, 41 exo utero, 39, 47–49 gene expression analysis, 49 ICR mice, 42 plasmids and siRNA, 42–43 in utero, 39, 43–47 K Knock-in approach, Sleeping Beauty transposon cell culture and gene targeting, 85 embryo production, tetraploid complementation, 87 ES cells, in vitro mobilization of, 85–86 insertion sites determination, 86 local hopping, exploitation of, 82 site-specific recombination, 86–87

vector design enhancer detection, 84 homology arm for targeted recombination, 82–83 selection scheme, 84 site-specific recombination for, 84–85 Knockout (KO) mouse models, 368 L LacZ-neo cassette preparation, recombineering, 136 LacZ-neo stop cassette, 129 Lentivirus, shRNA gene knockdowns design and cloning, shRNA, 360–361 ESC infection, 362–363 lentiviral particles generation, 361–362 pLKO.puro and pLKO.pig vectors, 359–360 quantification, knockdown efficiency, 364 vs. siRNA approach, 358 Lentivirus transgenesis characterization of PCR analysis, 11 quantitative real-time PCR, 12–13 Southern blot analysis, 11–12 generation of denuded embryos, 10–11 embryo culture medium, 9 high-titer lentivector preparations, 6–8 subzonal injection, 10 3rd generation vector construct of, 5–6 technologies for, 5 vectors, 4 Ligands, GIFM administration preparation of, 172 routes of, 172 volumes injected, 172–173 toxicity, 174 Linker ligation-mediated PCR (LM-PCR) oligonucleotide sequence and reaction condition for, 81 procedure for, 78–80 protocol of, 80, 82 Local hopping enhancer detector (LHED) system, 82–83 LoxP and FRT sites, gene targeting, 146–147 LoxP-flanked stop element insertion, 405–407 in vitro deletion, 407–408 30 LoxP site, 129–130 M MADM. See Mosaic analysis with double markers Metabolic phenotypes, 334 Microarray data analysis methods, 474–475

576

Subject Index

Microarray (cont.) experiments, oocytes or preimplantation embryos, 460–461 Microarrays, mouse embryos gene expression profiling data analysis and interpretation functional annotation, 535–536 genes finding and, 532–533 principal component analysis, 533–535 statistical analysis, 531–532 data submission, 536–537 experimental strategies data sets checking, 515 embryonic materials analysis, 515–518 replication, number of, 519–520 methods of mRNA abundance, RNA-seq, 514–515 noncoding/microRNAs and proteins, 513 spatial resolution, 513–514 quality control of results cooperative hybridization issues, 530–531 Rank-Plot method, 526–529 replications, correlation of, 529 signal intensities calibration, 529–530 RNAs, expression profiling cy3-labeled target preparation, 521–524 cy5-labeled target preparation, 524 hybridizations, washing, and scanning, 524, 526 labelled targets, quality control of, 524–525 purification of, 524 Microinjection and exo utero electroporation, 48 and in utero electroporation, 45 Minimal information about a microarray experiment (MIAME), 477, 536–537 Minos transposons, 58–59 Mosaic analysis with double markers (MADM), 161–162, 177–178 Mouse genetics, transposable elements (TEs). See Transposable elements (TEs) Multipotent GS (mGS) cells, 26–27 Multipurpose alleles, 113 Mutagens, forward genetic screens chemical agents, 221, 223 comparison of, 222 DNA transposons, 224–225 physical agents, 223 retroviral vectors, 223 N Neomycin resistance gene (neo), 146 N-ethyl-N-nitrosourea (ENU) mutagenesis action mechanism, 299–300 breeding schemes, mutations recovery autosomal dominant phenotypes, 300–301 autosomal recessive phenotypes, 300–304

chemical structure of, 299 forward genetic screens (see also Forward genetic screens, ENU mutagenesis screens) design, 315–316 execution, 316–322 gene identification, 323–326 materials and methods breeding scheme, 306–307 inactivating method, 306 injection, 305–306 mutation identifying method, 308–309 phenotype mapping, 308 phenotype screening, 307–308 preparation, 304–305 mouse mutagens, history of, 298–299 phenotype-driven analysis mutant ascertainment, 339–340 mutation identification, 340–343 mutation validation, 343–345 screen design, 331–336 treatment, 336–339 Nonsense-mediated mRNA decay (NMD), 275–276 Northern blot protocol, 423–424 Nuclear receptors (nuR), 157 O Oligonucleotides LM-PCR, 81 PCR, recombineering, 133–134 shRNA annealing, 379 OO. See Ovulated oocytes Oocytes and preimplantation embryos biochemical analyses, total RNA decreasing, 483–485 expressed sequences, libraries of, 486–487 gene expression profiling (see Gene expression profiling, oocytes and preimplantation embryos) isoform-specific changes alignment-based analysis, 500–501 read counts for, 500 Spin trancript, 501 large-scale expression, comparative analysis annotated gene content, 496 experimental and computational approaches, 495–496 FGO and OO sequencing datasets, 496 FGO cDNA library genes, 498–499 maturation and development controlling genes identification, 486–487 molecular studies, mRNA fate, 485–486 RNA content, 484 RNA relative quantification artifacts, 493–494 cDNA synthesis and probe position, interacting effects, 490

577

Subject Index

computational and experimental normalization, 493–494 gene expression patterns, 489 normalization to luciferase qPCR, 491 normalization to oocyte, 491 oligo(dT)-primed cDNA synthesis, 488 QADB, 488 qPCR assay, 492 Ovulated oocytes (OO), 483, 496–499, 506 P p15A-pTK-DTA-ampR, subcloning vector, 130–131 PCA. See Principal component analysis (PCA) Phenotype-driven analysis, ENU mutagenesis mutant ascertainment, 339–340 mutation identification, 340–343 mutation validation, 343–345 screen design behavioral phenotypes, 335–336 developmental phenotype, 332–334 general principle, 331–332 metabolic phenotypes, 334 physiological and cellular phenotypes, 335 suppressor/enhancer phenotypes, 336 three-generation strategy, recessive mutations, 330 treatment breeding and mapping, 339 concentration, 337 mouse strain choice, 337–338 percent survival, mice, 338 Phosphoglycerate kinase (PGK) gene, 146 Physiological phenotypes, 335 PiggyBac (PB) transposons ES cells electroporation, 258–259 molecular characteristics of, 59–60 Splinkerette-PCR, 233–235 Plasmids for PCR, 134 and siRNA, 42–43 Pluripotency, ESC lentivirus-based system, 358–364 siRNA-mediated method, 354–358 Pluripotent stem cells, induced, 65 PolyA trapping, 249–250 vectors, 277 Polymerase chain reaction (PCR) lentivirus transgenesis, 11 recombineering cycle settings for, 136 oligonucleotides, 133–134 plasmids for, 135 primer sequences for, 132 purification and yield, 134 reactions, 133

Preimplantation embryos. See Oocytes and preimplantation embryos Principal component analysis (PCA), 533–535 3D plots, 535 singular value decomposition (SVD), 534 Promoter trapping, 246, 248–249 Ptet. See Tet-responsive promoter (Ptet) Putative cancer genes validation, 101–102 Q Quality control of labeled targets, by NanoDrop, 525 Quantitative amplification and dot blotting (QADB), 488 Quantitative real-time PCR, lentivirus transgenesis, 12–13 R Rank-Plot method, microarray data quality control of, 528 steps for, 526–527 Real-time PCR expression analysis, 422–423 protocol, 424–426 transcript abundance evaluation, 503–506 Recessive screens, BLM-deficient cells homozygote mutants generation, mitotic recombination, 229–231 homozygous mutant library generation, 231–232 tet-off cassette, 231 Recombinase-mediated cassette exchange (RMCE), 113, 389 construction, 402–403 DNA preparation, 379–380 positive ES cells clones identification, 404–405 into Rosa26 of ES cells, 403–404 transfection ES cells, 380–381 Recombinases applications of, 112 Cre zoo and conditional mutagenesis, 114–117 Dre recombinase, 117 vs. endogenous driver gene, in situ hybridization digoxigenin-labeled riboprobes synthesis, 200–201 postnatal mice, tissue preparation, 200 prenatal mice, tissue preparation, 199–200 RNA detection on, 202–204 Flp recombinase, 114 issues in, 118–119 tamoxifen administration, in mice and cultured cells, 117–118 Recombinase target sites (RTs), 110–111 Recombination site functionality confirmation, gene targeting materials, 147

578 Recombination site functionality confirmation, gene targeting (cont.) method, 147–148 recombinase-expressing bacteria, analysis of, 148–150 Recombineering, conditional targeting constructs allele design and in silico work exon selection, 127–129 lacZ-neo stop cassette, 129 30 loxP site, 129–130 p15A-pTK-DTA-ampR, subcloning vector, 130–131 scheme for, 128 BAC, 132 background resistance, 136–137 in silico, DNA cloning program, 131–132 lacZ-neo cassette preparation, 134 PCR cycle settings for, 136 oligonucleotides, 133–134 plasmids for, 135 primer sequences for, 132 purification and yield, 134 reactions, 133 plating, 139–140 protocol for, 137–139 sequencing oligos, 139 standard electroporation protocol, 140–141 Recurrent integration sites, 101 Region-specific chromosome engineering. See Knock-in approach, Sleeping Beauty transposon Replication number, for microarray analysis, 519–520 Reporter alleles, genetic fate mapping interchromosomal recombination-MADM, 161–162 intrachromosomal recombination, 159–161 promoters selection, 164–166 Reporter molecule, intersectional genetic fate mapping, 195 Retrotransposons, 55, 72, 93 Retroviral vectors gene trap mutagenesis ES cells infection and clone picking, 256–257 generation of, 255 virus supernatant production, 255–256 genome-wide mutagenesis, 223 RMCE. See Recombinase-mediated cassette exchange (RMCE) RNA detection of, in situ hybridization, 202–204 expression profiling cy3-labeled target preparation, 521–524 cy5-labeled target preparation, 524 microarray hybridizations, 524, 526 purification and quality control of, 524

Subject Index

isolation and cRNA target preparation cRNA fragmentation and gel electrophoresis, 472–473 double-stranded cDNA recovery, 469–472 equipment for, 465 reagents for, 466–467 total RNA and double-stranded cDNA synthesis, 466–468 oocytes and preimplantation embryos, content of, 484 relative quantification, transcriptional silence, 487–493 (see also Oocytes and preimplantation embryos) RNA interference applications, in mice. See also Gene knockdown, RNA interference conditional RNAi inducible approaches, 375–376 tissue-specific approaches, 373–375 doxycycline treatment, 381–382 generation and phenotypic analysis, 382 quantification, 382 RMCE DNA preparation, 379–380 transfection ES cells for, 380–381 short interfering RNAs, 369–370 shRNA design of, 376–377 expression vectors, 370–371 in vivo expression, 371–373 ligation and transformation, 379 oligonucleotides annealing, 379 pINV-7 vector, 377–378 T4 polynucleotide kinase treatment, 378–379 RNA sequencing (RNA-seq), 476–478, 514–515 Rosa26 locus, 389 S Self-renewal and pluripotency, ESC lentivirus-based system, 358–364 siRNA-mediated method, 354–358 Short hairpin RNAs (shRNA) design and cloning, 390–392 expression systems cre recombinase-inducible, 395–396 dox-inducible, 396–397 expression vectors, 370–371 genomic integration, 393 induced expression, determination northern blot protocol, 423–424 real-time PCR protocol, 424–426 in vivo doxcycline induction of, 419 in vivo expression oversaturation, 372 recombinase assisted integration, 372–373 pretest, transient transfection, 402, 411 protocols DNA preparation, RMCE, 379–380

579

Subject Index

doxycycline treatment, 381–382 generation and phenotypic analysis, 382 quantification, 382 shRNA cloning, 377–379 shRNA design, 376–377 transfection ES cells, RMCE, 380–381 stable transfection of, 402–405, 411–412 timescale and principle, of validation, 394–395 transgenic mice generation, 405, 409, 412 Short interfering RNA (siRNA) in vivo, 370 off-target effects, 370 selection, 369–370 Site-specific recombinase (SSR), 250–251. See also Genetic fate mapping, SSRs Site-specific recombination, 84–87, 113 Sleeping Beauty (SB) transposons, 60–61 genome-wide germline mutagenesis, transgenic approach generation and breeding of, 77–78 linker ligation-mediated PCR, 78–82 mutagenesis scheme, 74–76 transposition sites distribution, 74 vector construction, 76–77 vector structures and gene trap scheme, 76 region-specific chromosome engineering, knock-in approach cell culture and gene targeting, 85 embryo production, tetraploid complementation, 87 ES cells, in vitro mobilization of, 85–86 insertion sites determination, 86 local hopping, exploitation of, 82 site-specific recombination, 86–87 vector design, 82–85 transgenic vs. knock-in approaches, 73 Small interfering RNAs (siRNA), gene knockdown automated image scanning, 357–358 GFP expression detection, 355 image acquisition process, 357 quantitative image analysis, 357–358 screening parameters, 354 transfection, 355–356 Southern blot analysis, lentivirus transgenesis, 11–12 Spermatogonial stem cells (SSCs), 18–19. See also Germline stem (GS) cells Spermatogonial transplantation and offspring production donor cell preparation, 29 efferent duct, 31–33 glass needle, 31–32 recipient mice, 33–34 recipient preparation, 30 SSC activity, measurement of, 33 Splice acceptor sequence (SA) bgeo, 273, 275

splice donor (SD) combination, 275–276 Splinkerette-PCR, 100, 259–262, 290–291 piggyBac integration-site isolation, 233–235 sister-clone filtering, 235 Suppressor/enhancer phenotypes, 336 T Tamoxifen administration, in mice and cultured cells, 117–118 tet operator (tetO), 431 Tetracycline-controlled transcription, reversibility binary system integration, 446–447 CaMKIIa-tTA mice, 440–441 in cancer research, 439, 441–443 conditional knockdown, RNAi, 445–446 iPSC technology, secondary, 443 learning and memory controlling genes, 439 regulatory systems schematic presentation, 432 Tet-Off, 432 Tet-On, 432–434 transgenic mice, from repositories distribution, 435 tTA/rtTA genes expressing mouse lines, 436–438 transgenic rats, 444 Tetracycline-inducible shRNA mice cDNA generation, 422 induced expression, 423–426 in vivo doxcycline induction, 419 mouse tissues, harvesting and storage, 420 real-time PCR expression analysis, 422–423 tissue homogenization and RNA purification, 420–422 Tetraploid complementation, embryo production, 87 Tet repressor (TetR), 431 Tet-responsive promoter (Ptet), 431 Tissue homogenization and RNA purification high-throughput, 421–422 low-throughput, 420–421 medium-throughput, 421 Tissue-specific mutagenesis, 98–99 Tol2 transposons, 56, 58 Transcript abundance evaluation, using real-time PCR FGO and OO comparison, 506 oocyte and embryo isolation, 503 quantitative PCR, 504–506 RNA extraction and cDNA synthesis, 504 Transcriptional silence, RNA relative quantification, 487–493. See also Oocytes and preimplantation embryos Transcription factor binding sites (TFBS), 536

580 Transgenic approach, Sleeping Beauty transposon generation and breeding of GFP-positive mutant mice detection, 78 transgenic lines, 77–78 transposon lines, screening for, 77 linker ligation-mediated PCR, 78–82 mutagenesis scheme, 74–76 transposition sites distribution, 74 vector construction, 76–77 vector structures and gene trap scheme, 76 Transgenic mouse lines, CaMKIIa-tTA mice, 440–441 Transposable elements (TEs) DNA transposons cut-and-paste transposons, 57 Minos, 58–59 piggyBac (PB), 59–60 Sleeping Beauty (SB), 60–61 Tol2, 56, 58 in mouse genetics, applications of cancer genetics, 65–67 development of, 67 gene therapy, 64 germline mutagenesis, 61–63 pluripotent stem cells, induced, 65 transgenesis, 63–64 retrotransposons, 55 Transposase expression, inducible, 100 Transposon gene trap vector protocol design, 254–255 electroporation, ES cells, 258–259

Subject Index

plasmids generation, 258 U Universal mouse reference (UMR), 519, 524, 530–531 V Vectors design, SB transposon enhancer detection, 84 homology arm for targeted recombination, 82–83 selection scheme, 84 site-specific recombination for, 84–85 gene trap mutagenesis expression-dependent and-independent approaches, 274 NMD, 275–276 plasmid and retroviruses, 254 PolyA trap vectors, 276–277 reporter and selection genes, 253–254 splice acceptor vectors (SA), 273–275 splicing elements, 253 transposon, 254–255 W Whole-body (constitutive) screens, DNA transposons, 98